Randomization of deep neural networks for telecommunications

Abstract

End-to-end deep neural network (DNN) communication is provided between a first device (104a) and a second device (104b). At a first device, a transmission DNN (106) processes input communication data to generate an output communication signal (118). The first apparatus performs a scrambling DNN operation in response to a predicted transmission of the generated output communication signal not satisfying a white noise interference level, including: selecting neural network scrambling information (NNSI) for reconfiguring a transmission DNN to process input communication data to generate a scrambled output communication signal, the scrambled output communication signal when transmitted satisfies a white noise interference level. The first device transmits a control message comprising an indication of the NNSI and scrambling timing information for directing when the second device reconfigures a corresponding receiving DNN (109) to generate reconstructed communication data. The first device transmits the scrambled output communication signal based on the scrambling timing information.

Description

Background Technology

[0001] Conventional fourth-generation (4G) and fifth-generation (5G) communication systems feature complex transmitter and receiver processing chains with multiple processing components for encoding and modulating input communication data for wireless transmission from a first device and reception and reconstruction by a second device. The complexity of current transmitter and receiver processing chains can be reduced by training machine learning (ML) algorithms, such as deep neural networks (DNNs), to form transmitter (TX) DNN models (TX DNNs) and receiver (RX) DNN models (RX DNNs) capable of providing end-to-end communication (also known as transmit DNNs and receive DNNs). Such transmitter and receiver DNN models can potentially enhance and / or replace conventional transmitter and receiver processing chains. For example, a trained transmit DNN generates transmission waveforms suitable for efficiently overcoming the numerous channel environments, impairments, and interferences (e.g., multipath interference, multiple access interference, narrowband interference) found in current communication systems, thereby further enhancing performance. Furthermore, such a transmit and receive DNN model is well-suited for supporting end-to-end communication systems, where building conventional transmitter and receiver processing chains may be impractical.

[0002] Although the deployment of transmit and receive DNN models in advanced communication systems is being considered, conventional approaches typically do not provide sufficient control over the DNN model to quickly generate outputs that produce whitened physical transmit signals to reduce interference when necessary.

[0003] For example, although a transmission DNN model can be trained to generate an output communication signal for whitening the transmission signal, maintaining a specific level of white noise interference in the transmission signal (e.g., the amount of white noise interference the system can tolerate or a specific power spectral density (PSD) level of the white noise spectrum) may be impractical for different combinations of input communication data that the transmission DNN model can handle.

[0004] There is an opportunity to develop an efficient and dynamic mechanism that can control the transmission DNN model used to adjust the transmission signal, which is whitened during transmission or appears random to neighboring cells (e.g., whitening interference), and / or meets a certain level of white noise interference, while still maintaining the advantages of using transmission and reception DNNs for end-to-end communication. Summary of the Invention

[0005] In a first aspect, this disclosure provides a method performed by a first device communicating with a second device, the method comprising: processing input communication data with a transmitting deep neural network (DNN) to generate an output communication signal for transmission to the second device; performing a scrambling DNN operation in response to a predicted transmission of the generated output communication signal not satisfying a white noise interference level, the scrambling DNN operation further comprising: selecting neural network scrambling information (NNSI) for reconfiguring the transmitting DNN to process the input communication data to generate a scrambled output communication signal that satisfies a white noise interference level when transmitted; transmitting a control message to the second device, the control message indicating the NNSI and scrambling timing information for instructing the second device when to reconfigure the receiving DNN; and transmitting the scrambled output communication signal satisfying the white noise interference level to the second device based on the scrambling timing information.

[0006] In a second aspect, this disclosure provides a method performed by a second device communicating with a first device, the method comprising: receiving from the first device a control message indicating NNSI and scrambling timing information; receiving from the first device a communication signal transmitted according to the scrambling timing information; reconfiguring a receiving DNN of the second device according to the NNSI and the scrambling timing information; processing the received communication signal with the receiving DNN to generate reconstructed communication data represented by the received communication signal; and sending the reconstructed communication data to a data sink of the second device, or sending the reconstructed communication data to one or more upper protocol layers of the protocol stack of the second device.

[0007] A further aspect provides apparatus and systems for implementing the methods of the first and second aspects.

[0008] This method, apparatus, and system offer numerous advantages, including, for example, efficient design and control of the transmit and receive DNN structures capable of maintaining a transmit signal that meets white noise interference levels and reducing interference to other DNN or non-DNN receivers in the cell or area surrounding the first and second devices. Scrambling DNN operations on the transmit DNN of the first device maintain the white noise interference level of the transmit signal from the first device without generating unexpected transmission spikes due to various combinations of input communication data processed by the transmit DNN. Scrambling DNN operations mitigate, reduce, and / or prevent transmission spikes in the transmit signal of the first device when using the transmit DNN, while meeting the white noise interference level. Further advantages include efficient control of the reconfiguration of the receive DNN of the second device due to corresponding scrambling DNN operations on the transmit DNN of the first device. The transmit and receive DNNs of the first and second devices can be reconfigured efficiently, quickly, and dynamically in real time according to the scrambling DNN operations to change the white noise interference level of the transmit signal generated using the output communication signal of the transmit DNN, while maintaining the transmission power or bit / symbol error rate of the signal of interest. Further advantages include efficient synchronization between a first device using a transmitting DNN and a second device using a corresponding receiving DNN, enabling dynamic whitening of the transmitted signal from the first device and enabling the second device to receive and decode the dynamically whitened transmitted signal. Attached Figure Description

[0009] This disclosure will be better understood by referring to the accompanying drawings, and many features and advantages of this disclosure will be apparent to those skilled in the art. The same reference numerals are used in different drawings to indicate similar or identical items. Embodiments of the invention will be described by way of example with reference to the following drawings, wherein:

[0010] Figure 1 This is a schematic diagram illustrating a comparison between an example conventional transmitter and receiver structure according to some embodiments and an example end-to-end communication transmitter and receiver structure using a deep neural network.

[0011] Figure 2a This is a schematic diagram illustrating an example scrambled DNN communication system according to some embodiments;

[0012] Figure 2b This is a schematic diagram illustrating an example power spectral density of a transmitted signal that meets a white noise interference level according to some embodiments;

[0013] Figure 2c This is a schematic diagram illustrating another example power spectral density of a transmitted signal having transmission spikes that do not meet the white noise interference level, according to some embodiments;

[0014] Figure 2d This is a schematic diagram illustrating a further example power spectral density of another transmitted signal exceeding the level of white noise interference according to some embodiments;

[0015] Figure 2e This is a schematic diagram illustrating a further example power spectral density of another transmitted signal that meets the white noise interference level according to some embodiments;

[0016] Figure 3a This is a schematic diagram illustrating an example input-based scrambling configuration for transmitting and receiving a DNN according to some embodiments;

[0017] Figure 3b This is a schematic diagram illustrating an example output-based scrambling configuration for transmitting and receiving a DNN according to some embodiments;

[0018] Figure 3c This is a schematic diagram illustrating an example hidden layer-based scrambling configuration for transmitting and receiving a DNN according to some embodiments;

[0019] Figure 4a This is a flowchart illustrating an example DNN scrambling process according to some embodiments for generating an output communication signal from a transmitting DNN that satisfies a white noise interference level during transmission;

[0020] Figure 4b This is a flowchart illustrating an example process according to some embodiments for analyzing whether the spectral density of a transmitted signal representing an output communication signal from a transmitting DNN satisfies a white noise interference level;

[0021] Figure 4c This is a flowchart illustrating an example process for receiving one or more control messages, including neural network scrambling information, at a second device according to some embodiments;

[0022] Figure 4d This is a flowchart illustrating an example process for receiving the transmission of output communication signals from a transmission DNN of a first device at a second device and reconstructing communication data, according to some embodiments.

[0023] Figure 5 This is a schematic diagram illustrating an example first device transmitter with a transmission buffer and an example second device receiver with a reception buffer according to some embodiments;

[0024] Figure 6a This is a schematic diagram illustrating example random permutations and random inverse permutations (or depermutations) for scrambling one or more neural network layers of a transmitting DNN according to some embodiments;

[0025] Figure 6bThis is a schematic diagram illustrating an example random permutation sequence starting from an initial seed according to some embodiments;

[0026] Figure 6c This is a schematic diagram illustrating an example permutation matrix generated from a selected i-th random permutation sequence for scrambling / descrambling one or more neural network layers of a transmitting / receiving DNN, according to some embodiments;

[0027] Figure 6d This is a flowchart illustrating an example iterative process according to some embodiments for selecting the i-th random permutation sequence to randomize the order of neural network nodes of one or more neural network layers of a transmitting DNN such that the transmission of the output communication signal satisfies the white noise interference level.

[0028] Figure 7 This is a signal flowchart illustrating the enabling and disabling of a scrambled DNN communication session between a first device and a second device according to some embodiments;

[0029] Figure 8 This illustrates, according to some embodiments, in Figure 7 The diagram shows a signal flow chart of an example DNN scrambling communication between the first and second devices during a DNN communication session.

[0030] Figure 9 This illustrates, according to some embodiments, in Figure 7 The diagram shows another example of DNN scrambling communication between the first and second devices during a DNN communication session.

[0031] Figure 10 This is a signal flow diagram illustrating a scrambled DNN communication session between a first device, a second device, and a third device according to some embodiments;

[0032] Figure 11 This illustrates, according to some embodiments, in Figure 7 or Figure 10 The diagram shows a signal flow chart of an example DNN scrambling communication between the first, second, and third devices during a DNN communication session.

[0033] Figure 12 This is a signal flow diagram illustrating the enabling and disabling of uplink (UL) / downlink (DL) scrambling DNN communication sessions between a base station and a user equipment according to some embodiments;

[0034] Figure 13 This illustrates, according to some embodiments, in Figure 12 A signal flow diagram of an example DL DNN scrambling communication between a base station and a user equipment during a UL / DL DNN communication session;

[0035] Figure 14 This illustrates, according to some embodiments, in Figure 12 A signal flow diagram of an example UL DNN scrambled communication between a user equipment and a base station during a UL / DL DNN communication session;

[0036] Figure 15 This illustrates, according to some embodiments, in Figure 12 Another example of UL DNN scrambled communication between a user equipment and a base station during a UL / DL DNN communication session;

[0037] Figure 16 This is a signal flow diagram illustrating another example DNN scrambling communication session between a base station and a user equipment according to some embodiments;

[0038] Figure 17 This is a schematic diagram of an example computer-readable medium according to some embodiments. Detailed Implementation

[0039] Figure 1 A comparison is shown between an example conventional transmitter and receiver structure for a first conventional communication device 102a and a second conventional communication device 102b, and an example end-to-end communication transmitter and receiver structure using a transmit DNN structure 106 and a receive DNN structure 108 (TX DNN and RX DNN) for a first deep neural network (DNN) device 104a and a second deep neural network (DNN) device 104b (also referred to herein as first device 104a and second device 104b), respectively. Conventional 4G and 5G communication systems have complex transmitter and receiver processing chains with multiple processing components for encoding and modulating input communication data 101a for wireless transmission from the first conventional device 102a and for reception by the second conventional communication device 102b as reconstructed communication data 101b.

[0040] In this example, the transmitter processing chain of the first conventional communication device 102a processes input communication data 101a (e.g., bit blocks, bit streams, or other digital data) from a data source (not shown) for transmission from the first conventional communication device 102a to the second conventional communication device 102b. The transmitter processing chain includes an arrangement of processing blocks such as coding blocks, interleaving blocks, scrambling blocks, precoding blocks, and modulation blocks, which process the input communication data 101a into an output communication signal 118 for transmission. The first communication device 102a includes a radio frequency (RF) front end comprising an RF analog transmitter / transmitter (TX) assembly (RF analog TX) 103a for processing the output data signal (e.g., digital-to-analog conversion and / or RF up-conversion, etc.) for RF up-conversion and transmission via an antenna as a transmission signal 105a on the communication channel 105. The second conventional communication device 102b receives the transmission signal 105a. The second communication device 102b includes an RF front-end comprising an RF analog receiver / receiver (RX) component (RF analog RX) 103b for receiving a transmitted signal 105a (e.g., analog-to-digital conversion and / or down-conversion to baseband), and a receiver processing chain generating reconstructed communication data 101b from the received transmitted signal 105a. The receiver processing chain includes an arrangement of processing blocks such as demodulation, descrambling, deinterleaving, and decoding components for processing the received transmitted signal 105a and restoring the input communication data 101a to the reconstructed communication data 101b. As communication systems evolve (e.g., from 5G to 6G communication standards) to provide higher capacity and lower latency, and with the convergence of various technologies, the already complex transmitter and receiver processing chains will undergo various updates and modifications. This will result in tightly controlled, complex transmitter and receiver processing chains.

[0041] The complexity of current transmitter and receiver processing chains can be reduced by training machine learning (ML) algorithms, such as deep neural networks (DNNs), to integrate transmitter and receiver chain processing blocks into transmit DNN and receive DNN structures 106 and 108 (also referred to herein as transmitter DNN structure and receiver DNN structure, respectively), capable of providing end-to-end communication. Such transmit and receive DNN structures 106 and 108 potentially enhance and / or replace conventional transmitter and receiver processing chains used in the first conventional communication device 102a and the second conventional communication device 102b. For example, transmit DNN structure 106 includes a transmit DNN model 107 (TX DNN) trained to replace the transmitter processing chain comprising encoding, interleaving, scrambling, and modulation blocks of the first conventional device 102a. The first device 104a configures a transmission DNN structure 106 to process input communication data 101a' (e.g., blocks of input communication data / bits) and generates an output communication signal 118 for transmission as a transmission signal 105a' on the communication channel 105' via an RF-simulated TX 103a'. In this case, the transmission DNN model 107 of the transmission DNN structure 106 generates the output communication signal 118 during training. The RF-simulated TX 103a' processes the output communication signal 118 for transmission as a transmission signal 105a' on the communication channel 105'. The resulting transmission signal 105a' generated from the output communication signal 118 of the transmission DNN model 107 is a transmission waveform that, depending on the training / conditions, is suitable for efficiently addressing the numerous channel environments, impairments, and interferences (e.g., multipath interference, multiple access interference, narrowband interference) found in current communication systems.

[0042] The second device 104b receives a transmission signal 105a' via an RF analog RX 103b', which processes the received transmission signal 105a' (e.g., performs at least down-conversion) into a baseband receive communication signal 119 for input to a receive DNN structure 108. The receive DNN structure 108 includes a receive DNN model 109 (RX DNN) trained to generate reconstructed communication data 101b' when given a properly formatted baseband (or down-converted) input data signal. The second device 104b configures the receive DNN model 109 to perform the inverse operation of the transmission DNN model 107 to generate reconstructed communication data 101b' representing the input communication data 101a' input to the transmission DNN model 107.

[0043] In the example, the second device 104b has a protocol stack with multiple protocol layers. After generating reconstructed communication data 101b' for a specific time slot or for one or more time slots, the second device 104b sends the reconstructed communication data 101b' for each of the one or more time slots to one or more upper-layer protocols of the protocol stack of the second device 104b. In the protocol stack of the second device 104b, lower layers are responsible for providing services to upper layers, and the upper layers use those services to provide their own functionality. For example, the reconstructed communication data 101b' is generated at the physical layer of the protocol stack and is passed up and processed by each upper layer until the application layer of the protocol stack, where the corresponding reconstructed communication data 101b' is used, but not limited to, displaying to a user, further processing, and / or sending to one or more applications of the second device 104b for further processing and / or consumption of the reconstructed communication data 101b'.

[0044] In this example, the receive DNN model 109 replaces the receiver processing chain during training. This receiver processing chain includes, for example, demodulation, descrambling, deinterleaving, and decoding blocks to generate reconstructed communication data 101b' from the received transmitted signal 105a'. The reconstructed communication data 101b' represents the input communication data 101a'. The transmit and receive DNN models 107 and 109 are well-suited for supporting end-to-end communication systems where building conventional transmitter and receiver processing chains may be impractical. As the complexity and requirements of transmitter and receiver chains increase, the transmit DNN structure 106 and the receive DNN structure 108 will become key components of 5G advanced or even 6G and higher-level communication systems.

[0045] Although the transmit DNN model 107 is described as performing the functions of the coded blocks, interleaved blocks, scrambling blocks, and modulation blocks of the transmitter chain, this is merely illustrative and not limited thereto. Those skilled in the art will understand that the transmit DNN model 107 is trained to perform any one or more functions of the transmitter processing chain, including encoding, interleaving, scrambling, precoding, modulation, and the like, combinations thereof, modifications thereof, and / or at least one or more of these as required by the application. Although the receive DNN model 109 is described as being trained to perform the functions of the receiver processing chain, including demodulation, descrambling, deinterleaving, and decoding, this is merely illustrative and not limited thereto. Those skilled in the art will understand that the receive DNN model 109 is trained to perform one or more functions of the receiver processing chain, including demodulation, descrambling, deinterleaving, decoding, and / or any other receiver processing chain functions and the like, combinations thereof, modifications thereof, and / or at least one or more of these as required by the application. For example, the transmit DNN model 107 is trained to perform many or most (if not all) functions of the transmit processing chain, excluding the modulation block and the RF analog TX 103a', while the corresponding receive DNN model is trained to perform most (if not all) functions of the receiver processing chain, excluding the RF analog RX 103b' and the demodulation block.

[0046] As described herein, each of the corresponding transmit DNN models 107 and receive DNN models 109 of the transmit DNN structure 106 and receive 108 has been trained to respectively replace the functions of the conventional transmit / receive chain, and / or to overcome various channel conditions, and / or meet the performance requirements of 5G, 6G and future communication standards and the like. The transmit DNN models 107 and 109 can be deployed to configure the transmit DNN structure 106 and receive DNN structure 108 respectively for performing DNN communication between the first device 104a and the second device 104b. For example, the first DNN model 107 includes at least an input neural network layer, one or more hidden neural network layers, and an output neural network layer. The first DNN model 107 processes the input communication data and generates an output communication signal for processing by the RF analog TX 103a' (e.g., digital-to-analog conversion and RF up-conversion, etc.) and transmitted as a transmission signal 105a'. For example, the second DNN model 109 includes at least an input neural network layer, one or more hidden neural network layers, and an output neural network layer. RF analog RX 103b' processes the transmitted signal 105a' (e.g., analog-to-digital conversion and / or down-conversion to baseband) to generate the received communication signal 119 for input to a second DNN model 109, which processes the received communication signal to generate reconstructed communication data representing the input communication data incorporated in the transmitted signal 105a'. In another example, a pair of first DNN models 107 and / or second DNN models 109 may be selected by a first device 104a depending on the communication performance requirements of the DNN communication session with the second device 104b and the like, and the first device 104a may transmit the selection of the first DNN model and / or the second DNN model to the second device 104b during the establishment of the DNN communication session.

[0047] The ML model algorithms / architectures used to train the transmitting DNN model 107 and receiving DNN model 109 as described herein are based on or include (by example only, but not limited to) one or more of the following: neural networks, fully connected neural networks, convolutional neural networks, long short-term memory (LSTM) neural networks, and Transformer neural networks, and / or any other suitable DNN architecture, combinations thereof, modifications thereof (as described herein), and / or supervised and / or unsupervised training may be performed according to application requirements. For example, supervised training of the first DNN model 107 and the second DNN model 109 used by the first device 104a and the second device 104b may use, for example, gradient-based backpropagation techniques to update, for example, the nodes of the corresponding neural network layers of the first DNN model 107 and / or the second DNN model 109 and the weights / parameters of any other DNN architecture components using, for example, an appropriate or suitable loss function. Assumptions as referenced Figures 3a to 3c The various DNN models and / or architectures, as well as the DNN model / structure arrangements, described for the first and second devices have been trained and determined. Multiple different first DNN models 107 and / or second DNN models 109 for different communication scenarios can be stored in a storage device accessible to the first device 104a and / or the second device 104b and / or mapped to appropriate identifiers / indexes in that storage device for retrieval when establishing DNN communication between the first device 104a and the second device 104b to configure the first DNN structure 106 and the second DNN structure 108 of the first device 104a and the second device 104b.

[0048] While replacing one or more functions of conventional transmit and receive chains with transmit DNN structure 106 and receive DNN structure 108 offers significant advantages, several considerations exist when implementing transmit DNN structure 106 and receive DNN structure 108 to meet the performance requirements of 5G, 6G, and future communication standards. For example, rapidly generating whitened physical transmit signals with reduced interference or meeting a certain level of white noise interference is beneficial for increasing capacity and reducing latency in communication systems. Currently, the scrambling / descrambling blocks of conventional 4G / 5G transmitter and receiver chains are controlled separately to rapidly whiten physical transmit signals. However, this functionality is difficult to incorporate into current transmit and receive DNN structures. Although the transmission DNN model 107 and the reception DNN model 109 of the transmission DNN structure 106 and the reception DNN structure 108 can be trained together (e.g., jointly) to generate an output communication signal 118 that, when up-converted and transmitted as a transmission signal 105a', satisfies a certain level of white noise interference (e.g., the amount of white noise interference tolerated by the system or a specific PSD level of the tolerated white noise spectrum), this specific level of white noise interference may not be maintained for subsequent input communication data blocks due to variations in the input bitstream.

[0049] Furthermore, different receivers in a cell or area can tolerate different levels of interference, where the transmitter chain of the first device 104a dynamically adjusts the white noise interference level of the transmitted signal. For the transmit DNN structure 106, which has replaced the conventional transmitter processing chain, the remaining components, such as, for example, the RF-simulated TX 103a', dynamically adjust the final transmitted signal to meet the white noise interference level. However, using the RF-simulated TX 103a' results in a coarse adjustment of the transmission power and carries the risk of increasing the symbol or bit error rate on the communication link, while subsequently reducing throughput due to increased retransmissions, for example, between the first device 104a and the second device 104b. Another possible approach is to train multiple transmit DNN models (and corresponding receive DNN models), each configured to generate an output communication signal 118 that satisfies different white noise interference levels for the same input communication data 101a' when transmitted by the RF-simulated TX 103a'. The first device 104a selects to generate a transmission DNN model (and a corresponding reception DNN model) for an output communication signal 118 with the lowest white noise interference level predicted for transmission. Since the number of transmission DNN models and corresponding reception DNN models is enormous, it is impossible to satisfy all possible white noise interference levels, thus making this infeasible. This is also an inefficient and impractical use of computational resources at both the first device 104a and the second device 104b.

[0050] Multiple different transmit DNN models (and corresponding receive DNN models) are trained to whiten the transmit signal 105a' to satisfy different levels of white noise interference. Depending on the white noise interference level, transmit and receive DNN pairs are selected for use by the transmit DNN structure 106 and receive DNN structure 108 of the first device 104a and the second device 104b, respectively. This is a resource-intensive process requiring significant computational, storage, and transmission resources to reliably satisfy all types of interference and white noise interference levels. This means that the first and second devices have multiple transmit and receive DNN structures stored thereon for retrieval.

[0051] The aforementioned problem is addressed by incorporating scrambling operations into the transmission DNN structure 106 of the first device 104a, which controls the generation of an output communication signal 118 from input communication data 101a'. This output communication signal, when processed and transmitted by the RF analog TX 103a', generates a transmission signal 105a' that satisfies a white noise interference level. The receive DNN structure 108 of the second device 104a uses the inverse operation of the transmission DNN structure 106 to perform descrambling to generate reconstructed communication data 101b'. The scrambling DNN operation can be controlled using Neural Network Scrambling Information (NNSI). The first device 104a selects an NNSI from a set of NNSIs, each NNSI being associated with a different white noise interference level. For example, each NNSI describes the type of scrambling and / or where the scrambling occurs within one or more neural network layers of the transmission DNN model 107 of the transmission DNN structure 106. When the first device 104a predicts that the obtained transmission signal 105a' meets a specific white noise interference level, the first device 104a selects an NNSI from the set of NNSIs to scramble the output communication signal 118. Before transmitting the corresponding output communication signal 118 as the transmission signal 105a', the first device 104a transmits the selected NNSI to the second device 104b. This allows the second device 104b to reconfigure the receiving DNN structure 108 to generate reconstructed communication data 104b' corresponding to the input communication data 101a' represented by the transmission signal 105a'.

[0052] For example, when the first device 104a communicates with the second device 104b, the transmission DNN structure 106 of the first device 104a uses the transmission DNN model 107 of the transmission DNN structure 106 to process the input communication data 101a' to generate an output communication signal 118 for transmission to the second device 104b. When the first device 104a analyzes the generated output communication signal 118 and estimates or predicts that the resulting transmission will not meet the white noise interference level, in response, the first device 104a and the second device 104b perform a scrambling DNN operation. The scrambling DNN operation at the first device 104a includes the first device 104a selecting an NNSI to reconfigure the transmission DNN model 107 of the transmission DNN structure 106 to process the input communication data 101a' and generate a scrambled output communication signal 118 that meets the white noise interference level when transmitted as a transmission signal 105a'. The first device 104a transmits the selected NNSI to the second device 104b in a control message. The control message includes indications for NNSI and scrambling timing information. The scrambling timing information guides the second device 104b on when to reconfigure the receive DNN model 109 of the receive DNN structure 108 to generate reconstructed communication data 101b' when the receive DNN model receives a transmission signal 105a' corresponding to the scrambled output communication signal 118. At the appropriate time, based on the scrambling timing information, the first device 104a processes the scrambled output communication signal 118 into a transmission signal 105a' to the second device 104b via an RF analog TX component, wherein the transmission of the scrambled output communication signal 118 satisfies a white noise interference level.

[0053] The scrambling DNN operation at the second device 104b includes the second device 104b receiving a control message indicating the NNSI and the corresponding scrambling timing information. The RF analog RX 103b' of the second device 104b receives the transmission signal 105a' from the communication channel 105 from the first device according to the scrambling timing information, and outputs the received communication signal 119 for processing by the receiving DNN model 109 of the second device 104b. Before processing the received communication signal 119, the second device 104b reconfigures the receiving DNN model 109 using the received NNSI and the associated scrambling timing information. After reconfiguration, the receiving DNN model 109 processes the received communication signal 119 and generates reconstructed communication data 101b' represented by the received communication signal 119. The second device 104b sends the reconstructed communication data 101b' to its data sink, or sends the reconstructed communication data 101b' to one or more upper protocol layers of the protocol stack of the second device 104b (e.g., to the application protocol layer of the protocol stack for use by one or more applications executing on the second device 104b). The scrambling DNN operation continues when the transmission signal 105a' transmitted by the first device 104a meets the white noise interference level and / or when additional input communication data for transmission is present.

[0054] During a communication session, when the first device 104a performs multiple scrambling DNN operations, it selects different NNSIs because different input communication data 101a' causes the transmission DNN model 107 to generate an output communication signal 118 that will not meet the current white noise interference level when transmitted by RF-simulated TX 103a' as a transmission signal 105a'. The first device 104a sends the different selected NNSIs and associated scrambling timing information to the second device 104b in a control message for use in the corresponding scrambling DNN operation. When the white noise interference level changes, the first device 104a selects different NNSIs, and therefore, the victim device may experience intolerable interference from the first device 104a, or the first device 104a may receive a request to adjust the white noise interference level to a tolerable level.

[0055] The scrambling DNN operation performed by the first device 104a and the second device 104b provides the following advantages: the first device 104a does not transmit the output communication signal 118 of the transmission DNN until the obtained transmission signal meets the white noise interference level set by the first device 104a. This means that the transmission of the first device 104a will meet the white noise interference level without transmission spikes interfering with neighboring devices. The scrambling DNN operation also synchronizes the transmission DNN structure 106 and the receiving DNN structure 108 to operate together to recover the input communication data 101a' at the second device 104b.

[0056] The first device 104a and the second device 104b can be any type of communication device used in the communication system 100, such as, but not limited to, any combination of radio access network (RAN) elements including, for example, base stations (BS), network devices, user equipment (UE), or other RAN elements within the communication system 100. For example, the first device 104a and the second device 104b can be two BS, or two UE, or BS and UE, or UE and BS, or any other combination of communication devices according to application requirements. Figure 2a A scrambled DNN communication system 200 is shown, wherein the first device 210 and the second device 220 are the BS and the UE, respectively.

[0057] Figure 2a An example scrambled DNN communication system 200 is illustrated, in which a first device 210 communicates with a second device 220. In this case, the first device 210 is a BS (Browser Base Station) and the second device 220 is a UE (User Equipment). The BS 210 is connected to the core network (not shown) of the scrambled DNN communication system 200 via one or more interfaces. For example, the communication system could be a 5G / 6G or New Radio (NR) communication system. The UE 220 and BS 210 communicate on wireless communication channel 205 via downlink transmission signal 205a (e.g., downlink transmission) and uplink transmission signal 205b (e.g., uplink transmission). The wireless communication channel 205 may include: a downlink communication channel (e.g., a physical downlink shared channel (PDSCH)) on which the BS 210 transmits downlink transmission signals 205a to the UE 220; and an uplink communication channel (e.g., a physical uplink shared channel (PUSCH)) on which the UE 220 transmits uplink transmission signals 205b to the BS 210. The downlink communication channel may also include a downlink control channel (e.g., a physical downlink control channel (PDCCH)), and the uplink communication channel may also include an uplink control channel (e.g., a physical uplink control channel (PUCCH)).

[0058] BS 210 can be implemented as a computing system / device for performing any of the corresponding methods, scrambling DNN operations, scrambling / randomization / descrambling operations or processes described herein, and / or for implementing any of the corresponding systems, units, and / or devices as described herein. BS 210 includes an RF front-end 203a / b (including RF analog TX and RX components / antennas, etc.), one or more transceivers 211, one or more processors 212, and a memory unit 213 connected together. Those skilled in the art will understand that other types of computing devices / systems / platforms can alternatively be used to implement BS 210 and the methods described herein, such as distributed computing systems depending on application requirements. BS 210 includes one or more processors 212. The one or more processors 212 control the operation of other components of BS 210 (such as the RF front-end 203a / b, one or more transceivers 211, memory unit 213, and the like). The one or more processors 212 can be single-core or multi-core devices. One or more processors 212 may include a central processing unit (CPU), one or more CPUs, a graphics processing unit (GPU), and / or one or more GPUs and the like. Alternatively, one or more processors 212 may include dedicated processing hardware, such as a reduced instruction set computer (RISC) processor or programmable hardware with embedded firmware. Multiple processors may be included in BS 210. In some embodiments, one or more processors 212 may be part of a distributed computing system such as a cloud computing system and / or a cloud computing platform.

[0059] One or more processors 212 of BS 210 may be connected to a network interface, such as a transceiver 211 including, for example, a transmitter (TX) and a receiver (RX), for communicating via RF front-end 203a / b on the network's wireless communication channel 205 with other devices and systems, other communication devices, network devices, RAN entities or devices, operators, and / or any other devices, services, systems, and / or devices as required by the application. One or more processors 212 may optionally be connected to a user interface (UI) for user or operator input to instruct or use BS 210 and / or the underlying computing system and / or for outputting data from it. One or more processors 212 may optionally be connected to a display for displaying output to a user or operator.

[0060] BS 210 includes a memory system or memory cell 213, which includes working or volatile memory. One or more processors 212 can access the volatile memory to process data and can control the storage of data in the memory. The volatile memory can include any type of random access memory (RAM), such as static RAM (SRAM), dynamic RAM (DRAM), or the volatile memory can include flash memory, such as a secure digital card (SD). In some embodiments, memory cell 213 and / or one or more volatile memories can include multiple of a plurality of memories forming part of a distributed computing system such as a cloud computing system and / or a cloud computing platform and the like. BS 210 also includes non-volatile memory. The non-volatile memory may store a set of operating system instructions for controlling the operation of processor 212 in the form of computer-readable instructions and / or store software instructions in the form of computer-readable instructions that, when executed on one or more processors, cause the processors to implement methods, procedures, operations, and / or functions such as scrambling DNN operations, scrambling / randomization operations, procedures, and / or methods as described herein. Non-volatile memory can be any type of memory, such as read-only memory (ROM), flash memory, SD drive, magnetic drive memory, or disk drive memory, and similar types depending on application requirements. In some embodiments, non-volatile memory may include multiple non-volatile memories forming part of a distributed computing system such as a cloud computing system and / or a cloud computing platform and the like.

[0061] The non-volatile memory of memory unit 213 of BS 210 includes computer program code and / or instructions for implementing BS DNN controller (DNNC) 214 and / or BS downlink transmit DNN structure (BS DL TX DNN) 206 or BS uplink receive DNN structure (BS ULRX DNN) 208. When executed on one or more processors 212, BS DNNC 214 uses BS DL TX DNN 206 and / or BS UL RX DNN 208, BS neural network scrambling information (NNSI) storage / table / buffer 215a stored in memory unit 213, and BS transmit / receive (TX / RX) DNN storage or table 215b (also referred to as BS TX DNN / RX DNN storage or table 215b) stored in memory unit 213 to control scrambling DNN operations between BS 210 and UE 220. Although BS DNNC 214 is shown as part of memory cell 213, this is merely an example, and BS DNNC 214 is not limited thereto. Those skilled in the art will understand that BS DNNC 214 can be implemented in the hardware and / or software of BS 210 according to application requirements.

[0062] At least one processor 212, together with at least one memory unit 213 and computer program code or instructions stored thereon, is arranged to cause the computing system of BS 210 to perform at least, for example, as per [the relevant information]. Figures 1 to 17 The diagrams, flowcharts, or operations described in any of the above and their related features disclose at least the corresponding operations, methods, and / or processes.

[0063] Similarly, UE 220 can be implemented as a computing system / device for performing any of the corresponding methods, scrambling DNN operations, scrambling / randomization / descrambling operations or processes described herein, and / or for implementing any of the corresponding systems, units, and / or devices as described herein. UE 220 includes an RF front-end 203a / b, one or more transceivers 221, one or more processors 222, and a memory unit 223 connected together. Those skilled in the art will understand that other types of computing devices / systems / platforms can alternatively be used to implement UE 220 and the methods described herein. UE 220 includes one or more processors 222 (e.g., CPU). The one or more processors 222 control the operation of other components of UE 220, such as the RF front-end 203a / b, one or more transceivers 221, memory unit 223, and the like. The one or more processors 222 can be a single-core device or a multi-core device. The one or more processors 222 can include a CPU and / or a GPU. Alternatively, the one or more processors 222 can include dedicated processing hardware, such as a RISC processor or programmable hardware with embedded firmware. Multiple processors can be included in UE 220.

[0064] One or more processors 222 of UE 220 may be connected to a network interface, such as a transceiver 221 including, for example, a transmitter (TX) and a receiver (RX), for communicating via RF front-end 203a / b on the network's wireless communication channel 205 with other devices and systems such as BS 210, other communication devices, network devices, RAN entities or devices, users or operators, and / or any other devices, services, systems, and / or devices. One or more processors 222 may optionally be connected to a UI for user input to indicate or use UE 220 and / or the underlying computing system and / or for outputting data from it. One or more processors 222 may optionally be connected to a display for displaying output to the user.

[0065] UE 220 includes a memory system or memory unit 223, which includes working or volatile memory. One or more processors 222 can access the volatile memory to process data and can control the storage of data in the memory. The volatile memory can include any type of RAM, such as SRAM, DRAM, or it can include flash memory, such as an SD card. UE 220 also includes non-volatile memory. The non-volatile memory can store a set of operating system instructions and / or software instructions in the form of computer-readable instructions for controlling the operation of processor 222, which, when executed on one or more processors 222, cause processor 222 to perform corresponding methods, procedures, operations, and / or functions of scrambling DNN operations, scrambling / randomization / descrambling operations, and / or methods as described herein at UE 220. The non-volatile memory can be any kind of memory, such as ROM, flash memory, SD drive, magnetic drive memory, or disk drive memory, and similar types as required by the application.

[0066] The non-volatile memory of memory unit 223 of UE 220 includes computer program code and / or instructions for implementing the UE DNN controller (UEDNNC) 224 and / or the UE uplink transmission DNN structure (UE UL TX DNN) 226 and / or the UE downlink reception DNN structure (UE DL RX DNN) 228. When executed on one or more processors 212, UE DNNC 224 uses UE UL TX DNN 226 and / or UE DL RX DNN 228, UE NNSI storage / table / buffer 225a stored in memory unit 223, and UE TX / RX DNN storage / table 225b stored in memory unit 223 to control scrambling DNN operations between BS 210 and UE 220. Although UE DNNC 224 is shown as part of memory unit 223, this is merely an example, and UE DNNC 224 is not limited thereto. Those skilled in the art will understand that UE DNNC224 can be implemented in any combination of hardware and / or software of UE 220 and / or according to application requirements.

[0067] During operation, BS 210 and UE 220 establish a DL / UL DNN communication session with each other. The DL / UL DNN communication session includes DL DNN communication from BS 210 to UE 220 and UL DNN communication from UE 220 to BS 210. During the establishment of the DL / UL DNN communication session, BS 210 and UE 220 communicate with each other to define, agree on, and / or configure the types of BS DL TX DNN 206 and UE DL RX DNN 228 pairs used in DL DNN communication, and the types of UE UL TX DNN 226 and BS uplink receive DNN structure (BS UL RX DNN) 208 pairs used in UL DNN communication. For example, BS 210 selects the appropriate BS DL TX DNN 206 and UE DL RX DNN 228 pairs from the BS TX DNN / RX DNN store or table 215b (e.g., the DNN configuration table) for DL ​​DNN communication. Similarly, BS 210 selects the appropriate UE UL TX DNN 226 and BS UL RX DNN 208 pair from the BS TX DNN / RX DNN storage or Table 215b for UL DNN communication. Alternatively, as an option, UE 220 may select the appropriate UE UL TX DNN 226 and BS UL RX DNN 208 pair from the UE TX DNN / RX DNN storage 225b for UL DNN communication.

[0068] The BS TX / RX DNN store or table 215b includes a set of DL transmit / receive DNN structures and / or their pairs, and a set of UL transmit / receive DNN structures and / or their pairs, each of which is mapped to a DL / UL DNN identifier and stored in the BS TX / RX DNN store or table 215b at BS 210 (e.g., a lookup table). Each transmit / receive DNN structure is trained to transmit / receive a waveform that, depending on the training / conditions, is adapted to efficiently overcome, for example, a specific channel environment, one or more specific channel impairments, and / or one or more different types of interference (e.g., multipath interference, multiple access interference, narrowband interference) found in the current communication system, thereby further enhancing performance. DL / UL TXDNNs and corresponding DL / UL RX DNNs are trained to overcome one or more types of channel impairments and / or channel environments using supervised DNN training on multiple scenarios. For example, supervised DNN training jointly trains a pair of DL TX DNNs / DL RX DNNs and a pair of UL TX DNNs / UL RX DNNs. For example, BS 210 and UE 220 use DL transport DNN and receive DNN pairs specifically trained for downlink communication channels (e.g., PDSCH), and UE 220 and BS 210 use uplink transport DNN and receive DNN pairs specifically trained for uplink communication channels (e.g., PUSCH).

[0069] UE 220 also has a corresponding set of UL / DL transmit / receive DNN structures that are also mapped to the same UL / DL DNN identifier and stored in the UE TX DNN / RX DNN storage 225b (e.g., lookup table) at UE 220. BS 210 selects the DL transmit and receive DNN structures and / or UL transmit and receive DNN structures based on the DL and / or UL communication channel / environment, the communication performance requirements for the DL and / or UL DNN connections, and the type of DL and / or UL data communication in the communication session (e.g., voice communication, data communication, multimedia streaming, and the like).

[0070] BS 210 sends one or more control messages indicating a selected DL / UL receive DNN structure (e.g., DL / UL DNN identifier) ​​for DL ​​and UL communication channels (e.g., PDSCH and PUSCH) and DL and UL control channels (e.g., PDCCH / PUCCH) for DL ​​and UL DNN communication between UE 220 and BS 210. For DL ​​DNN communication, BS 210 and UE 220 configure their respective BS DL TX DNN 206 and UE DL RX DNN 228 based on the selected DL DNN identifier. For UL DNN communication, BS 210 and UE 220 configure their respective BS UL RX DNN 208 and UE UL TX DNN 226 based on the selected UL DNN identifier. After BS 210 and UE 220 establish a DL / UL DNN communication session, BS 210 and UE 220 use BS DL TX DNN 206 and UE DL RX DNN 228 to perform DL DNN communication for one or more time slots. UE 220 and BS 210 also use UE UL TX DNN 226 and BS UL RX DNN 208 respectively to perform UL DNN communication for one or more time slots.

[0071] During DL DNN communication, when BS 210 detects that a transmission from BS 210 does not meet a specific white noise interference level, BS 210 performs DL DNN scrambling. During UL DNN communication, when UE 220 detects that a transmission from UE 220 does not meet a specific white noise interference level, UE 220 performs UL DNN scrambling. For DL ​​DNN scrambling communication from BS 210 to UE 220, the TX DNN controller of BS DNNC 214 controls the DL DNN scrambling operation at BS 210, and the RX DNN controller of UE DNNC 224 controls the corresponding DL DNN scrambling operation at UE 220. For UL DNN scrambling communication from UE 220 to BS 210, the TX DNN controller of UE DNNC 224 controls the UL DNN scrambling operation at UE 220, and the RX DNN controller of BS DNNC 214 controls the corresponding UL DNN scrambling operation at BS 210.

[0072] For example, in DL DNN scrambling, when the output communication signal from BS DL TX DNN 206 does not meet the white noise interference level when predicted for transmission on PDSCH, BS 210 enables the execution of scrambling DLDNN operation for one or more time slots. The DL TX DNN controller of BS 210 detects that BS DL TX DNN 206 generates an output communication signal from the input communication data for a specific time slot, which will result in a downlink transmission signal 205a on PDSCH that does not meet the white noise interference level. When this occurs, the DL scrambling DNN operation is enabled, and the DL TX DNN controller of BS 210 selects Neural Network Scrambling Information (NNSI), which can be retrieved from the BS NNSI store 215a or iteratively determined by randomizing one or more neural network layers of the BS DL TX DNN 206 for reconfiguring the BS DL TX DNN 206 to generate an output communication signal from the same input communication data for a specific time slot. This will result in a downlink transmission signal 205a on the PDSCH that satisfies the white noise interference level set for the PDSCH. For example, the NNSI (e.g., randomization / scrambling parameters and / or randomized specific / selected neural network layers of the transmission DNN) is selected to randomize or scramble the order of a set of network nodes in one or more specific neural network layers of the BS DL TX DNN 206 of BS 210, such that the resulting output communication signal of the reconfigured BS DL TX DNN 206 produces a downlink transmission signal 205a that satisfies the white noise interference level of the PDSCH. Specify the randomized neural network layer of BS DL TX DNN 206 in NNSI.

[0073] Similarly, when UE 220 (or BS 210) detects that the uplink transmission signal 205b on the PUSCH will not meet the white noise interference level set for the PUSCH, UL scrambling DNN operation is enabled. For example, BS 210 selects an NNSI to reconfigure UE 220's UE UL TX DNN 226 to generate an output communication signal from it, which will result in an uplink transmission signal 205b on the PUSCH that meets the white noise interference level. In the example, when BS 210 selects an NNSI for UE 220, BS 210 uses random UE UL input communication data and UE UL TX DNN 226 to simulate the UL to generate an output communication signal, and selects an NNSI that produces an output communication signal that meets the white noise interference level of the UL. BS 210 sends the selected NNSI to UE 220 in a control message. The NNSI does not necessarily depend on specific UE UL input communication data, but primarily on the UE UL TX DNN 226 used in the UE UL. For example, NNSI (e.g., randomization / scrambling parameters and / or randomized specific layers of the transmission DNN) is selected to randomize or scramble the order of a set of network nodes in one or more specific neural network layers of the UE UL TX DNN 226 of the UE 220, such that the resulting output communication signal of the reconfigured UE UL TX DNN 226 produces an uplink transmission signal 205b that satisfies the white noise interference level of PUSCH.

[0074] For DL ​​scrambling communication, after selecting the NNSI for generating a downlink transmission signal 205a that satisfies the white noise interference condition for the BS DL TX DNN 206, the BS 210 sends a control message to the UE 220 including an indication of the NNSI and scrambling timing information. This control message is used by the UE DNNC 224 to reconfigure the UE DL RX DNN 228 according to the scrambling timing information for processing the downlink transmission signal 205a generated from the output communication signal generated from the reconfigured BS DL TX DNN 206. The received scrambling timing information instructs the UE DNNC 224 when to use the selected NNSI to reconfigure the UE DL RX DNN 228. For example, the UE DL RX DNN 228 processes the received downlink transmission signal 205a to generate reconstructed communication data for a specific time slot, corresponding to the input communication data transmitted for that specific time slot. As an example, UE 220 may send the reconstructed communication data to the data sink of UE 220, or send the reconstructed communication data to one or more upper protocol layers of the protocol stack of UE 220 (e.g., to the application protocol layer of the protocol stack for use by one or more applications running on UE 220).

[0075] Similarly, for UL scrambling communication, after selecting the NNSI for generating the uplink transmission signal 205b that satisfies the white noise interference condition for the UE UL TX DNN 226, the BS 210 sends a control message to the UE 220 including an indication of the NNSI and scrambling timing information. This control message is used by the UE DNNC 224 to reconfigure the UE UL TX DNN 226 according to the scrambling timing information for processing the uplink transmission signal 205b for transmission to the BS 210 on the PUSCH. The BS 210 receives the uplink transmission signal and processes it using the reconfigured BS UL TX DNN 208 based on the selected NNSI and the corresponding scrambling timing information. The scrambling timing information instructs the BS DNNC 214 when to reconfigure the BS DL TX DNN 208 using the selected NNSI. For example, the BS UL RX DNN 208 processes the received uplink transmission signal 205b to generate reconstructed communication data for a specific time slot, corresponding to the input communication data transmitted for that specific time slot.

[0076] Before transmitting the output communication signal generated by BS DL TX DNN 206, the BS TXDNNC of BS DNN controller 214 performs analysis on the output communication signal to identify whether the output communication signal will whiten the downlink transmission signal 205a and whether it meets the white noise interference level of PDSCH. When the analysis indicates that RF processing of the output communication signal will produce a transmission signal that does not meet the white noise interference level (e.g., a predicted transmission signal), then BS 210 and UE 220 perform the NNSI selection and scrambling DNN operations mentioned above.

[0077] The DL / UL scrambling process offers the following advantages: BS 210 / UE 220 does not transmit the output communication signal of the DL / UL transmission DNN until the obtained predicted downlink / uplink transmission signals 205a / 205b meet the white noise interference level set by BS 210 / UE 220. DL / UL transmissions from BS 210 and UE 220, respectively, will meet the corresponding white noise interference levels without transmission spikes causing interference to neighboring devices and / or cells. The DL / UL scrambling process also synchronizes the DL / UL transmission DNNs 206 / 226 and the corresponding DL / UL receive DNNs 228 / 208 to operate together, and recovers the corresponding input communication data at UE 220 or BS 210, respectively.

[0078] The DL / UL scrambling communication performed by BS 210 and UE 220 offers numerous advantages, including, for example, the efficient design and control of BS DL TX DNN 206 and UE DL RX DNN 228 and / or UE UL TX DNN 226 and BS UL RX DNN 208 to maintain a downlink transmission signal 205a or uplink transmission signal 205b that meets white noise interference levels, while reducing interference to other DNN or non-DNN receivers in the cell or area surrounding BS 210 and UE 220. Scrambling DNN operation enables BS 210 or UE 220 to maintain white noise interference levels when transmitting each of the downlink transmission signal 205a and / or uplink transmission signal 205b without generating unexpected transmission spikes due to various different combinations of input communication data processed by each of BS DL TX DNN 206 or UE UL TX DNN 226. Therefore, when the BS DL TX DNN 206 or UE UL TX DNN 226 and the corresponding UE DL RX DNN 228 or BS UL RX DNN 208 are reconfigured respectively, transmission spikes in the transmitted signals from BS 210 or UE 220 are mitigated, reduced, and / or prevented from occurring, while satisfying the white noise interference level. The BS DL TX DNN 206 (or UE UL TX DNN 226) and the corresponding UE DL RX DNN 228 (or BS UL RX DNN 208) of BS 210 and UE 220 (or UE 220 and BS 210) are efficiently, quickly, and dynamically reconfigured in real time to change the white noise interference level of the downlink transmission signal 205a / uplink transmission signal 205b, while maintaining the transmission power or bit / symbol error rate of the signal of interest. Further advantages include efficient synchronization between BS 210 and UE 220 during DL / UL scrambling communication, which enables dynamic whitening of transmitted signals 205a / 205b from BS 210 or UE 220 and allows the corresponding UE 220 or BS 210 to receive and decode the dynamically whitened transmitted signals.

[0079] Although reference Figure 1 or Figure 2aAnd / or wireless communication networks / systems described herein are merely examples and are not limited thereto. Those skilled in the art will understand that any type of communication network / system is applicable, such as, for example, any telecommunications network; any wired communication network; any wireless communication network; satellite network; peer-to-peer communication network; communication networks using third-generation (3G), fourth-generation (4G), fifth-generation (5G) and / or sixth-generation (6G) and higher standard technologies; Wi-Fi communication network; optical communication network; fiber optic communication network; and / or any other network used for communication between a first device and a second device; combinations thereof, modifications thereof, and / or as required by the application. Although referenced... Figure 2a And / or the first device described herein is described as BS 210, but this is merely an example and not a limitation. Those skilled in the art will understand that the first device can be any type of communication device capable of communicating with the second device, such as, but not limited to, UE, BS, satellite, mobile phone or smartphone, laptop computer, computing device, device using 3G, 4G, 5G and / or 6G and higher standard technologies, and / or any other device used to communicate with the second device; combinations thereof, modifications thereof, and / or as required by the application. Although referenced... Figure 2a And / or the second device described herein is described as UE 220, but this is merely an example and is not limited thereto. Those skilled in the art will understand that the second device can be any type of communication device capable of communicating with the first device, such as, but not limited to, UE, BS, satellite, mobile phone or smartphone, laptop computer, computing device, device using 3G, 4G, 5G and / or 6G and higher standard technologies, and / or any other device used to communicate with the first device; combinations thereof, modifications thereof, and / or as required by the application.

[0080] Figure 2b An example power spectral density (PSD) graph 230 is shown, representing the PSD of the transmitted signal 235 satisfying white noise interference level 232. In this example, the RF analog transmission component of the RF front end processes the output communication signal generated by the transmission DNN model and upconverts the resulting transmitted signal 235 to a carrier frequency f with a bandwidth 231 of 2f1. c The transmitted signal 235 has a PSD lower than the white noise interference level 232 across its entire bandwidth. In this example, Figure 2b The white noise interference level 232 is shown as the white noise interference level at the bandwidth of interest (i.e., frequency f). c -f1 and f cThe white noise level has a constant amplitude over the frequency of the bandwidth 231 between +f1 and the flat white noise power spectral density. Although the white noise level is described as a flat white noise power spectral density, this is merely an example, and those skilled in the art will understand that the white noise level is any suitable measure of white noise interference or disturbance, such as, for example, the total power of the flat white noise power spectral density over the frequency of the bandwidth of interest, and / or any other suitable measure of disturbance and the like. In this case, spectral analysis of the transmitted signal 235 indicates that the output communication signal generated by the transmission DNN model will produce a transmitted signal 235 that satisfies the white noise interference level.

[0081] Figure 2c Another PSD diagram 240 is shown, representing the PSD of the transmitted signal 245 that does not meet the white noise interference level 232. In this example, the RF analog transmission component processes the output communication signal generated by the transmission DNN model and upconverts the resulting transmitted signal 245 to a carrier frequency f with a bandwidth 231 of 2f1. c The PSD of the transmitted signal 245 exhibits significant transmission spikes in the form of PSD peaks 246a, 246b, 246c, and 246d that exceed the white noise interference level 232. If the first device 210 processes the output communication signal for transmission, these PSD peaks 246a, 246b, 246c, and 246d will cause significant interference to other devices in the area of ​​the first device. In this case, spectral analysis of the transmitted signal 245 indicates that the output communication signal generated by the transmission DNN model will produce a transmitted signal 245 that does not meet the white noise interference level. As another example, even if the average PSD of the transmitted signal 245 is lower than the white noise interference level 232, the analysis of the transmission spikes or PSD peaks 246a, 246b, 246c and 246d of the transmitted signal 245 can determine that the average PSD of the PSD peaks 246a, 246b, 246c and 246d is higher than the tolerable transmission spike PSD threshold, and therefore, it can be indicated that the transmitted signal 245 does not meet the white noise interference level 232.

[0082] Figure 2d A further example PSD diagram 250 is shown, representing the PSD of the transmitted signal 255, which also does not meet the white noise interference level 232. In this example, the RF analog transmission component processes the output communication signal generated by the transmission DNN model and upconverts the resulting transmitted signal 255 to a carrier frequency f with a bandwidth 231 of 2f1. cThe transmitted signal 255 has a PSD higher than the white noise interference level 252 across its entire bandwidth. In this case, spectral analysis of the transmitted signal 255 indicates that the output communication signal generated by the transmission DNN model will produce a transmitted signal 255 that exceeds the white noise interference level and therefore does not meet the white noise interference level.

[0083] Figure 2e A further example PSD diagram 260 is shown, representing the PSD of the transmitted signal 265 satisfying white noise interference level 232. In this example, the RF analog transmission component processes the output communication signal generated by the transmission DNN model and upconverts the resulting transmitted signal 265 to a carrier frequency f with a bandwidth 231 of 2f1. c The PSD of the transmitted signal 265 has small transmission spikes in the form of PSD peaks 266a, 266b, 266c, and 266d that exceed the white noise interference level 232. If the first device 210 processes the output communication signal for transmission, these small PSD peaks 266a, 266b, 266c, and 266d should not cause significant interference to other devices in the area of ​​the first device. In this case, spectral analysis of the PSD of the transmitted signal 265 will indicate that the output communication signal generated by the transmission DNN model will produce a transmitted signal 265 that meets the white noise interference level. For example, the average PSD of the transmitted signal 265 is below the white noise interference level 232, and analysis of the small transmission spikes or PSD peaks 266a, 266b, 266c, and 266d of the transmitted signal 265 can determine that the average PSD of the PSD peaks 266a, 266b, 266c, and 266d is below a tolerable transmission spike PSD threshold. Given these two conditions, the analysis of the PSD of the transmitted signal 265 can indicate that the transmitted signal 265 meets the white noise interference level 232.

[0084] Figure 3a An example communication system 300a is illustrated, comprising a first device 310 and a second device 320, which implement scrambling of the input layer of a transmitting DNN structure 306 and reciprocal descrambling of the output layer of a receiving DNN structure 308. The transmitting DNN structure 306 of the first device 310 includes a transmitting DNN model comprising an input neural network layer 316 and further DNN layers 307 (e.g., one or more hidden layers and an output layer) represented by blocks labeled DNN1. The input neural network layer 316 receives input communication data 301a, wherein the further DNN layer 307 processes the output of the input neural network layer 316 to generate an output communication signal 318. A TX RF front-end component 303a processes the output communication signal 318 and transmits it as a transmission signal 305 via an antenna.

[0085] In this example, the transport DNN model of the transport DNN structure 306 is reconfigured based on a randomization operation performed on the ordering of the neural network nodes of the input neural network layer 316. In this case, a scrambling DNN operation scrambles the nodes of the input neural network layer. Equivalent to scrambling the input neural network layer, the randomization operation can be used to scramble the input communication data 301a before it is input to the input neural network layer of the transport DNN model of the transport DNN structure 306. For example, the first device 310 selects the NNSI that specifies the input layer of the transport DNN model to be randomized / scrambled, and uses this configuration of the transport DNN structure 306 to select one or more specific time slots. For example, the selected NNSI includes a neural network layer indicator that specifies that the configuration of the transport DNN structure 306 will be used to randomize / scramble the input layer of the transport DNN model for one or more specific time slots. The first device 310 sends a control message to the second device 320, which includes the selected NNSI and the one or more time slots specified when scrambling occurs when using the selected NNSI.

[0086] The receiving DNN structure 308 of the second device 320 includes a receiving DNN model comprising a DNN layer 309 (e.g., an input layer and one or more hidden layers) and an output neural network layer 317, represented by blocks labeled DNN2. The DNN layer 309 of the receiving DNN model receives a communication signal 319 output from the RX RF front-end component 303b after receiving a transmission signal 305 in a specific time slot. The DNN layer 309 processes the received communication signal 319 and outputs scrambled, reconstructed communication data in the output neural network layer 317 of the receiving DNN structure. In this case, the receiving DNN model of the receiving DNN structure 308 has been reconfigured using NNSI, wherein the output neural network layer 317 performs an inverse randomization operation (descrambling operation) on the neural network nodes of the output neural network layer 317. The inverse randomization operation corresponds to the randomization operation performed on the neural network nodes of the input neural network layer 316 of the transmission DNN of the transmission DNN structure 306. The receiving DNN structure 308 sends the descrambled and reconstructed communication data 301b corresponding to the input communication data 301a to the data sink of the second device 320, or sends the reconstructed communication data 301b to one or more upper protocol layers of the protocol stack of the second device 320 (e.g., to the application protocol layer of the protocol stack for use by one or more applications executing on the second device 320).

[0087] The input neural network layer 316 of the transport DNN structure 306 comprises a set of N neural network nodes, where N > 1. Reconfiguring the transport DNN structure 306 includes performing a randomization operation on the input neural network layer 316 by randomizing the set of N neural network nodes (or a subset of those N nodes). When the NNSI includes one or more random permutation parameters (e.g., seed, random function type, etc.), randomizing the set of N neural network nodes of the input neural network layer 316 may include performing random permutations on the set of N neural network nodes of the input neural network layer 316. For example, the randomization operation uses random permutation parameters to generate an N-dimensional permutation matrix using a random permutation sequence of length N. Randomizing the set of N neural network nodes of the input neural network layer 316 is based on randomizing the order of the N neural network nodes by multiplying the order of the N neural network nodes by the N-dimensional permutation matrix. In another example, this is equivalent to multiplying the input communication data 301a (or its sorting) with an N-dimensional permutation matrix before it is fed into the input neural network layer 316.

[0088] After the second device 320 has received the NNSI for a specific time slot, it reconfigures the receive DNN structure 308 before processing the received communication signal 319 corresponding to that specific time slot. Given that the NNSI for that specific time slot includes a neural network layer indicator specifying that the input neural network layer 316 of the transmission DNN structure 306 is randomized / scrambled, the receive DNN structure 308 of the second device 320 is reconfigured by performing an inverse randomization operation on the set of neural network nodes of the output neural network layer 317 of the receive DNN structure 308. The output neural network layer 317 generates descrambled reconstructed communication data 301b corresponding to the input communication data 301a. In this example, the second device 320 sends the reconstructed communication data 301b to its data sink, or sends the reconstructed communication data 301b to one or more upper protocol layers of the protocol stack of the second device 320.

[0089] The output neural network layer 317 of the receiving DNN structure 308 includes a set of N neural network nodes, where N > 1. When the NNSI includes random permutation parameters (e.g., seed, random function type, etc.), a descrambling or inverse randomization operation is performed on the set of N neural network nodes of the output neural network layer 317. For example, this includes performing a random inverse permutation (or depermutation) on the set of N neural network nodes of the output neural network layer 317. For example, the descrambling or inverse randomization operation uses the random permutation parameters received in the NNSI for a specific time slot to generate an N-dimensional permutation matrix using a random permutation sequence of length N. The N-dimensional permutation matrix is ​​inverted to generate an inverse N-dimensional permutation matrix. The descrambling operation is performed by multiplying the N-dimensional output vector of the output neural network layer with the inverse N-dimensional permutation matrix, wherein the receiving DNN structure 308 generates descrambled and reconstructed communication data 301b corresponding to the input communication data 301a.

[0090] Randomizing the input neural network layer and / or one or more hidden neural network layers provides the following advantages: increasing the NNSI search space and increasing the possibility of iteratively determining an NNSI suitable for the transmission DNN structure 306" to generate an output communication signal with a spectrum that meets the white noise interference level when transmitted.

[0091] Figure 3b An example communication system 300b is shown, comprising a first device 310 and a second device 320, wherein Figure 3b Further revisions Figure 3a The first device 310 and the second device 320 are configured to implement scrambling / descrambling of the output layer of the transmitting DNN structure 306' and reciprocal descrambling of the input layer of the receiving DNN structure 308'. In this example, the transmitting DNN structure 306' of the first device 310 includes a transmitting DNN model comprising an output neural network layer 316' and a DNN layer 307' (e.g., an input layer and one or more hidden layers) represented by DNN1. The DNN layer 307' receives and processes input communication data 301a, while the output neural network layer 316' receives the input communication data 301a processed by DNN1 to generate an output communication signal 318. The TX RF front-end component 303a processes the output communication signal 318 and transmits it as a transmission signal 305 via an antenna.

[0092] In this example, based on and reference Figure 3aThe scrambling or randomization of the neural network nodes of the output neural network layer 316' is performed in a similar manner to the neural network nodes of the input neural network layer 316' to reconfigure the transmission DNN structure 306'. The receiving DNN structure 308' of the second device 320 includes a receiving DNN model comprising a DNN layer 309' (e.g., one or more hidden layers and an output layer) and an input neural network layer 317', represented by blocks labeled DNN2. The descrambling operation is described in reference... Figure 3a The descrambling operation of the output neural network layer 317 describes the descrambling of the input neural network layer 317'. This effectively descrambles the received communication signal 319. The DNN layer 309' processes the descrambled received communication signal 319 and outputs descrambled and reconstructed communication data 301b corresponding to the input communication data 301a.

[0093] The output neural network layer 316' of the transmission DNN structure 306' includes a set of N neural network nodes, where N>1. The reconfiguration of the transmission DNN structure 306' includes, as per the reference pair... Figure 3a The scrambling or randomization of the input neural network layer 316 describes the scrambling or randomization operation performed on the output neural network layer 316' of the transmitting DNN structure 306'. For example, randomizing the set of N neural network nodes of the output neural network layer 316' is based on randomizing the order of the N neural network nodes by multiplying the order of the N neural network nodes of the output neural network layer 316' by an N-dimensional permutation matrix. In another example, this is equivalent to multiplying the output communication signal 318 (or its order) by an N-dimensional permutation matrix before inputting it to the TX RF front-end component 303a.

[0094] After the second device 320 has received the NNSI for a specific time slot, the second device 320 reconfigures the receive DNN structure 308' before processing the received communication signal 319 corresponding to the specific time slot. Given that the NNSI for this specific time slot includes a neural network layer indicator specifying that the output neural network layer 316' of the transmission DNN structure 306' is randomized / scrambled, this is done by referring to... Figure 3aThe descrambling or derandomization operation on the set of neural network nodes of the output neural network layer 317 is performed in a similar manner to the descrambling or derandomization operation on the set of neural network nodes of the input neural network layer 317' of the receiving DNN structure 308' to reconfigure the receiving DNN structure 308' of the second device 320. The remaining DNN layers 309' represented by DNN2 of the receiving DNN structure 308' generate descrambled and reconstructed communication data 301b corresponding to the input communication data 301a. The second device 320 sends the reconstructed communication data 301b to the data sink of the second device 320, or sends the reconstructed communication data 301b to one or more upper protocol layers of the protocol stack of the second device 320.

[0095] Randomizing the output neural network layer of the transmission DNN structure 306' provides the following advantages: reducing the computational resources required to iteratively determine the updated NNSI, since only the output neural network layer representing the output communication signal needs to be processed to determine whether the potential NNSI produces an output communication signal that can be tuned to meet the white noise spectrum or white noise interference level.

[0096] Figure 3c An example communication system 300c is shown, comprising a first device 310 and a second device 320, wherein Figure 3c Further revisions Figure 3a or Figure 3b The first device 310 and the second device 320 more generally implement scrambling / descrambling of one or more hidden neural network layers 316" and 317" of the transmitting DNN structure 306" and the receiving DNN structure 308", respectively. The transmitting DNN structure 306" of the first device 310 includes a transmitting DNN model that includes one or more hidden neural network layers 316" for hidden layer scrambling operations and further DNN layers 307" (e.g., an input layer, one or more hidden layers (if any), and an output layer) represented by blocks labeled DNN1. The DNN layer 307" receives and processes input communication data 301a and passes the processed data to one or more hidden neural network layers 316" for scrambling accordingly. The one or more hidden neural network layers pass the scrambled, processed data to the output neural network layer of DNN1 to generate an output communication signal 318. The TX RF front-end component 303a processes the output communication signal 318 for transmission as a transmission signal 305 via an antenna.

[0097] In this example, based on and respectively refer to Figure 3a or Figure 3bThe scrambling or randomization operations performed on the ordering of neural network nodes of one or more hidden neural network layers 316" in a manner similar to those described for the input neural network layer 316 or the output neural network layer 316' are used to reconfigure the transmission DNN structure 306".

[0098] The receiving DNN structure 308" of the second device 320 includes a receiving DNN model comprising a DNN layer 309" (e.g., an input neural network layer, one or more hidden layers (if any), and an output neural network layer) represented by DNN2, and one or more hidden neural network layers 317" for descrambling (i.e., performing a descrambling operation). The input neural network layer of DNN2 309" receives the communication signal 319 output from the RX RF front-end component 303b after receiving the transmission signal 305 in a specific time slot. In this case, the receiving DNN structure 308" has been reconfigured using NNSI, wherein the one or more hidden neural network layers 317" are descrambled to be referenced to respectively. Figure 3a or Figure 3b The inverse randomization operation (descrambling operation) is performed on the corresponding hidden neural network layer 317" in a manner similar to that described for the neural network nodes of the output neural network layer 317" or the input neural network layer 317'. This effectively descrambles the received communication signal 319, where the hidden neural network layer 317" and the DNN layer 309" process the communication signal 319 and output descrambled and reconstructed communication data 301b corresponding to the input communication data 301a. The symmetric DNN architecture used for transmission and reception of DNN structures 306" and 308" simplifies the scrambling and descrambling operations.

[0099] Randomizing one or more hidden neural network layers offers the following advantages: it increases the NNSI search space and increases the possibility of iteratively determining an NNSI suitable for transmitting the DNN structure 306" to generate an output communication signal with a spectrum that satisfies the level of white noise interference when transmitted.

[0100] refer to Figures 3a to 3cWhile the scrambling of the input layer, output layer, and 1st or (L-1+1)th hidden layer of the transport DNN structure 306 is described separately, this is merely illustrative, and those skilled in the art will understand that one or more of the input layer, output layer, and hidden layer of the transport DNN structure 306 of the first device 310 may be scrambled and / or randomized. More generally, the NNSI, together with scrambling parameters, specifies which neural network layers are scrambled, wherein 3GPP standards and the like may predefine the neural network layers to be scrambled. Alternatively or additionally, the NNSI specifies a neural network layer indicator (e.g., a flag, bit, or field) that specifies one or more neural network layers of the transport DNN structure 306 for scrambling. More generally, the NNSI includes a neural network layer indicator that specifies the scrambling of one or more neural network layers of the transport DNN structure 306. The transport DNN structure 306 includes at least one input layer, one or more hidden layers, and an output layer. The first device 310 uses the NNSI to scramble / randomize the indicated one or more neural network layers. For each l-th neural network layer in one or more neural network layers, for 1≤l≤L and L is the number of neural network layers, where l=1 is the input layer and l=L is the output layer, the l-th neural network layer of the transmitting DNN structure 306 is randomized / scrambled, and the second device 320 uses the corresponding NNSI (received in the control message for a specific time slot) to reconfigure the receiving DNN structure 308 by performing an inverse randomization operation on the group of neural network nodes of the (L-l+1)-th neural network layer of the receiving DNN.

[0101] When the NNSI includes random permutation parameters, the l-th neural network layer of the transmission DNN structure 306 is reconfigured by: generating an N-dimensional permutation matrix based on the random permutations of the l-th neural network layer, where N > 1 and N is the number of neural network nodes in the l-th neural network layer; and multiplying a specific order of the group of N neural network nodes of the l-th neural network layer by the N-dimensional permutation matrix to form a randomized order of the group of N neural network nodes. The first device 310 uses the reconfigured transmission DNN structure 306 to process the input communication data 301a and generate a scrambled output communication signal 318. The TX RF front-end component 303a processes the output communication signal 318 and transmits it as a transmission signal 305 to the second device 320 via an antenna in a specific time slot. The transmission signal 305 satisfies the white noise interference level because the NNSI is selected to meet this criterion.

[0102] The second device 320 receives a corresponding NNSI, which includes random permutation parameters, one or more specific neural network layers, and specific time slot or scrambling timing information associated with when the receiving DNN structure 308 should be reconfigured for descrambling the corresponding received communication signal 319. At an appropriate time, for example, before processing the received communication signal 319 corresponding to a specific time slot, each of the specified one or more neural network layers of the receiving DNN structure 308 is reconfigured for descrambling based on the corresponding NNSI. The (L-1+1)th neural network layer of the receiving DNN structure 308 has a set of N neural network nodes, N>1, wherein a specific ordering is reconfigured by: generating an N-dimensional permutation matrix based on the random permutations of the (L-1+1)th neural network layer; inverting the N-dimensional permutation matrix to generate an inverse N-dimensional permutation matrix; and multiplying the specific ordering of the set of N neural network nodes of the (L-1+1)th neural network layer by the inverse N-dimensional permutation matrix to form an inverse randomized ordering of the set of N neural network nodes. After reconfiguration, the receiving DNN structure 308 processes the received communication signal 319 received in a specific time slot to generate reconstructed communication data 301b, which corresponds to the input communication data 301a transmitted in that specific time slot.

[0103] Figure 4a An example transmission DNN scrambling process 400 is shown for generating a scrambled output communication signal from a transmission DNN model that satisfies the white noise interference level during transmission. Figure 1 The reference numerals in the figures were reused Figures 4a to 4d Similar or identical components, features, and / or devices. In this example, the first device 104a communicates with the second device 104b, wherein the first device 104a includes a transmission DNN structure 106 having a transmission DNN model 107, and the second device includes a receive DNN structure 108 having a receive DNN model 109. The transmission DNN model 107 has been trained to override several functions of the transmission processing chain (e.g., encoding, interleaving, scrambling, precoding, etc.). The receive DNN model 109 has also been trained to perform the inverse operation of the transmission DNN model 107 to generate reconstructed communication data 101b'. The transmission DNN scrambling process 400 includes the following steps:

[0104] In step 402, the first device 104a retrieves input communication data 101a' for transmission in a specific time slot. The input communication data 101a' can be a bit stream from a data source or any type of digital data.

[0105] In step 404, the first device 104a processes the retrieved input communication data 101a' using the transmission DNN structure 106 to generate an output communication signal 118 for transmission to the second device 104b.

[0106] In step 406, the first device 104a checks whether the transmission of the generated output communication signal 118 will meet the white noise interference level. If the transmission of the generated output communication signal 118 will meet the white noise interference level (e.g., 'yes'), the process continues to step 412; otherwise (e.g., 'no'), the process continues to step 408. The first device 104a performs a spectrum analysis to determine whether the output communication signal 118 will produce a transmission signal 105a' that meets the white noise interference level. That is, the first device 104a uses spectrum analysis to predict whether the output communication signal 118 will produce a transmission that meets the white noise interference level. For example, the spectrum analysis includes calculating the power spectral density of the predicted transmission signal that will be generated by transmitting the output communication signal 118 after RF analog TX processing. The predicted transmission signal is not transmitted; instead, it is an estimate of the transmission signal that will be generated when the output communication signal 118 is transmitted by the first device 104a.

[0107] In step 408, in response to the fact that the transmission of the generated output communication signal does not meet the white noise interference level, the first device 104a performs a scrambling DNN operation by selecting the NNSI of the transmission DNN model 107 for reconfiguring the transmission DNN structure 106 to process the input communication data 101a' to generate a scrambled output communication signal that meets the white noise interference level when transmitted.

[0108] For example, the NNSI may include data representing the randomization of the order of the inputs, outputs, or a set of neural network nodes in one or more specified neural network layers of the transmitted DNN model. As an example, the NNSI may include one or more random permutation parameters for randomizing the order of the set of neural network nodes in one or more specified neural network layers of the transmitted DNN model 107 of the first device 104a. Randomizing the order is performed using the random permutation parameters to randomize the order of the set of neural network nodes in one or more specified neural network layers of the transmitted DNN model 107.

[0109] In step 410, the first device 104a reconfigures the transport DNN model 107 based on the selected NNSI. Proceeding to step 404, the reconfigured transport DNN model 107 processes the input communication data 101a' retrieved in step 402. For example, reconfiguring the transport DNN model includes using the NNSI to reconfigure the transport DNN model 107 to randomize the ordering of the inputs, outputs, or a set of neural network nodes of one or more specific neural network layers, as specified by the NNSI.

[0110] In step 412, the first device 104a checks whether the NNSI has been updated. If the NNSI has been updated (e.g., 'yes'), that is, the updated NNSI is different from the previously selected NNSI, then proceed to step 414; otherwise (e.g., 'no'), then proceed to step 416.

[0111] In step 414, a control message is transmitted from the first device 104a to the second device 104b. This control message includes data indicating an indication of NNSI and scrambling timing information, which includes specific time slots (e.g., one or more time slots) to guide the second device 104a on when to reconfigure the receiving DNN model 109 to generate reconstructed communication data 101b' corresponding to the input communication data 101a'. The process continues to step 416.

[0112] In step 416, the first device 104a transmits scrambled output communication signal 118 to the second device 104b with scrambling timing information, wherein the transmission of the scrambled output communication signal satisfies the white noise interference level. The process continues to step 402 to retrieve further input communication data 101a' for transmission.

[0113] Figure 4b It shows Figure 4a The example spectrum analysis procedure in step 406 is used to analyze whether the spectral density of the transmitted signal representing the output communication signal generated by the transmitted DNN model 107 meets the white noise interference level. The spectrum analysis procedure includes the following steps:

[0114] In step 406a, the spectral density of the predicted transmitted signal, representing the output communication signal 118 to be transmitted after processing, is estimated. Since this transmitted signal has not yet been transmitted, it is a predicted transmitted signal. For example, given the output communication signal 118, the spectral density of the predicted transmitted signal is estimated by simulating or modeling the characteristics of the RF front-end TX component, antenna, and communication channel. Alternatively, the estimated spectral density of the predicted transmission is estimated for each antenna output of the RF front-end TX component and the like. The estimated spectral density of the predicted transmitted signal can be estimated based on a combined estimate of the spectral density of the predicted transmission at the output of each of the antennas in the RF front-end TX component.

[0115] In step 406b, it is checked whether the estimated spectral density of the predicted transmitted signal meets the white noise interference level. If the estimated spectral density of the predicted transmitted signal meets the white noise interference level (e.g., 'Yes'), the process continues to step 406d; otherwise (e.g., 'No'), the process continues to step 406c.

[0116] The check further includes detecting or identifying whether the estimated spectral density meets the white noise interference level. For example, when the predicted transmitted signal forms one or more interference spikes above the white noise interference level, the analysis of the power spectral density of the transmitted signal detects or identifies that the predicted transmitted signal's spectral density does not meet the white noise interference level. In another example, when the power spectral density of the transmitted signal on the bandwidth of interest is below the white noise interference level or within a predetermined threshold region of the white noise interference level, the analysis of the power spectral density of the predicted transmitted signal detects or identifies that the predicted transmitted signal's spectral density does indeed meet the white noise interference level.

[0117] In a further example, the inspection and analysis in step 406b includes comparing the estimated spectral density of the predicted transmitted signal with the white noise spectral density associated with the white noise interference level. If the estimated spectral density of the predicted transmitted signal is less than or substantially matches the white noise spectral density associated with the white noise interference level, then proceed to step 406d. Otherwise, if the estimated spectral density of the predicted transmitted signal is greater than or substantially deviates from the white noise spectral density associated with the white noise interference level, then proceed to step 406c.

[0118] In step 406c, the indication data represents an indication that the white noise interference level is not met. This indication may be based on setting a predetermined negative flag / field value (e.g., '0', 'N', negative binary value) that indicates that the white noise interference level is met.

[0119] In step 406d, the indication data represents an indication that the white noise interference level is met. This indication can be based on setting a predetermined positive flag / field value (e.g., '1', 'Y (yes)', positive binary value) that indicates that the white noise interference level is met.

[0120] For example, the processor or other computing device (e.g., BS DNNC 214 or UE DNNC 224 in Figure 2) executes automatically. Figure 4a The steps of the spectrum analysis process in step 406. Although several examples of performing checks related to whether the predicted transmitted signal meets the white interference noise level in step 406b have been described, these are merely examples, and those skilled in the art will understand that other suitable automated methods or techniques are applicable to checking whether the predicted transmitted signal meets the white interference noise level and the like.

[0121] Figure 4c It is shown that the second device 104b is used to receive including Figure 4a Example control message procedure 430 for one or more control messages of NNSI transmitted from the first device 104a in step 414. Control message procedure 430 includes the following steps:

[0122] In step 432, a control message including NNSI and scrambling timing information is received at the second device 104b from the first device 104a. The scrambling timing information includes a specific time slot associated with the NNSI. This specific time slot indicates when the first device 104a will transmit a transmission signal 105a' representing the output communication signal generated from the transmission DNN model 107 during the scrambling DNN operation.

[0123] In step 434, the second device 104b stores the received NNS mapped to a specific time slot (or scrambling timing information). The second device 104b retrieves the stored NNSI mapped to the specific time slot to reconfigure the receive DNN model 109 before the second device 104b processes the received communication data signal corresponding to the reception of the signal 105a' transmitted in the specific time slot.

[0124] Figure 4d An example descrambling DNN process 440 is shown for receiving the output communication signal of the transmission DNN model 107 from the first device 104a at the second device 104b and generating reconstructed communication data 101b'. The descrambling DNN process 440 includes the following steps:

[0125] In step 442, the second device 104b receives the communication data signal 119 based on the reception of the transmission signal 105a' transmitted from the first device according to scrambling timing information including a specific time slot. For example, the received communication data signal 119 is based on the reception of the transmission signal 105a' transmitted from the first device. Figure 4a The transmission signal 105a' generated in the DNN scrambling process 400 is the output communication signal generated in the transmission signal, wherein the transmission of the output communication signal occurs in a specific time slot or according to specific scrambling timing information.

[0126] In step 444, the second device 104b retrieves the NNSI (if any) for a specific time slot of the scrambling timing information. For example... Figure 4c As described, the NNSI is received from one or more control messages transmitted by the first device 104a and stored at the second device 104b. The second device 104b maps the received NNSI to a corresponding specific time slot for which the NNSI will be used.

[0127] In step 446, the second device 104b checks whether any NNSI associated with a specific time slot has been retrieved. For example, the second device 104b checks whether the NNSI has changed from a previous NNSI for the specific time slot. If the NNSI has not been retrieved or has not changed (e.g., 'No'), the process proceeds to step 450, where the current receive DNN model 109 is used to descramble the received communication data signal 119. If the NNSI has been retrieved or changed (e.g., 'Yes'), the process proceeds to step 448, where the receive DNN model 109 is reconfigured according to the NNSI.

[0128] In step 448, the second device 104b configures (if it is the first time) or reconfigures the receiving DNN model 109 of the second device 104b according to the NNSI to process the received communication data signal 119 according to the scrambling timing information or a specific time slot. Proceed to step 450.

[0129] For example, NNSI includes data representing the randomization of the order of inputs, outputs, or a set of neural network nodes of one or more specified neural network layers of the transmitting DNN model 107 of the first device 104a. Reconfiguration of the receiving DNN model 109 of the second device 104b further includes using NNSI to reconfigure the receiving DNN model 109 of the second device 104b to reverse the randomization applied to the specified neural network layers of the transmitting DNN model 107.

[0130] In step 450, the second device 104b uses the receiving DNN model 109 to process the received communication data signal 119 in order to reconstruct the communication data represented by the received communication data signal 119.

[0131] In step 452, the second device 104b sends the reconstructed communication data 101b' to the data sink of the protocol stack of the second device 104b and / or one or more upper layers.

[0132] Figure 5 An example DNN communication system 500 with a first device 510 communicating with a second device 520 is shown. In this example, the first device 510 includes a transmission DNN structure 506 (or TX DNN 506), a TX DNN controller 514, an output communication (OC) signal buffer 536a (also referred to as OC buffer 536a), an NNSI buffer 537a, and an RF front-end TX component 503a. Input communication data 501a to be transmitted in each specific time slot (e.g., time slot q(TS(q)), TS(q+1), ..., TS(q+Q) etc.) is applied to the input of the TX DNN 506. The TX DNN 506 generates corresponding OC signals 518a-518q for transmission in each of the specific time slots (e.g., TS(q), TS(q+1), ..., TS(q+Q) etc.). Each of the generated OC signals 518a to 518q passes through a TX DNN controller 514, which operates as a gate to determine whether OC signal 518a should be passed to the RF front-end TX component 503a for transmission in TS(q) as a transmission signal 505a that satisfies the white noise interference level. The TX DNN controller 514 implements the scrambling DNN operation performed at the first device 510, as referenced... Figure 1 a to Figure 4bAs described and / or as described herein. If the TX DNN controller 514 detects that the generated OC signal 518a will meet the white noise interference level when the first device 104a transmits it as a transmission signal 505a in TS(q), then the TX DNN controller 514 sends the OC signal 518a to the OC buffer 536a for transmission in a specific time slot TS(q). If the TX DNN controller 514 detects that the generated OC signal will not meet the white noise interference level if the first device 504a transmits the generated OC signal 518a as a transmission signal 505a in a specific TS(q), then the TX DNN controller 514 does not allow the generated OC signal 518a to enter the OC buffer 536a. Instead, the TX DNN controller 514 selects the NNSI for a specific time slot TS(q) to reconfigure the TX DNN 506 to generate an OC signal 518a for the specific time slot TS(q), which is predicted to satisfy the white noise interference level when transmitted in the specific time slot TS(q).

[0133] For example, the TX DNN controller 514 selects an NNSI signal 527a from an NNSI table (not shown) in a storage device, where each NNSI entry is associated with a specific white noise interference level characteristic, which may ensure that the resulting OC signal 518a meets the white noise interference level when transmitted. The TX DNN controller 514 guides the TX DNN 506 to be reconfigured for scrambling DNN operation based on the selected NNSI signal 527a, as referenced. Figure 1 a to Figure 4bAs described and / or as described herein, the reconfigured TX DNN 506 reprocesses the input communication data 501a for a specific time slot TS(q) to generate an updated or rescrambled OC signal 518a. If the TX DNN controller 514 detects that the generated OC signal 518a will still not meet the white noise interference level if the first device 510 transmits it as a transmission signal 505a, the TX DNN controller 514 iteratively performs the selection of a new NNSI signal until the OC signal 518a meets the white noise interference level when it is transmitted. If the TX DNN controller 514 detects that the generated OC signal 518a will now meet the white noise interference level if the first device 510 transmits it as a transmission signal 505a, the TX DNN controller 514 sends the OC signal 518a to the OC buffer 536a for transmission in the specific time slot TS(q). The TX DNN controller 514 also sends the corresponding selected NNSI signal 527a to the NNSI buffer 537a. Each NNSI signal 527a includes data indicating the specific time slot to which it applies (e.g., TS(q)) and an indication for reconfiguring the randomization / scrambling parameters of the TX DNN 506, which the RX DNN controller 524 can use to reconfigure the RX DNN 508 to descramble and generate reconstructed communication data 501b. The first device 510 transmits each NNSI signal 527a in the NNSI buffer 537a to the second device 520 in a control message 505b at an appropriate time (e.g., before the transmission of the corresponding OC signal 518a) to instruct the second device 520 when it should reconfigure the RX DNN 508 to process and descramble the corresponding received OC signal for the specific time slot TS(q) to generate reconstructed communication data 501b for the specific time slot TS(q). Typically, the first device 510 transmits the selected NNSI signal 527a for a specific time slot TS(q) before transmitting the OC signal 518a in that time slot. In other examples, the first device 510 transmits the selected NNSI signal 527a in a control message after the transmission of the OC signal 518a, wherein the second device 520 uses the received NNSI signal 527a to reconfigure the RX DNN 508 before processing the corresponding received OC signal.

[0134] Each of the OC signals 518a-518q and / or NNSI signals 527a-527p in the OC buffer 536a and NNSI buffer 537a is output for RF processing at an appropriate time. For example, each of the OC signals 518a-518q will be RF processed by the RF front-end TX component 503a in a specific time slot TS(q), TS(q+1), ..., TS(q+Q) of the OC signal for transmission, and will be transmitted as a data transmission signal 505a on the data communication channel. For example, each of the NNSI signals 527a-527p will be RF processed by the RF front-end TX component 503a for transmission before the specific time slot TS(q), TS(q+1), ..., TS(q+Q) used for the transmission of the OC signals 518a-518q, and will be transmitted as a control message 505b on the control communication channel. In another example, each of the NNSI signals 527a-527p will be RF-processed by the RF front-end TX component 503a for transmission before the second device 520 processes the corresponding received OC signals 518a-518q, and will be transmitted as one or more control messages 505b on the control communication channel.

[0135] For example, when the first device 510 is a base station and the second device 520 is a user equipment, the data communication channel includes a downlink data channel, such as the Physical Downlink Shared Channel (PDSCH). Alternatively, when the first device 510 is a user equipment and the second device 520 is a base station, the data communication channel includes an uplink data channel, such as the Physical Uplink Shared Channel (PUSCH). Similarly, when the first device 510 is a base station and the second device 520 is a user equipment, the control communication channel includes a downlink control channel, such as the Physical Downlink Control Channel (PDCCH). Alternatively, when the first device 510 is a user equipment and the second device 520 is a base station, the control communication channel includes an uplink control channel, such as the Physical Uplink Control Channel (PUCCH).

[0136] The second device 520 includes a receiving DNN structure 508 (or RX DNN 508), an RX DNN controller 524, a received output communication (Rx OC) signal buffer 536b (also referred to as Rx OC buffer 536b), an NNSI buffer 537b, and an RF front-end RX component 503b. The RF front-end RX component 503b receives a transmission signal 505a on the data communication channel in a specific time slot (e.g., TS(Q+1)) and processes the received transmission signal 505a to generate a baseband receive OC data signal (Rx OC signal) 519a. The RF front-end RX component 503b feeds the Rx OC signal 519a into the Rx OC buffer 536b for a specific time slot (e.g., TS(Q+1)). In this example, the Rx OC buffer 536b is a first-in-first-out (FIFO) queue, but it can be any type of buffer / queue implementation. Occasionally, the RF front-end RX component 503b also receives control messages 505b transmitted on the control communication channel, including an NNSI signal 529a for a specific time slot (e.g., TS(Q+1)), and processes the received control messages 505b to generate the NNSI signal 529a. The RF front-end RX component 503b feeds the NNSI signal 529a to the NNSI buffer 537b for the specific time slot (e.g., TS(Q+1)). The Rx OC buffer 536b is connected to the RX DNN controller 524, which receives from the Rx OC buffer 536b an Rx OC signal 519q corresponding to a specific time slot (e.g., TS(Q+P)). The RX DNN controller 524 implements the inverse descrambling DNN operation performed at the first device 510, as referenced. Figure 1 a to Figure 4d As described and / or as described herein.

[0137] For example, the RX DNN controller 524 controls when to input the Rx OC signal 519q for a specific time slot TS(Q+P) into the RX DNN 508 to generate reconstructed communication data 501b corresponding to the input communication data 501a for the specific time slot TS(Q+P). In this case, the RX DNN controller 524 recognizes the presence of an NNSI signal 529q for the specific time slot TS(Q+P), which is used to reconfigure the RX DNN 508 to descramble the Rx OC signal 519q received in time slot TS(Q+P). Before processing the Rx OC signal 519q for time slot TS(Q+P), the RX DNN controller 524 uses the NNSI signal 529q for time slot TS(Q+P) to reconfigure the RX DNN 508. After reconfiguration, the RX DNN controller 524 inputs the RxOC signal 519q into the reconfigured RX DNN 508 to generate reconstructed communication data 501b for a specific time slot TS(Q+P). The second device 520 sends the reconstructed communication data 501b for the specific time slot TS(Q+P) to its data sink, or sends the reconstructed communication data 501b to one or more upper protocol layers of the protocol stack of the second device 520.

[0138] As can be seen, the second device 520 receives several NNSI signals 529a, 529b, and 529p in the NNSI buffer 537b, and these NNSI signals correspond to Rx OC signals 519a, 519b, and 519p for specific time slots TS(Q+1), TS(Q+2), and TS(Q+P-1). The RX DNN controller 524 reconfigures the RX DNN 508 using the corresponding NNSI signals 529a, 529b, 529o, and 529p before processing each of the corresponding Rx OC signals 519a, 519b, and 519p. The TX DNN controller 514 may not need to update the NNSI signals for each time slot, because the TX DNN controller 514 may find that the same NNSI applies to multiple consecutive time slots. In this case, the RX DNN controller 524 only needs to reconfigure the RX DNN 508 when it receives a control message with updated NNSIs for some subsequent time slots. For example, there may be no further NNSI changes between TS(Q+P-2) and TS(Q+2), so the RX DNN controller 524 only needs to use the NNSI signal 529o to reconfigure the RXDNN 508 until the Rx OC signal 519b will be processed at TS(Q+2), so the Rx OC signal 519c up to but not including the Rx OC signal 519p is processed by the same configuration of the RX DNN 508 for time slots TS(Q+3) to TS(Q+P-2).

[0139] Figure 6a An example permutation operation 600 is shown for performing a random permutation on a specific order of one or more neural network nodes 641 of a neural network layer transmitting a DNN model to perform a scrambling DNN operation, and for performing a random inverse permutation (also known as a descrambling permutation) on a specific order of a set of neural network nodes 642 of a neural network layer receiving a DNN model to perform a descrambling DNN operation.

[0140] In this example, a set of neural network nodes 641 of a specific neural network layer of the transmitted DNN model has a specific ordering, for example, indicated by node labels 1, 2, 3, 4, and 5. The transmitted DNN model is reconfigured by randomizing this specific ordering of the set of neural network nodes 641 of the specific neural network layer using a random permutation operation 616. The random permutation operation 616 permutes or randomizes the specific ordering of the neural network nodes 641, thereby producing a permuted or randomized ordering of the neural network nodes 642 of that specific neural network layer of the transmitted DNN model. After performing the random permutation operation 616, the specific ordering of the set of neural network nodes 642 of that specific neural network layer is indicated, for example, by the reordered node labels 2, 1, 3, 5, and 4. Before the random permutation operation 616, the input connection to neural network node 1 is now fed to neural network node 2; before the random permutation operation 616, the input connection to neural network node 2 is now fed to neural network node 1; before the random permutation operation 616, the input connection to neural network node 3 is still fed to neural network node 3; before the random permutation operation 616, the input connection to neural network node 4 is now fed to neural network node 5; and before the random permutation operation 616, the input connection to neural network node 5 is now fed to neural network node 4. Even though the outputs of this group of neural network nodes 642 (e.g., nodes 1, 2, 3, 4, and 5) are still connected to the same neural network nodes in other neural network layers of the reconfigured transport DNN model, the random permutation of the inputs to this group of neural network nodes 642 results in scrambling of the output of the resulting transport DNN model. This can be referred to as a scrambling DNN operation.

[0141] For a specific random permutation operation 616, the NNSI includes the necessary data for inverting the random permutation operation 616, for use by the second device in reconfiguring the corresponding receiving DNN model to invert the random permutation operation 616. The second device uses the NNSI to identify the reciprocal neural network layers of the receiving DNN model and reconfigures the receiving DNN model using a random inverse permutation operation or a random depermutation operation 617 with respect to a specific ordering of the set of neural network nodes of the identified neural network layer of the receiving DNN model. In this example, a set of neural network nodes 643 of a specific neural network layer of the receiving DNN model has a specific ordering, for example, indicated by node labels 2, 1, 3, 5, 4. The receiving DNN model is reconfigured by randomizing the specific ordering of the set of neural network nodes 643 of the specific neural network layer using the random depermutation operation 617. The random depermutation operation 617 performs a depermutation on the specific ordering of the neural network nodes 643, thereby producing a depermutated or derandomized ordering of the neural network nodes 644 of the specific neural network layer of the receiving DNN model. After performing the random permutation operation 617, the specific ordering of the set of neural network nodes 644 in a particular neural network layer is indicated by the reordered node labels 1, 2, 3, 4, and 5. This means that the input connection to neural network node 1 before the random permutation operation 617 is now input to neural network node 2, the input connection to neural network node 2 before the random permutation operation 617 is now input to neural network node 1, the input connection to neural network node 3 before the random permutation operation 617 remains input to neural network node 3, the input connection to neural network node 4 before the random permutation operation 617 is now input to neural network node 5, and the input connection to neural network node 5 before the random permutation operation 617 is now input to neural network node 4. Even though the outputs of neural network nodes 1, 2, 3, 4, and 5 are still connected to the same neural network nodes in other neural network layers of the reconfigured receiving DNN model, the random permutation of the inputs to this set of neural network nodes 644 results in the descrambling of the output of the resulting receiving DNN model. This can be referred to as the descrambling DNN operation.

[0142] NNSI can include, for example, randomization parameters / functions for randomizing the order of the inputs, outputs, or a set of neural network nodes of one or more specified neural network layers to be randomized in the transmitting DNN. The one or more specific neural network layers of the transmitting DNN include, but are not limited to, selections of one or more of the following: the input neural network layer (or input layer), the output neural network layer (or output layer), and / or the hidden neural network layers of the transmitting DNN model, and / or combinations thereof. The specified neural network layers of the transmitting DNN model are randomized / scrambled. NNSI is used to reconfigure the transmitting and receiving DNN models to perform scrambled DNN operations.

[0143] The second device uses the NNSI to reconfigure the receive DNN model synchronously with the reception of the output communication signal representing the reconfigured transmit DNN model. To perform random permutation operations, the NNSI further includes data representing one or more random permutation parameters and / or functions, which include one or more of the following: seed data, random permutation iterations or sequence numbers, or identification of a seed generation and pseudo-randomization function for performing a specific ordering of random permutations of a set of neural network nodes in a specified neural network layer of the transmit DNN model. Specific seed data may include, but is not limited to, for example, the identifier of the second device (e.g., UE), the identifier of the first device (e.g., BS), the identifier of the cell in which the second device is located, the identifier of the cell in which the first device is located, time slot information, frame identification number, and / or any other information associated with the first or second device, a specific random permutation iteration, or a sequence number. For example, an initial seed or initial seed value can be generated from the seed data using seed generation and pseudo-randomization functions to randomize and / or derandomize neural network nodes in a specified one or more neural network layers of the transmit DNN model (TX DNN) and / or the receive DNN model (RX DNN), respectively.

[0144] As an example, the second device receives an NNSI including one or more random permutation parameters, such as seed data, random permutation iterations or numbers, or identification of seed generations, an indication of one or more neural network layers scrambled at the first device, and a pseudo-randomization function for performing random permutations on the set of neural network nodes of the indicated one or more neural network layers. The second device also receives corresponding timing information associated with when to reconfigure the RX DNN (e.g., descramble the corresponding neural network layers of the RX DNN). The timing information includes one or more time slots. At a specific time slot, reconfiguring the RX DNN of the second device may include: generating an inverse random permutation sequence corresponding to the random permutation iteration or number for each neural network layer of the RX DNN corresponding to each indicated neural network layer of the TX DNN. The inverse random permutation sequence has a length equal to the number of neural network nodes in the set of neural network nodes of each neural network layer of the RX DNN. For each neural network layer of the RX DNN corresponding to each neural network layer of the TX DNN, the set of neural network nodes of each neural network layer of the RX DNN is derandomized or inverse permuted by applying the generated inverse random permutation sequence to the set of neural network nodes. Alternatively, an inverse random permutation matrix can be generated and applied to the sorting of the nodes in the neural network and similar methods.

[0145] Figure 6bAn example random permutation sequence 650 is shown, starting from an initial seed 652 (also referred to as the initial seed value 652). For a specific neural network layer with a set of N neural network nodes, the initial seed 652 is input to a random permutation function to generate a first random permutation sequence 651a of size N. The first random permutation sequence 651a can be a permutation of integers in the range [1, N], where the integers represent the node labels of the set of N neural network nodes. The first random permutation sequence 651a is generated by permutation iteration 1 of the random permutation function with the initial seed 652. Further iterations or loops of the random permutation function with the initial seed 652 generate different random permutation sequences of size N, such as, for example, the second random permutation sequence 651b in the second permutation iteration 2, the third random permutation sequence 651c in the third permutation iteration 3, ..., the i-th random permutation sequence 651i in the i-th permutation iteration i, ..., the n-th random permutation sequence 651n in the n-th permutation iteration n, and so on. Given an initial seed value of 652 and a random permutation function pair, the i-th random permutation sequence (or the i-th cycle of a random permutation sequence of length N) can be generated by simply specifying the i-th random permutation iteration number and iterating through the random permutation sequence generated by using the initial seed 652 and the random permutation function until the i-th cycle.

[0146] As an example, the initial control message may include an already referenced structure containing an initial seed value, a random permutation function indication, a random permutation iteration count, and the transmitted DNN model. Figure 1 a to Figure 6a The NNSI describes the randomized data of one or more specified neural network layers. The first device transmits an initial control message to the second device to reconfigure the corresponding neural network layer of the receiving DNN model using the received initial seed value, random permutation function indication, random permutation iteration number, and specified neural network layer, as shown in the reference. Figures 3a to 3c The method described is used to generate a corresponding random permutation sequence associated with the number of random permutation iterations. The generated random permutation sequence generates a random depermutation (or inverse permutation) operation 617 for reconfiguring the corresponding neural network layer of the receiving DNN model at the second device, for descrambling the DNN operation and generating reconstructed communication data. The second device sends the reconstructed communication data to its data sink, or to one or more upper protocol layers of the protocol stack of the second device. Furthermore, the first device sends a subsequent update to the NNSI for reconfiguring the transmission DNN model in a subsequent control message, which includes an indication of the updated number of random permutation iterations. These are transmitted in further control messages or one or more further control messages when the initial seed and / or random permutation function are updated or changed.

[0147] Figure 6c The permutation and unpermutation (or inverse permutation) operations 660 are illustrated, wherein the random permutation operation 616 uses a permutation method derived from the permutation method. Figure 6b The selected i-th random permutation sequence 652i generates a random permutation matrix 616', which is used to scramble / descramble one or more neural network layers of the transmitting / receiving DNN model. The i-th random permutation sequence 652i is used to generate the i-th N×N random permutation matrix P. i 616'. For example, the i-th N×N random permutation matrix P is computed in the following way. i 616': Generate an N×N identity matrix I, where each column is labeled consecutively from 1 to N; and then use the generated i-th random permutation sequence 652i to permutate the columns of the identity matrix I. This produces a random permutation matrix P. i 616', this random permutation matrix represents the i-th random permutation sequence 652i. By using the random permutation matrix P... i Multiplying 616' with an N-dimensional vector performs a random permutation operation 616, resulting in a permuted N-dimensional vector permuted according to the i-th random permutation sequence 652i. The random permutation matrix P has already been computed. i After 616', based on taking the permutation matrix P i The matrix inverse of 616' is used to derive the random solution permutation operation 617, thereby producing the solution permutation matrix D. i 617', where D i =(P i ) -1 This can be achieved by permuting the solution matrix D. i 617' is multiplied by the N-dimensional vector that needs to be deperformed to perform a random deperformation operation 617, thereby producing a deperformed N-dimensional vector.

[0148] For example, for the selected NNSI indicating the number of iterations of the i-th random permutation sequence 652i, the i-th neural network layer of the transmitting DNN model is reconfigured as follows, where 1 ≤ l ≤ L and l = 1 is the input layer, l = L is the output layer, and L is the number of neural network layers including one or more hidden neural network layers: generating the i-th random permutation sequence 652i; using the i-th random permutation sequence 652i to generate an N-dimensional random permutation matrix P for the i-th neural network layer. i 616', N>1 and N is the number of neural network nodes in the I-th neural network layer; and a specific sorting vector representing the current specific sorting of the N neural network nodes in the I-th neural network layer and an N-dimensional random permutation matrix P. i Multiplying by 616' produces a randomized sorted vector representing a permuted set of N neural network nodes.

[0149] For example, the first device selects an NNSI that indicates the number of iterations of the i-th random permutation sequence 652i and indicates that the i-th neural network layer of the transmitting DNN model is to be reconfigured. As described herein, the first device sends the selected NNSI to the second device in a control message. Given the selected NNSI, the second device reconfigures the (L-1+1)-th neural network layer receiving the DNN by: generating the i-th random permutation sequence 652i; and using the i-th random permutation sequence 652i to generate an N-dimensional random permutation matrix P for the (L-1+1)-th neural network layer. i 616', N>1 and N is the number of neural network nodes in the (L-l+1)th neural network layer; by applying the N-dimensional random permutation matrix P i The inverse of 616' is used to generate the N-dimensional solution matrix D. i 617', and the specific sorting vector representing the current specific sorting of the N neural network nodes of the (L-l+1)th neural network layer is combined with the N-dimensional solution permutation matrix D. i Multiplying by 617' produces a sorted vector (or inverse random permutation sequence) representing a set of N neural network nodes that have undergone unscrambled permutations. This vector is used to reconfigure the (L-1+1)th neural network layer of the receiving DNN model.

[0150] Figure 6d An example iterative NNSI selection process 670, performed by a first device, is illustrated for selecting an i-th random permutation sequence to perform a randomization operation to randomize the order of neural network nodes in one or more neural network layers of a transmitted DNN model. The i-th random permutation sequence produces a reconfigured transmitted DNN model that, when given input communication data, generates an output communication signal that satisfies a white noise interference level during transmission. The iterative NNSI selection process 670 performed by the first device includes one or more of the following steps:

[0151] In step 672, the first device selects or generates an i-th random permutation sequence (or i-th randomization operation) for randomizing one or more neural network layers of the transmitted DNN model. For example, the first device selects an i-th random permutation iteration from an NNSI table mapping to characteristics of white noise interference levels that satisfy the current white noise interference level. Alternatively, the first device generates an i-th random permutation iteration (e.g., selects a value i) to generate the i-th random permutation sequence. The i-th random permutation iteration (selected or generated) is used to generate an i-th cycle (referred to as the i-th random permutation sequence) of random permutation sequences using an initial seed and a specific random permutation function. The i-th random permutation sequence has a length N > 1, equal to the number of neural network nodes in the set of neural network nodes of a specific neural network layer of the transmitted DNN model. The i-th random permutation sequence is used to randomly permutate the set of neural network nodes of the specific neural network layer.

[0152] In step 674, the first device reconfigures the transmission DNN model using the selected / generated i-th random permutation sequence from step 672.

[0153] In step 676, the first device processes the input communication data using a reconfigured transmission DNN model to generate an output communication signal for transmission.

[0154] In step 678, the first device analyzes and determines whether the transmission of the generated output communication signal will satisfy the white noise interference level. This may include: if the first device transmits the generated output communication signal, analyzing and / or estimating the spectral density of the predicted transmitted signal waveform. When the generated output communication signal is determined to produce a predicted transmitted signal waveform that satisfies the white noise interference level if transmitted (e.g., 'yes'), the process continues to step 682; otherwise (e.g., 'no'), the process continues to step 680.

[0155] For example, as referenced Figure 1 a to Figure 4b ,in particular Figures 3a to 3b and Figure 4b The description pertains to the determination in step 678, which involves analyzing the spectral density of a predicted transmission signal generated from the processed output communication signal for transmission. A first device (or a component thereof) automatically analyzes the output communication signal to determine whether the transmission of the output communication signal will satisfy a white noise interference level. For example, a trained spectral density estimation model processes the output communication signal to predict or estimate the power spectral density of the transmission signal generated from the output communication signal. In another example, the output communication signal is input into a simulation model to generate the predicted transmission signal, wherein, as referenced... Figures 3a to 4bThe power spectral density of the transmitted signal is predicted by analysis of the white noise level, as described and / or as described herein. In another example, a simulation of the transmission of the output communication signal over an analog communication channel is performed, wherein the simulation performs an analysis of whether the simulated transmission meets the white noise interference level and indicates the results.

[0156] In step 680, the first device updates the iteration number i to i = i + 1, and the iterative NNSI selection process 670 continues to step 672 for generating / selecting another i-th random permutation sequence by the first device. That is, for the next i-th iteration, the generation, reconfiguration processing, and analysis steps 672, 674, 676, and 678 are repeated.

[0157] In step 682, the first device instructs the selected / generated i-th permutation sequence or i-th permutation iteration to reconfigure the transmission DNN model to generate an output communication signal that satisfies the white noise interference level when transmitted.

[0158] For example, in step 682, indicating the selected / generated i-th permutation sequence includes indicating the i-th random permutation iteration to be included in the NNSI, wherein the first device sends the NNSI including the i-th random permutation iteration to the second device in a control message, as described herein. Initially, the NNSI includes at least an initial seed, a random permutation function, the i-th random permutation iteration, and one or more neural network layers of the transport DNN model for generating the i-th random permutation sequence for reconfiguring the transport DNN. Subsequent updates to the NNSI may include indications of the selected / generated i-th random permutation iteration.

[0159] In addition to selecting the i-th random permutation sequence / i-th random permutation iteration, step 672 further includes selecting a set of one or more neural network layers by the first means, wherein the selected set of one or more neural network layers is different from the previous selection of one or more neural network layers. NNSI includes the selected one or more neural network layers that produce an output communication signal that satisfies the white interference level when transmitted.

[0160] The iterative NNSI selection process 670 is performed by the first device or its controller component (e.g., Figure 2aThe process is performed using the BS DNNC 214, wherein, based on the analysis performed in step 678, the mapping between the whitening characteristics of the output communication signal during transmission and the i-th random permutation sequence / i-th random permutation iteration and / or other NNSI data (e.g., one or more selected neural network layers) can be used to populate an NNSI lookup table accessible to the first device. Further modifications include, for example, in step 672, the first device selecting the i-th random permutation sequence by retrieving from the NNSI lookup table a random permutation sequence or random permutation iteration that maps to white noise interference characteristics that may satisfy the white noise interference level. This is based on the corresponding whitening characteristics indicating the likelihood that the selected NNSI will produce the transmission of an output communication signal that satisfies the white noise interference level for the transmitted DNN. In another example, the first device uses the NNSI table to guide the iterative NNSI selection process 670 when searching, for example, the i-th random permutation sequence / i-th random permutation iteration and / or other NNSI data (e.g., one or more selected neural network layers), thereby producing an output communication signal that satisfies the white noise interference level during transmission.

[0161] Optionally, the one or more neural network layers of the transmitting DNN model include one or more neural network layers from the group consisting of: an input neural network layer, an output neural network layer, and one or more hidden neural network layers. One or more neural network layers can be predefined or pre-selected before the scrambling DNN operation. The selection of one or more neural network layers can take into account the capabilities of the first and / or second devices. For example, if the second device does not have the capability to descramble the hidden neural network layers, the input and / or output neural network layers can be selected to reduce complexity and / or computational resource consumption. Alternatively or additionally, one or more neural network layers can be randomly selected before the scrambling DNN operation and / or when generating a new NNSI for whitening the transmitted signal.

[0162] Figure 7 An example signal stream of scrambled DNN communication 720 during a communication session between the first device 104a and the second device 104b is shown. Figure 1 The first device 104a and the second device 104b are used as referenced. Figures 1 to 6c Any aspect described can be used to perform scrambled DNN communication 720. For example, the first device 104a and the second device 104b are shown in Figure 2 as base station 210 and user equipment 220, or as... Figures 3a to 3c The first device 310 and the second device 320 perform scrambled DNN communication 720. The signal flow of the scrambled DNN communication 720 in the communication session between the first device 104a and the second device 104b includes the following signal flow operations:

[0163] In operation 721a, the first device 104a and the second device 104b establish a DNN connection during a communication session between them. During the establishment of the DNN connection, the first device and the second device communicate with each other to define, agree on, and / or configure the types of transmit DNN structures and receive DNN structures used to perform end-to-end communication therebetween.

[0164] For example, after the first device 104a initiates a standard communication session with the second device 104b, the first device 104a selects the type of transport DNN structure to be used in the DNN connection with the second device 104b, depending on the communication channel conditions / environment, the communication performance requirements of the DNN connection, and the type of data communication being performed (e.g., voice communication, data communication, multimedia streaming, and the like). The first device 104a requests the machine learning processing capabilities of the second device 104b to assist in the selection of the transport DNN structure and / or the receive DNN structure to be used during the DNN connection. The first device 104a accesses a DNN lookup table (or neural network table) that includes a set of transport / receive DNN structures / models and / or pairs thereof mapped to DNN identifiers stored in the table. In this example, the first device 104a stores the DNN lookup table thereon. The second device 104b accesses a corresponding DNN lookup table that has a corresponding set of transport / receive DNN structures / models also mapped to the same DNN identifiers. In the example, the corresponding DNN lookup table stores each receiving DNN structure / model and a mapping to the corresponding DNN identifier of the transmitting / receiving DNN structure / model pair stored in the DNN lookup table accessible by the first device 104a. In the example, the second device stores the corresponding DNN lookup table thereon.

[0165] In this example, the first device 104a selects a suitable transmit DNN structure / model and / or corresponding receive DNN structure / model for the DNN connection. After selecting the transmit DNN structure and / or the corresponding receive DNN structure, the first device 104a initiates a DNN connection by sending a DNN connection request message to the second device 104b. This DNN connection request message has fields requesting the establishment of a DNN connection, the selected transmit DNN structure / receive DNN structure and / or its corresponding DNN identifier, or an indication of the type of receive DNN structure that the second device 104b should use and / or the type of transmit DNN structure that the first device 104a will use. Therefore, the second device 104b selects an appropriate receive DNN structure. In some examples, when performing full-duplex communication, i.e., when performing both downlink and uplink communication, the first device 104a has a transmit DNN structure and a receive DNN structure for communication to and from the second device 104b, and the second device 104b also has a corresponding receive DNN structure and a corresponding transmit DNN structure for communication to and from the first device 104b. The first device 104a and the second device 104b establish a DNN connection, wherein the first device 104a configures a selected transmit DNN structure (and / or its receive DNN structure) for DNN communication with the second device 104b on a communication channel in one or more time slots. The second device 104b configures a corresponding receive DNN structure (and / or its corresponding transmit DNN structure) for DNN communication with the first device 104a on a communication channel in one or more time slots.

[0166] In operation 721a, the white noise interference level is also initially set by the first device 104a, or during the establishment of the DNN connection, according to a request from the second device 104b to improve block error rate performance (e.g., to request an increase in the white noise interference level).

[0167] In operation 721b, after establishing DNN connections with a first device 104a operating a transmission DNN structure (and / or a reception DNN structure) and a second device 104b operating a corresponding reception DNN structure (and / or a corresponding transmission DNN structure), the first device 104a uses the configured transmission and reception DNN structures of the first device 104a and the second device 104b, respectively, to perform DNN communication with the second device 104b for one or more time slots. In this example, the first device 104a uses the transmission DNN structure to process input communication data for transmission to the second device 104b, which uses the reception DNN structure to process the received transmission to generate reconstructed communication data. Alternatively, the second device 104b uses the transmission DNN structure to process input communication data for transmission to the first device 104a, which uses the reception DNN structure to process the received transmission to generate reconstructed communication data. The second device 104b sends the reconstructed communication data to its data sink, or sends the reconstructed communication data to one or more upper protocol layers of its protocol stack.

[0168] In operation 792, when the first device 104a detects that the transmission from the first device 104a does not meet a specific white noise interference level, the first device 104a and the second device 104b perform DNN scrambling operations. For scrambled DNN communication (e.g., DL DNN scrambling) from the first device 104a to the second device 104b, the TX DNN controller of the first device 104a controls the scrambling DNN operation of the first device 104a, and the RX DNN controller of the second device 104b controls the descrambling DNN operation of the second device 104b, as referenced. Figures 1 to 6d (Specifically, Figures 4a to 5 As described above. For scrambling DNN communication (e.g., UL DNN scrambling) from the second device 104a to the first device 104b, the TX DNN controller of the second device 104b controls the scrambling DNN operation of the second device 104b, and the RX DNN controller of the first device 104a controls the descrambling DNN operation of the first device 104a. For simplicity and by way of example only, the following steps describe scrambling DNN communication from the first device 104a to the second device 104b, wherein similar or identical operations apply to scrambling DNN communication from the second device 104a to the first device 104b, wherein the first device 104a and the second device 104b exchange roles.

[0169] In this example, the white noise interference level is set by the first device 104a according to a request from the second device 104b for improved block error rate performance (e.g., a request for an increase in the white noise interference level), or according to a request from a third device (not shown) experiencing intolerable white noise interference caused by transmissions from the first device 104a during DNN communication. The first device 104a and the second device 104b perform the scrambling DNN operation in operation 792 based on, but not limited to, the following scrambling DNN operation:

[0170] In operation 722a, when the output communication signal from the transmission DNN structure does not meet the white noise interference level during transmission, the first device 104a enables the execution of scrambling DNN operation for one or more time slots. For example, the TX DNN controller of the first device 104a detects that the transmission DNN model of the transmission DNN structure generates an output communication signal from the input communication data, which will produce a transmission signal that does not meet the white noise interference level. In this case, scrambling DNN operation is enabled, and the TX DNN controller of the first device 104a, as referenced... Figure 1 Figure 2 and Figures 3a to 6d (in particular Figures 3a to 6d The NNSI described above is used to reconfigure the transmission DNN structure of the transmission DNN model (referred to as reconfiguring the transmission DNN structure) to generate an output communication signal from the same input communication data, which will produce a transmission signal that meets the white noise interference level.

[0171] In operation 723a, the first device 104a communicates when and how to enable scrambling DNN operation by sending a control message to the second device 104b. This control message includes a selected NNSI and scrambling timing information regarding when the receive DNN structure at the second device 104b should be reconfigured according to the scrambling DNN operation. For example, the first device 104a sends a control message to the second device 104b containing data representing the selected NNSI and indications of one or more time slots indicating when the first device 104a uses the selected NNSI to reconfigure the scrambling timing information of the transmit DNN structure. Initially, the selected NNSI includes information representing initial seed information, the type of random permutation function, one or more neural network layers of the transmit DNN structure to be reconfigured for scrambling, data of random permutation iterations, and any other data enabling the second device 104b to reconfigure the receive DNN structure (referred to as reconfiguring the receive DNN structure) for descrambling and generating reconstructed communication data. Subsequent control messages for further scrambling DNN operations include updated NNSI, such as, for example, the number of selected random permutation iterations, or one or more selected neural network layers that have been reconfigured for scrambling.

[0172] In operation 723b, the second device 104b acknowledges receipt (e.g., ACK) of the selected NNSI configuration and the successful enabling of scrambling DNN operation at the second device 104b.

[0173] In operations 724a and 724b, the first device 104a and the second device 104b initiate a scrambling DNN operation based on scrambling timing information. For example, for a scrambling DNN operation from the first device 104a to the second device 104b, the first device 104a prepares to perform a scrambling TX DNN operation at the first device 104a for transmission to the second device 104b, and the second device 104b prepares to perform a descrambling RX DNN operation at the second device 104b to receive transmissions from the first device 104a based on scrambling timing information (e.g., one or more time slots). For example, the scrambling TX DNN operation includes the first device 104a reconfiguring the transmission DNN structure based on the NNSI selected for each specific time slot associated with the scrambling timing information. The descrambling RX DNN operation includes the second device 104b reconfiguring the receive DNN structure based on the NNSI selected for each specific time slot before the receive DNN structure processes the transmissions arriving in each specific time slot associated with the scrambling timing information. Similarly, if there is a DNN scrambling operation from the second device 104b to the first device 104a (e.g., an uplink scrambling DNN operation), the second device 104b is prepared to perform a scrambling TX DNN operation similar to that described above with reference to the first device, and the first device 104a is prepared to perform a descrambling RX DNN operation as described above with reference to the second device 104b.

[0174] In operation 725, the first and second devices communicate with each other using corresponding transmit DNN and receive DNN structures, and perform scrambling DNN operations based on scrambling timing information (e.g., one or more specific time slots where scrambling is performed). For example, the first device 104a performs a scrambling TX DNN operation at the first device 104a to transmit to the second device 104b. The second device 104b performs a reciprocal descrambling RX DNN operation at the second device 104b based on the scrambling timing information (e.g., one or more time slots) to receive the transmission from the first device 104a. For example, the scrambling TX DNN operation includes the first device 104a reconfiguring the transmit DNN structure based on the NNSI selected for each specific time slot. The scrambling RX DNN operation includes the second device 104b reconfiguring the receive DNN structure based on the NNSI selected for each specific time slot before the receive DNN structure processes the transmissions arriving in each specific time slot. If the first device 104a and the second device 104b are performing scrambled DNN communication from the second device 104b to the first device 104a (e.g., uplink scrambled DNN operation), the second device 104b performs a scrambled TX DNN operation in a manner similar to that described above for the first device 104a, and the first device 104b performs a scrambled RX DNN operation in a manner similar to that described above for the second device 104b.

[0175] Furthermore, during communication between the first device 104a and the second device 104b, when performing a scrambling DNN operation, the first device 104a iteratively determines a subsequent NNSI for transmission in one or more subsequent time slots using a common / initial seed and a random function already transmitted in the first control message of operation 723a. This subsequent NNSI includes the minimum randomization information necessary to enable the second device 104b to reconfigure its receive DNN structure for processing transmissions received in subsequent time slots. For example, the subsequent NNSI includes, for instance, an identified random permutation sequence number or order. The first device 104a transmits the subsequent NNSI to the second device 104b in one or more fields of the subsequent control message. The subsequent NNSI also includes, for example, one or more specific neural network layers with randomization / scrambling. The subsequent scrambling DNN operation only requires an updated random permutation iteration or sequence number when the output communication signal generates a transmission signal that does not meet the white noise interference level and is used to synchronize the reconfiguration of the receive DNN structure of the second device 104b for processing transmissions received in subsequent time slots. The second device 104b reuses the common / initial seed data, common seed generation, and pseudo-randomization function instructions transmitted in the initial control message.

[0176] In operation 726, the first device 104a disables the scrambling DNN operation.

[0177] In operation 727a, the first device 104a sends a notification to the second device 104b that the scrambling DNN operation is disabled / turned off.

[0178] In operation 727b, the second device 104b acknowledges the receipt of the communication (e.g., ACK).

[0179] In operation 728a, upon receiving an acknowledgment (ACK) from the second device 104b in operation 727b, the first device 104a disables the scrambling DNN operation and restores the transmission DNN structure to its original configuration. If the first device 104a and the second device 104b are performing scrambling DNN communication from the second device 104b to the first device 104a (e.g., an uplink scrambling DNN operation), where the second device 104b has a transmission DNN structure and the first device 104b has a corresponding receive DNN structure, then the first device 104a also restores their receive DNN structures to their original configuration.

[0180] In operation 728b, when communication for disabling scrambling DNN operation is received in operation 727a, the second device 104b disables the scrambling DNN operation and restores the received DNN structure to its original configuration. If the first device 104a and the second device 104b are performing scrambling DNN communication from the second device 104b to the first device 104a (e.g., uplink scrambling DNN operation), where the second device 104b has a transmit DNN structure and the first device 104b has a corresponding receive DNN structure, then the second device 104b also restores their transmit DNN structures to their original configuration.

[0181] In operation 729, after the first device 104a and the second device 104b disable the scrambling DNN connection, the first device 104a and the second device 104b continue to perform DNN communication with the first device 104a operating the original transmit DNN structure (and / or receive DNN structure) and the second device 104b operating the original corresponding receive DNN structure (and / or corresponding transmit DNN structure). For example, the first device 104a uses the configured transmit and receive DNNs of the first device 104a and the second device 104b to perform DNN communication with the second device 104b for one or more time slots. In this example, the first device 104a uses the transmit DNN structure to process input communication data for transmission to the second device 104b, which uses the receive DNN structure to process the received transmission to generate reconstructed communication data. Alternatively, the second device 104b uses a transmit DNN structure to process input communication data for transmission to the first device 104a, which uses a receive DNN structure to process the received transmission to generate reconstructed communication data. The first device 104a and the second device 104b resume standard or regular communication sessions after terminating DNN communication.

[0182] Figure 8 It shows in Figure 7 An example signal stream of scrambled DNN operation 825 between the first device 104a and the second device 104b during operation 725 of the scrambled DNN communication 720 shown. The first device 104a and the second device 104b use, as referenced... Figures 1 to 6c To perform the function of any of the described aspects Figure 1 The scrambling DNN operation 825 is performed for communication from the first device 104a to the second device 104b. In other examples, the base station 210 and user equipment 220 of FIG2 perform the scrambling DNN operation 825. In another example, Figures 3a to 3c The first device 310 and the second device 320 of either of them perform the scrambling DNN operation 825, or Figure 5 The first device 510 and the second device 520, and the like, perform scrambling DNN operation 825. Although the first device 104a and the second device 104b perform scrambling DNN operation 825, this is merely illustrative and not limiting. Those skilled in the art will understand that scrambling DNN operation 825 is applicable to communication from the second device 104b to the first device 104a, wherein the roles of the first device 104a and the second device 104b may be interchanged for some or all of the operations of scrambling DNN operation 825. In this example, the first device 104a and the second device 104b establish a scrambling DNN operation as described in operation 792, and specifically as... Figure 7Operations 722a, 723a, 723b, 724a, and 724b are described. The signal flow of the scrambling DNN operation 825 from the first device 104a to the second device 104b includes the following signal flow operations:

[0183] In operation 802, the first device 104a retrieves input communication data from the data source for input into the transmission DNN structure of the first device 104a.

[0184] In operation 804, the transmission DNN structure of the first device 104a processes the input communication data to generate an output communication signal for transmission to the second device 104b in a specific time slot.

[0185] In operation 806, before the generated output communication signal is transmitted in a specific time slot, the first device 104a performs spectral analysis to determine whether the predicted transmission of the output communication signal will meet the white noise interference level. For example, Figure 1 , Figure 2a , Figures 2b to 2d and / or Figures 4a to 4d and / or Figure 5 And / or perform spectrum analysis as described. In response to the predicted transmission of the generated output communication signal not meeting the white noise interference level, operations 808 to 834 are performed; otherwise, scrambling DNN operation 825 continues with operation 816 for transmitting the output communication signal.

[0186] In operation 808, the first device 104a selects the updated NNSI for reconfiguring the transmission DNN structure to generate an output communication signal that satisfies the white noise interference level during transmission. The updated NNSI can be determined iteratively and / or from an NNSI lookup table, as referenced. Figures 1 to 6d As described. During the selection of the updated NNSI, the transmission DNN structure is reconfigured and the input communication data is reprocessed to generate the output communication signal. After determining / selecting the updated NNSI that causes the reconfigured transmission DNN structure to generate an output communication signal that meets the white noise interference level, the signal flow of scrambling DNN operation 825 continues to operation 810.

[0187] In operation 810, the first device 104a reconfigures the transmission DNN structure using an updated NNSI for a specific time slot.

[0188] In operation 814, the first device 104a transmits updated NNSI and specific timeslot information to the second device 104b. For example, the first device 104a transmits a control message to the second device 104b on a control channel. This control message includes an indication of the updated NNSI and specific timeslot information, which guides the second device 104a on when to reconfigure its receive DNN structure to process transmissions from the first device 104a corresponding to the specific timeslot information. Upon receiving the control message from the first device 104a, the second device 104b performs operation 834, whereby the second device 104b stores the updated NNSI and specific timeslot information in an NNSI buffer or NNSI table for use at the appropriate time associated with the specific timeslot information.

[0189] In operation 816, when the transmission of the generated output communication signal meets the white noise interference level, the first device 104a transmits the output communication signal of the reconfigured transmission DNN structure according to specific time slot information.

[0190] When an output communication signal is received in a specific time slot, the second device 104b can buffer the received output communication signal until it is ready for processing by the receiving DNN structure. In operation 848, before processing the received output communication signal corresponding to the specific time slot, the second device 104b reconfigures the receiving DNN structure using the updated NNSI for that specific time slot. The second device 104b retrieves the updated NNSI for the specific time slot from a storage device (e.g., from an NNSI buffer or an NNSI table) and uses it to reconfigure the receiving DNN structure, as shown, for example, by referring to... Figures 1 to 6d ,in particular Figures 3a to 3c , Figure 4c and Figure 4d as well as Figure 5 As described and / or as described herein. The reconfiguration of the receive DNN structure modifies the receive DNN structure to perform the inverse operation performed by the reconfigured transmit DNN structure, i.e., a descrambling operation, to generate reconstructed communication data corresponding to the input communication data transmitted in a specific time slot.

[0191] In operation 850, the received DNN structure processes the received output communication signal for a specific time slot and generates reconstructed communication data corresponding to the input communication data transmitted in the specific time slot. The second device 104b sends the reconstructed communication data to the data sink at the second device 104b, or sends the reconstructed communication data to one or more upper protocol layers of the protocol stack of the second device 104b.

[0192] In operation 852, upon successful generation of reconstructed communication data, the second device 104b sends an acknowledgment to the first device 104a indicating successful reception of the input communication data for a specific time slot. In operation 853, the second device 104b continues to prepare to receive the next transmission from the first device 104a in one or more subsequent time slots.

[0193] In operation 825a, when the first device 104a receives an acknowledgment in operation 852, it continues to the next transmission of further input communication data (if any) from the data source in one or more subsequent time slots, wherein scrambling DNN operation 825 is repeated until it is determined that scrambling DNN communication should be disabled, as... Figure 7 As described in operation 726.

[0194] Figure 9 It shows in Figure 7 The signal flow diagram of another example of the scrambling DNN operation 925 between the first device 104a and the second device 104b during operation 725 of the scrambling DNN communication 720 shown is illustrated. The scrambling DNN operation 925 is further modified by an insertion operation 917. Figure 8 The scrambling DNN operation 825 involves the first device 104a identifying a further NNSI that satisfies the white noise interference level between transmissions. The signal flow of the scrambling DNN operation 925 from the first device 104a to the second device 104b includes the following signal flow operations:

[0195] At the first device, operations 902-916 substantially correspond to Figure 8 Operations 802-816 in the middle. Similarly, at the second device, operations 934, 948, 950, 952 and 953 essentially correspond to Figure 8 Operations 834, 848, 850, 852, and 853 are described in the first device 104a. Following operation 916, the first device 104a performs operation 917, whereby the first device 104a continues to identify further NNSIs to aid in reconfiguring the transmission DNN structure for generating an output communication signal predicted to satisfy a white noise interference level during transmission. The first device 104a stores any identified NNSIs mapped to NNSI identifiers and the associated or estimated white noise interference level in an NNSI lookup table for use in operation 908 (or...). Figure 8 This can be used in operation 808. This can guide the search for the updated NNSI in operation 908, thus providing a faster selection of the updated NNSI, which will ensure that the generated output communication signal meets the white noise interference level when transmitted. In the example, the first device 104a performs operation 917 in the background and / or between operations 916 and 925a.

[0196] In operation 925a, when the first device 104a receives an acknowledgment in operation 952, it continues to the next transmission of further input communication data (if any) from the data source in one or more subsequent time slots, wherein scrambling DNN operation 925 is repeated until it is determined that scrambling DNN communication should be disabled, as... Figure 7 As described in operation 726.

[0197] Optionally, the first device 104a sends any identified NNSI, NNSI identifier, and / or associated or estimated white noise interference level updates to the second device 104b via one or more control messages for use by the second device 104b when updating its NNSI lookup table with the NNSI and NNSI identifier and the like. This provides the advantage that, for the NNSI selected by the first device 104a, the control message for reconfiguring the receive DNN structure of the second device 104b can send the NNSI identifier and timing information of the selected NNSI, wherein the second device 104b can retrieve the selected NNSI accordingly.

[0198] Figure 10 The signal flow of another example of scrambled DNN communication 1000 between the first device 104a, the second device 104b, and the third device 1003 is shown. The scrambled DNN communication 1000 is further modified by insertion operations 1005a and 1005b. Figure 7 The scrambled DNN communication 720, wherein the third device 1003 communicates with the first device 104a regarding a tolerable level of white noise interference. The signal flow of the scrambled DNN communication 1000 from the first device 104a to the second device 104b includes the following signal flow operations:

[0199] At the first device 104a and the second device 104b, operations 1021a, 1021b, 1092, 1026, 1027a, 1027b, 1028a, 1028b and 1029 substantially correspond as shown in the reference. Figure 7Operations 721a, 721b, 792, 726, 727a, 727b, 728a, 728b, and 729 are described. In operation 1021a, the white noise interference level is initially set by the first device 104a, or during the establishment of the DNN connection, according to a request from the second device 104b to improve block error rate performance (e.g., requesting an increase in the white noise interference level). After the first device 104a and the second device 104b have established a connection for DNN communication and performed DNN communication between them in operation 1021b, the third device 1003 may experience intolerable white noise interference caused by transmissions from the first device 104a (or the second device 104b) during DNN communication in operation 1021b.

[0200] In operation 1005a, the third device 1003 sends a notification to the first device 104a (or the second device 104b) indicating that the transmission of output communication signals from the first device 104a to the second device 104b exceeds an acceptable level of white noise interference associated with the third device 1003. In this example, the third device 1003 is a victim device experiencing interference from the transmissions of the first device 104a. In this example, when the first device 104a is a base station, the third device 1003 contacts the first device 104a via an uplink channel. In another example, the third device 1003 is also a base station of another cell and requests the first device 104a to reduce interference to other user equipment in its cell via the core network. In this example, the notification from the third device 1003 indicates a tolerable or acceptable level of white noise interference. In another example, the notification from the third device 1003 indicates that the current level of white noise interference output by the first device 104a is unacceptable.

[0201] In operation 1005b, when the first device 104a receives a notification from the third device 1003 in operation 1005a, the first device 104a adjusts the white noise interference level by reducing the white noise interference level. In one example, if the notification from the third device 1003 indicates a tolerable or acceptable white noise interference level, the first device 104a adjusts the white noise interference level to a tolerable or acceptable level. In another example, if the notification from the third device 1003 indicates that the current white noise interference level output by the first device 104a is unacceptable, the first device 104a adjusts the white noise interference level by incrementally decreasing the white noise interference level or until the third device 1003 stops notifying the reduction of the white noise interference level. For example, the third device 1003 (e.g., the victim UE) moves away from the first device 104a, so that the first device does not experience an unacceptable white noise interference level. In another example, as the third device 1003 moves away from the first device 104a, it sends further notifications to the first device 104a indicating a further tolerable level of white noise interference. As the distance the third device 1003 moves away from the first device 104a increases, these further white noise interference levels may increase beyond the previous white noise interference levels.

[0202] The first device 104a and the second device 104b perform operation 1092 using an adjusted white noise interference level. Figure 10 The remaining operations 1026, 1027a, 1027b, 1028a, 1028b, and 1029 essentially correspond to... Figure 7 Operations 726, 727a, 727b, 728a, 728b and 729.

[0203] Figure 11 It shows in Figure 7 or Figure 10 The signal flow graph of another example of scrambling DNN operation 1125 between the first device 104a and the second device 104b during operation 725 or 1092 of the scrambling DNN communication 720 or 1000 shown is shown. Figure 8 Scrambling DNN operation 825 or Figure 9 The scrambling DNN operation 925 involves the third device 1103 instructing the first device 104a to adjust the white noise interference level to an acceptable level. The signal flow of the scrambling DNN operation 1125 from the first device 104a to the second device 104b includes the following signal flow operations:

[0204] At the first device 104a, operations 1102, 1104, 1106, 1108, 1110, 1114, 1116, and 1125a substantially correspond to Figure 8 Operations 802, 804, 808, 810, 814, 816, and 825a, or Figure 9 Operations 902, 904, 908, 910, 914, 916, and 925a. At the second device 104b, operations 1134, 1148, 1150, 1152, and 1153 substantially correspond to the reference. Figure 8 The operations described are 834, 848, 850, 852, and 853, or refer to [reference]. Figure 9 Operations 934, 948, 950, 952, and 953 are described. After the first device 104a and the second device 104b perform scrambling DNN operation 1125, the first device 104a and the second device 104b perform operations 1102, 1104, 1106, 1108, 1110, 1114, and 1116, and operations 1134, 1148, 1150, 1152, and 1153, respectively, for one or more time slots. During these one or more time slots, regardless of the scrambling DNN communication between the first device 104a and the second device 104b, the third device 1103 experiences an intolerable level of white noise interference in which the transmit DNN structure of the first device 104a generates an output communication signal that satisfies the current level of white noise interference when transmitted. This means that the current white noise interference level set by the first device 104a is too high, and the transmission from the first device still results in an intolerable or unacceptable level of interference at the third device 1103.

[0205] like Figure 10 As described in operation 1005a, in operation 1105a, the third device 1103 sends a notification to the first device 104a (or the second device 104b) indicating that the transmission of output communication signals from the first device 104a to the second device 104b exceeds an acceptable level of white noise interference associated with the third device 1103. Figure 10As described in operation 1005b, in operation 1105b, when the first device 104a receives a notification from the third device 1103 in operation 1105a, it adjusts the white noise interference level by reducing it to an acceptable amount (if indicated by the third device 1103) or increasing it. If the latter, the third device 1103 sends a further notification indicating that the adjusted white noise interference level is still intolerable or unacceptable, wherein the first device 104a adjusts the further increase until the third device 1103 stops sending any further notification or sends a notification indicating that the white noise interference level is acceptable. Further examples as described with respect to operations 1005a and 1005b also apply. Figure 11 Operations 1105a and 1105b.

[0206] In operation 1125a, when the first device 104a receives an acknowledgment from the second device 104b in operation 1152, it continues the next transmission of further input communication data (if any) from the data source in one or more subsequent time slots, wherein the scrambled DNN operation 1125 signal stream is repeated until it is determined that scrambled DNN communication is disabled, as... Figure 7 Operation 726 or Figure 10 As described in operation 1026.

[0207] Figure 12 The following diagram illustrates an example of the signal flow for scrambling DL / UL DNN communication 1220 during a DL / UL DNN communication session between BS 210 and UE 220. In this example, Figure 2a The reference numerals are used for the same or similar parts. Scrambled DNN communications 720 and 1020 are further modified to include scrambled DL / UL DNN communications 1220 between the BS and the UE during a DL / UL DNN communication session between the BS and the UE. The BS 210 and UE 220 in Figure 2 use as referenced. Figures 1 to 11 Any aspect described herein shall be used to perform scrambling DL / UL DNN communication 1220. Specifically, operations 1221a, 1221b, 1292, 1226, 1227a, 1227b, 1228a, 1228b, and 1229 at BS 210 and UE 220 are further modified as described in references. Figure 7 The corresponding operations described are 721a, 721b, 792, 726, 727a, 727b, 728a, 728b and 729, and / or Figure 10The corresponding operations 1021a, 1021b, 1092, 1026, 1027a, 1027b, 1028a, 1028b, and 1029 are used for the DL / UL DNN communication session between BS210 and UE 220. The signal flow of the scrambled DL / UL DNN communication 1220 for the communication session between BS210 and UE 220 includes the following signal flow operations:

[0208] In Operation 1221a, the BS and UE perform Radio Resource Control (RRC) DNN connection establishment to establish a DL / UL DNN communication session between them. The DL / UL DNN communication session includes DL DNN communication from BS 210 to UE 220 and UL DNN communication from UE 220 to BS 210. During the establishment of the DL / UL DNN communication session, BS 210 and UE 220 communicate with each other to define, agree on, and / or configure the types of DL / UL transmit DNN structures and DL / UL receive DNN structures, each of which will be used in the DL DNN and UL DNN communications to perform end-to-end communication between the BS and the UE.

[0209] For example, when UE 220 is in the RRC_IDLE state, UE 220 sends an RRC DNN connection request to BS 210 to establish a DL / UL DNN communication session. The RRC DNN connection request may include the UE identifier and the UE's capabilities, enabling BS 210 to select appropriate DL / UL transmit DNN structures and DL / UL receive DNN structures for DL ​​and UL communication channels (e.g., Physical Downlink Shared Channel (PDSCH) and Physical Uplink Shared Channel (PUSCH)), while also defining control channels (e.g., Physical Downlink Control Channel (PDCCH) or Physical Uplink Control Channel (PUCCH)) for DL ​​and / or UL DNN communication between UE 220 and BS 210. BS 210 has a set of DL transmit / receive DNN structures and / or pairs thereof, and a set of UL transmit / receive DNN structures and / or pairs thereof, each of which is mapped to a DL / UL DNN identifier and stored in a DNN lookup table at BS 210. UE 220 has a corresponding set of UL / DL transmit / receive DNN structures that are also mapped to the same UL / DL DNN identifier and stored in a DNN lookup table at UE 220. BS 210 selects the DL transmit and receive DNN structures and / or UL transmit and receive DNN structures based on the DL and / or UL communication channels / environment, the communication performance requirements for the DL and / or UL DNN connection, and the type of DL and / or UL data communication (e.g., voice communication, data communication, multimedia streaming, and the like). BS 210 sends an RRC DNN connection establishment message, which includes data representing the selected DL / UL receive DNN structure (e.g., DL / UL DNN identifier) ​​for the DL and UL communication channels (e.g., PDSCH and PUSCH) and the DL and UL control channels (e.g., PDCCH / PUCCH) for DL ​​and UL DNN communication between UE 220 and BS 210. When UE 220 receives an RRC DNN connection establishment message, it can use the DL DNN identifier and UL DNN identifier from the DNN lookup table to select and configure the corresponding DL receive DNN structure and UL transmit DNN structure. UE 220 can send an RRC connection establishment complete message to BS 210, indicating that UE 220 has configured the corresponding DL receive DNN structure and UL transmit DNN structure accordingly and is ready to communicate with BS 210 via DL and UL DNN. BS 210 also configures the corresponding DL transmit DNN structure and UL receive DNN structure for DL ​​and UL DNN communication with UE 220.

[0210] Alternatively, when UE 220 is in the RRC_CONNECTED state, BS 220 sends an RRC DNN connection reconfiguration message to UE 220 to modify an existing RRC connection into an RRC DNN connection. This RRC DNN connection reconfiguration message includes data indicating the appropriate DL / UL transport DNN structure and DL / UL receive DNN structure for the DL and UL communication channels (e.g., PDSCH and PUSCH), and also defines the control channels (e.g., PDCCH / PUCCH) for DL ​​and UL communication between UE 220 and BS 210. UE 220 may send an RRC reconfiguration complete message to BS 210, indicating that UE 220 has configured the corresponding DL receive DNN structure and UL transport DNN structure accordingly and is ready to perform DL and UL DNN communication with BS 210. BS 210 also configures the corresponding DL transport DNN structure and UL receive DNN structure for DL ​​and UL DNN communication with UE 220.

[0211] In operation 1221a, during or after the establishment of the DL / UL DNN connection, BS 210 sets the white noise interference level of the DL communication channel and / or UL communication channel to a default setting or a certain white noise interference level, depending on the performance requirements of the DL / UL DNN communication session. Alternatively or additionally, in the example, UE 220 requests a certain white noise interference level to improve the block error rate performance of DL and / or UL DNN communication (e.g., requesting an increase in the white noise interference level).

[0212] In Operation 1221b, after BS 210 and UE 220 establish a DL / UL DNN communication session, BS 210 and UE 220 respectively use the configured DL transmit and DL receive DNN structures to perform DL DNN communication for one or more time slots. UE 220 and BS 210 also respectively use the configured UL transmit and UL receive DNN structures to perform UL DNN communication for one or more time slots.

[0213] In operation 1292, when BS 210 detects that a transmission from BS 210 does not meet a specific white noise interference level for DL ​​DNN communication, or a transmission from UE 220 does not meet a specific white noise interference level for UL DNN communication, BS 210 and UE 220 perform DL and / or UL DNN scrambling. For DL ​​DNN scrambling communication from BS 210 to UE 220, the TXDNN controller of BS 210 and the RX DNN controller of UE 220 control the scrambling DL DNN operation, as referenced... Figures 1 to 11 (Specifically, Figures 4a to 5 As described in the description. Similarly, for UL DNN scrambling communication from UE 220 to BS 210, the TX DNN controller of UE 220 and the RX DNN controller of BS 210 control the scrambling UL DNN operation.

[0214] As described above, the white noise interference level is determined by BS 210 based on UE 220's request for improved block error rate performance, or by another UE or BS experiencing intolerable white noise interference caused by transmissions from BS 210 during DL DNN communication or from UE 220 during UL DNN communication (e.g., see the reference to third device 1003 or 1103). Figure 10 or Figure 11 The request is used to set the DL / UL. BS 210 and UE 220 perform DNN operations on DL / UL based on, but not limited to, the following scrambling DL / UL DNN operations 1292:

[0215] In operation 1222a, when the output communication signal from the DL transmission DNN structure of BS 210 is predicted to not meet the white noise interference level of the DL communication channel if transmitted on the DL communication channel (e.g., PDSCH), BS 210 enables the execution of scrambling DL DNN operation for one or more time slots. For example, the DL TX DNN controller of BS 210 detects that the DL transmission DNN structure predicts the output communication signal from the input communication data to produce a transmission signal on the PDSCH that does not meet the white noise interference level. In this case, the DL scrambling DNN operation is enabled, and the DL TX DNN controller of BS 210 selects NNSI, as referenced. Figure 1 Figure 2 and Figures 3a to 6d (in particular Figures 3a to 6d As described, this is used to reconfigure the DL transmission DNN structure to generate an output communication signal from the same input communication data, which will produce a transmission signal on the PDSCH that meets the white noise interference level. Similarly, when UE 220 or BS 210 detects that the transmission signal on the UL communication channel (e.g., PUSCH) does not meet the white noise interference level set for the UL communication channel, UL scrambling DNN operation is enabled. BS 210 selects an NNSI for reconfiguring the UL transmission DNN structure for use by UE 220 to generate an output communication signal from it, which will produce a transmission signal on the PUSCH that meets the white noise interference level.

[0216] In operation 1223a, BS 210 transmits information on when and how scrambling DL and UL DNN operations will be enabled by sending, for example, RRC control messages (e.g., RRC connection reconfiguration), Media Access Control (MAC) messages, or Downlink Control Information (DCI) messages to UE 220 on the PDCCH. This message includes fields including the selected NNSI configuration and DL and / or UL timing information regarding when the DL receive DNN structure at UE 220 should be reconfigured and / or when the UL transmit DNN structure at UE 220 should be reconfigured. Initially, the selected NNSI includes initial seed information, the type of random permutation function, the specific layer of the DL and / or UL transmit DNN structure to be reconfigured, data for random permutation iterations, and any other data enabling UE 220 to reconfigure the DL receive DNN structure to generate reconstructed communication data and / or reconfigure the UL transmit DNN structure to generate output communication signals for transmission on the PUSCH. Subsequent RRC / MAC / DCI messages for further DL or UL scrambling DNN operations include updated NNSI configurations, such as the number of selected random permutation iterations, or one or more selected neural network layers that have been reconfigured for DL ​​and / or UL scrambling.

[0217] In operation 1223b, UE 220 acknowledges (e.g., ACKs) the reception of the selected NNSI configuration for scrambling / descrambling DL DNN operations (e.g., descrambling DL RX DNN operations performed by UE 220) and / or scrambling UL DNN operations (e.g., scrambling UL TX DNN operations performed by UE 220), and the enabling of scrambling DL / UL DNN operations.

[0218] In operations 1224a and 1224b, BS 210 and UE 220 initiate scrambling DL and / or UL DNN operations based on DL and / or UL timing information. For example, BS prepares to perform a scrambling DL TX DNN operation for DL ​​transmission to UE 220, while UE 220 prepares to perform a descrambling DL RX DNN operation for receiving DL transmission from BS 210 using DL timing information (e.g., one or more DL time slots). In another example, UE 220 prepares to perform a scrambling UL TX DNN operation for transmitting UL transmission from UE 220 to BS 210, while BS 210 prepares to perform a descrambling UL RX DNN operation for transmitting UL transmission from UE 220 using UL timing information (e.g., one or more UL time slots).

[0219] In operation 1225, BS 210 and UE 220 are in accordance with reference Figures 1 to 11(In particular, Figure 7 and Figure 10 The scrambling DL / UL DNN operation is performed in a similar manner to that described in the PDSCH and / or PUSCH communication with each other.

[0220] In Operation 1226, BS 210 disables scrambling DL / UL DNN operations after scrambling DL / UL DNN operations / communications are determined to be unnecessary.

[0221] In Operation 1227a, when BS 210 disables or turns off scrambling DL and / or UL DNN operations, BS 210 sends to UE 220, for example, an RRC message, a Media Access Control (MAC) message, or a DCI message, which notifies UE 220 that scrambling DL and / or UL DNN operations are being disabled / turned off.

[0222] In operation 1227b, UE 220 acknowledges (e.g., ACK) the receipt of the RRC / MAC / DCI message transmitted by BS 210 in operation 1227a.

[0223] In operation 1228a, when an acknowledgment (ACK) is received from UE 220 in operation 1227b, BS 210 disables scrambling DL / UL DNN operations and restores the DL transmit DNN structure and / or UL receive DNN structure to their original configuration. Alternatively, BS 210 retains the current DL transmit DNN structure and / or UL receive DNN structure (e.g., the most recent reconfiguration), which still allows any further DNN communication to potentially meet the white noise interference level.

[0224] In operation 1228b, when an RRC / DCI message for disabling scrambling DL / UL DNN operation is received in operation 1227a, UE 220 disables scrambling DL and / or UL DNN operation and restores the receive DL DNN structure and / or UL transmit DNN structure to their original configuration. Alternatively, if BS 210 has also instructed it to do so, UE 220 retains the current DL receive DNN structure and / or the current UL transmit DNN structure (e.g., the latest reconfiguration), which still allows any further DNN communication to potentially meet the white noise interference level.

[0225] In Operation 1229, after BS 210 and UE 220 disable the scrambling DNN connection, BS 210 and UE 220 continue to perform DNN communication. When DNN communication is no longer needed, BS 210 and UE 220 resume performing, for example, standard or regular 3G to 4G, 5G, or 6G communication sessions between them.

[0226] Figure 13It shows in Figure 12 An example signal flow of scrambled DL / UL DNN operation 1325 between BS 210 and UE 220 during operation 1225 in scrambled DL / UL DNN communication 1220. In this example, Figure 2a The reference numerals in the accompanying drawings are used for the same or similar parts. Example scrambling DNN operation 825 is further modified to include scrambling DL DNN operations between the BS and the UE during a DL DNN communication session between the BS 210 and the UE 220. Figure 2a The BS 210 and UE 220 are used as referenced. Figures 1 to 12 Any aspect described can be used to perform scrambling DL DNN operation 1325. Specifically, at BS 210, operations 1302-1316 substantially correspond to, as referenced... Figure 8 Operations 802-816 are described, but modified for scrambling DL DNN communication. Similarly, at the UE, operations 1334, 1348, 1350, 1352, and 1353 essentially correspond to those in the reference. Figure 8 Operations 834, 848, 850, 852, and 853 are described, but modified for scrambling DL DNN communication. In this example, Figure 12 DNN operation 1292 configures BS 210 and UE 220 to perform scrambling DL DNN operation 1325. The signal flow of scrambling DL DNN operation 1325 from BS 210 to UE 220 includes the following signal flow operations:

[0227] Operations 1302 to 1306 basically correspond to Figure 8 Operations 802 to 804, in addition to the BS 210's DL transmission DNN structure processing input communication data to generate output communication signals for transmission to the UE 220 on the PDSCH in a specific time slot.

[0228] Operation 1306 further modifies, for example, by using RRC / DCI messaging on the PDCCH to transmit any updated NNSI and specific timeslot information to the UE 220 via BS 210. Figure 8 Operation 806. Upon receiving an RRC / DCI message from BS 210, UE 220 communicates with... Figure 8 Operation 1334 is performed in a similar manner to operation 834, wherein the NNSI buffer or NNSI table accessible to the UE 220 stores updated NNSI and specific timeslot information for descrambling received transmissions on the PDSCH in relation to the specific timeslot information.

[0229] In operation 1316, when the transmission of the generated output communication signal meets the white noise interference level, BS 210 transmits the reconfigured DL transmission DNN structure's output communication signal on the PDSCH according to specific time slot information. When the transmission of the output communication signal on the PDSCH is received in a specific time slot, UE 220 can buffer the received output communication signal until it is ready for processing by the DL receive DNN structure.

[0230] Operations 1348 and 1350 essentially correspond to operations 848 and 850, except that the DL receive DNN structure is reconfigured using an updated NNSI for a specific time slot, and the DL receive DNN structure processes the received output communication signal for the specific time slot and generates reconstructed communication data corresponding to the input communication data transmitted in the specific time slot. UE 220 sends the reconstructed communication data to its data sink, or sends the reconstructed communication data to one or more upper protocol layers of UE 220's protocol stack.

[0231] Operations 1352 and 1325a basically correspond to Figure 8 Operations 852 and 825a. Repeat scrambling DL DNN operation 1325 until it is determined that scrambling DL / UL DNN communication should be disabled, as... Figure 12 As described in operation 1226.

[0232] Figure 14 The signal flow of an example of scrambled UL DNN communication 1420 between UE 220 and BS 210 is shown. In this example, Figure 2a The reference numerals are used for the same or similar parts. In this example, operations 1421a / b, 1422a, 1423a, 1423b, 1424a and 1424b correspond to operations 1221a / b, 1222a, 1223a, 1223b, 1224a and 1224b, except that UE 220 uses a UL transmit DNN structure (e.g., UL TX DNN) to perform scrambled UL DNN communication on the PUSCH, and BS 210 uses a UL receive DNN structure (e.g., UL RX DNN) to perform scrambled UL DNN communication on the PUSCH.

[0233] When UE 220 has input communication data for transmission, UE 220 uses the UL transmission DNN structure to generate UL output communication signals for transmission on the PUSCH. Operation 1406 essentially corresponds to Figure 8 or Figure 13Operations 806 or 1306, in addition to UE 220 performing these operations, involve UE 220 alternatively detecting whether the predicted transmission of the UL output communication signal does not meet the white noise interference level of the PUSCH. When a non-compliance with the white noise interference level is detected, UE 220 performs operations 1408, 1410, and 1414 based on the following:

[0234] Operation 1408 essentially corresponds to operation 808 or 1308, except that UE 220 selects an updated NNSI for reconfiguring the UL transmission DNN structure to generate a UL output communication signal that meets the white noise interference level in a specific time slot when transmitted on the PUSCH. This can be referenced as follows. Figures 1 to 6d and / or Figure 8 or Figure 13 The selection is iterative, as described in operations 808 or 1308. After UE 220 selects the updated NNSI, UE 220 proceeds to operation 1410.

[0235] Operations 1410 and 1414 essentially correspond to operations 810 and 814 or 1310 or 1314, except that UE 220 reconfigures the UL transmission DNN structure using an updated NNSI for a specific time slot, and UE 220 transmits the updated NNSI to BS 210 using uplink control signaling on the PUCCH. BS 210 already has the specific time slot information, which is determined by BS 210 during RRC / DCI message passing to UE 220. Upon receiving uplink control signaling, BS 210 performs operation 1434, which essentially corresponds to operation 834 or 1334, except that BS 210 stores the updated NNSI with the corresponding specific time slot in an NNSI buffer or NNSI table at BS 210 for descrambling received transmissions on the PUSCH using the UL receive DNN structure associated with the specific time slot information.

[0236] In operation 1416, when the transmission of the generated UL output communication signal meets the white noise interference level, UE220 transmits the reconfigured UL transmission DNN structure UL output communication signal on PUSCH according to a specific time slot.

[0237] Operations 1448 and 1450 basically correspond to Figure 8 Operations 848 and 850 or Figure 13In addition to BS210 performing these operations when receiving UL output communication signals transmitted on the PUSCH in a specific time slot, BS210 buffers these operations before processing the received UL output communication signals for a specific time slot using the UL receive DNN structure reconfigured with the updated NNSI for that specific time slot. The UL receive DNN structure processes the received UL output communication signals for that specific time slot and generates reconstructed communication data corresponding to the input communication data transmitted by UE 220 on the PUSCH in that specific time slot. BS210 sends the reconstructed communication data to its data sink or to one or more upper protocol layers of its protocol stack.

[0238] Operations 1452 and 1453 basically correspond to Figure 8 Operations 852 and 853 or Figure 13 Operations 1352 or 1353, except those performed by BS 210.

[0239] In operation 1420a, when UE 220 receives the acknowledgment in operation 1452, it continues to the next transmission of further input communication data (if any) from the data source at UE 220 on the PUSCH in one or more subsequent time slots, wherein scrambled UL DNN communication 1420 is repeated until it is determined that scrambled UL DNN communication should be disabled, as... Figure 12 As described in operation 1226.

[0240] Figure 15 The signal flow of another example of scrambled UL DNN communication 1520 between UE 220 and BS 210 is shown. In this example, Figure 2a The reference numerals in the accompanying drawings are used for the same or similar parts. In this example, operations 1521a / b, 1522a, 1523a, 1523b, 1524a, and 1524b correspond to operations 1421a / b, 1422a, 1423a, 1423b, 1424a, and 1424b. When UE 220 has input communication data for transmission, the UE uses the UL transmission DNN structure to generate UL output communication signals for transmission on the PUSCH. Operation 1506 is further modified based on the following modifications. Figure 14 Operation 1406:

[0241] In operation 1507, after detecting that the UL output communication signal would not meet the white interference noise level if transmitted, UE 220 uses uplink control signaling on PUCCH to transmit a request for the updated NNSI from BS 210.

[0242] Operation 1508 basically corresponds to Figure 8 or Figure 13 Operations 808 or 1308, except that BS 210 selects an updated NNSI for reconfiguring the UL transmission DNN structure of UE 220 to generate a UL output communication signal that satisfies the white noise interference level in a specific time slot when transmitted on the PUSCH. This can be done by BS 210 as referenced. Figures 1 to 6d and / or Figure 8 , Figure 13 and Figure 14 The iterative selection is described in operations 808, 1308, and 1408. In the example, see reference... Figure 2a As described, when BS 210 selects an NNSI for UE 220, BS 210 uses random UE UL input communication data and UE UL transmission DNN structure to simulate UL to generate output communication signal, and selects an NNSI that generates an output communication signal that satisfies the white noise interference level of that UL.

[0243] Operation 1514 essentially corresponds to operations 814 and 1314, except that BS 210, for example, uses RRC / DCI messaging on the PDCCH to transmit updated NNSI and specific timeslot information for UL to UE 220. In operation 1534a, BS 210 stores the updated NNSI and specific timeslot information in an NNSI buffer or NNSI table at BS 210 for descrambling received transmissions on the PUSCH in relation to the specific timeslot information. Similarly, upon receiving an RRC / DCI messaging from BS 210, UE 220 performs operation 1534b and stores the updated NNSI and specific timeslot information in an NNSI buffer or NNSI table at UE 220 for scrambling incoming communication data according to the specific timeslot information for transmission on the PUSCH. In operation 1510, UE 220 reconfigures its UL transmission DNN structure using the updated NNSI for the specific timeslot.

[0244] Operations 15 and 16 basically correspond to respectively Figure 8 or Figure 13 Operations 816 and 1316, in addition to UE 220 transmitting UL output communication signals on PUSCH according to specific time slots.

[0245] Operations 1516, 1548, 1550, 1552, 1554, and 1520a correspond to Figure 14Operations 1416, 1448, 1450, 1452, 1454, and 1420a. UE 220 continues to the next transmission on the PUSCH of further input communication data (if any) from the data source at UE 220 in one or more subsequent time slots, where scrambled UL DNN communication 1520 is repeated until it is determined that scrambled UL DNN communication should be disabled, as... Figure 12 As described in operation 1226.

[0246] Figure 16 As shown in the reference Figure 2a and Figures 12 to 15 The described signal flow of the scrambled DL / ULDNN communication session 1600 between BS 210 and UE 220. In this example, Figure 2a The reference numerals in the accompanying drawings are used for the same or similar components. BS 210 includes a BS DNN controller 214 (BS DNNC), a BS DL transmit DNN structure (BS DL TX DNN) for processing and scrambling input communication data for downlink transmission to UE 220 on PDSCH, and a BS UL receive DNN structure (BS UL RX DNN) for processing uplink transmissions received from UE 220 on PUSCH. UE 220 includes a UE DNN controller 224 (UE DNNC), a UE DL receive DNN structure (UE DLRX DNN) for processing downlink transmissions received from BS 210 on PDSCH, and a UE UL transmit DNN structure (UE UL TXDNN) for uplink transmissions to BS 210 on PUSCH. It is also assumed that BS 210 has assigned appropriate frequencies / time slots to UE 220 for control plane signaling on the corresponding PDCCH and PUCCH.

[0247] In operation 1621a, BS DNNC 214 establishes a UL / DL DNN communication session between BS 210 and UE 220. BS DNNC 214 selects a DL transmit and receive DNN structure (BS DL TX DNN and UE DL RX DNN) pair for scrambling the DL / UL DNN communication session 1600. In this scrambling DL / UL DNN communication session, BS 210 transmits an RRC establishment request message, which includes the DL DNN type or identifier associated with the selected DL TX / RX DNN pair (e.g., RRCDNN establishment request (DL DNN type / Id)). BS DNNC 214 retrieves the TX DNN configuration data of BS DL TX DNN 228 corresponding to the selected DL DNN type or identifier from the BS TX / RX DNN store or table 215b. In operation 1621b, BSDNNC 214 sends a configuration instruction (e.g., Cfg(DL DNN type)) to configure BS DL TX DNN structure 206. In operation 1621c, when UE DNNC 224 receives an RRC establishment request message including a DL DNN type or identifier associated with the selected DL TX / RX DNN pair, UE DNNC 224 retrieves RX DNN configuration data for UE DL RX DNN 228 corresponding to the received DL DNN type or identifier from UE TX / RX DNN storage / table 225b at UE 220, and sends a configuration instruction (e.g., Cfg(DL DNN type)) to UE DL RX DNN structure 228 to configure UE DL RX DNN 228 based on the retrieved RX DNN configuration data. In operation 1621d, after the configuration of UE DL RX DNN 228, UE DNNC 224 of UE220 sends an RRC response (e.g., RRC DNN establishment response (ACK)) to BS 210 indicating confirmation of the configuration of UE DL RX DNN 228.

[0248] Upon receiving confirmation, in operation 1602i, the input communication data (e.g., I_Data Xi) for transmission in a specific time slot TSi is applied to the BS DL TX DNN 206, which generates a DL output communication (OC) signal (e.g., OC_Data Xi) corresponding to I_Data Xi. In operation 1604i, DL OC_Data Xi is provided to the BSDNNC 214 for transmission as a transmission waveform signal to the UE 220 on the PDSCH in TSi. (See reference...) Figure 5As described, upon receiving OC_Data Xi, BS DNNC 214 detects whether the transmission of OC_Data Xi will meet the white interference noise level. If so, OC_Data Xi is buffered and transmitted to UE 220 as a transmission waveform signal on PDSCH in a specific TSi. Otherwise, NNSI is generated / selected to reconfigure DL TX DNN 206 to generate OC_Data Xi that meets the white interference noise level, as described in the reference. Figure 5 and / or Figures 1 to 4d and / or Figures 6a to 15 As described in either of the above. In operation 1616i, BS DNNC 214 transmits OC_Data Xi as a transmitted waveform signal in TSi to UE 220 in PDSCH (e.g., PDSCH RF TX WAVEFORM (OC_DataXi, TSi)). UE 220 receives the transmitted signal waveform in TSi on PDSCH and processes (e.g., downconverts) it to the received OC_Data Xi (e.g., RxOC Xi), and in operation 1650i-1, the Rx OC Xi of TSi is input to UE DL RX DNN 228 for processing. In operation 1650i-2, UE DL RX DNN 228 processes the Rx OC Xi and generates reconstructed communication data (e.g., R_IData Xi) for TSi corresponding to I_Data Xi. In operation 1652i, after UE DL RX DNN 228 has successfully generated R_IData Xi for TSi, UE DNNC 224 uses uplink control plane signaling to send an acknowledgment (e.g., ACK) to BS 210. Upon receiving the ACK from UE 220, operations 1602i to 1652i are repeated for further input communication data to be transmitted from BS 210 to UE 220 in subsequent time slots.

[0249] In operation 1622, if the BS DNNC 214 detects that the DL output communication signal will not meet the white interference noise level during transmission (as per reference)... Figures 1 to 15 If (as described), then scrambled DL DNN communication is enabled, and BS 210 selects NNSI for reconfiguring BS DL TX DNN 206 to generate a DL output communication signal that meets the white interference noise level (as described in the reference). Figures 1 to 15(As described). The selected NNSI includes the specific neural network layer to be scrambled (e.g., NN layer #), the initial seed (e.g., Seed), the identifier of the permutation random function (PRF), and the number of random permutation iterations (e.g., RPermIt #) (e.g., NN layer #, Seed, PRF ID, RPermIt #). In operation 1622a / 1623a, BS DNNC 214 enables scrambled DLDNN communication and transmits the selected NNSI (e.g., RRC scrambled DNN enable request (NN layer #, Seed, PRF ID, RPermIt #)) to UE 220 on the PDCCH using RRC control plane signaling. In operation 1624b, when UE DNNC 224 receives an RRC scrambled DNN enable request message including the selected NNSI associated with BS DL TX DNN 206, UE DNNC 224, based on the NNSI and as referenced Figures 3a to 3c and / or Figures 4a to 4b The described method generates configuration data for reconfiguring the UE DLRX DNN 228 and sends configuration instructions (e.g., Cfg (scrambling)) to scramble the configuration information of the UE DLRX DNN 228 for reconfiguring the UE 220.

[0250] In operation 1623b, after UE DL RX DNN 228 has been reconfigured, UE DNNC 224 of UE 220 uses uplink control plane signaling to send an RRC response message (e.g., RRC scrambling DNN enable response (ACK)) on the PUSCH to BS 210 indicating acknowledgment of the reconfiguration. Upon receiving the acknowledgment, in operation 1624c, BS DNNC 214 sends a configuration instruction (e.g., Cfg(Scramble)) with scrambling configuration information based on the selected NNSI to reconfigure BS DL TX DNN structure 206.

[0251] In operation 1602j, input communication data (e.g., I_Data Yj) for time slot TSj is applied to the reconfigured BS DL TX DNN 206, which generates a DL output communication signal (e.g., OC_Data Y) corresponding to I_Data Yj in TSj. In operation 1604j, BS DL TX DNN 206 outputs OC_Data Yj to BS DNNC 214 for transmission as a transmission waveform signal in TSj to UE 220 on the PDSCH. (See reference...) Figure 5As described, upon receiving OC_Data Yj, the BS DNNC 214 detects whether OC_Data Yj will meet the white interference noise level during transmission. If so, OC_Data Yj is buffered and transmitted as a transmission waveform signal in TSj on the PDSCH. In operation 1616i, the BS DNNC 214 transmits OC_Data Yj as a transmission waveform signal in TSj on the PDSCH (e.g., PDSCH RF TX WAVEFORM (OC_DataYj, TSj)). The UE 220 receives the transmission signal waveform for TSj on the PDSCH and processes (e.g., down-converts to baseband) it into the received OC_Data Yj (e.g., Rx OC Yj). In operation 1650j-1, the UE DNNC 224 applies or inputs the Rx OC Yj for TSj to the reconfigured UE RX DL DNN 228 for DNN processing. In operation 1650j-2, UE RX DL DNN 228 processes Rx OC Yj and generates reconstructed communication data (e.g., R_IData Yj) for TSj corresponding to I_Data Yj. In operation 1652j, after UE DL RX DNN 228 has successfully generated R_IData Yj, UE DNNC 224 sends an acknowledgment (e.g., ACK) to BS 210 on the PUCCH using uplink control plane signaling. Upon receiving the ACK from UE 220, operations 1602j to 1652j are repeated for further input communication data to be transmitted from BS 210 to UE 220 in subsequent time slots until it is detected that the transmission of the DL output communication signal for subsequent time slots generated by BS DL TX DNN 206 will not meet the white noise interference level.

[0252] In operation 1606a, BS DNNC 214 detects that the current DL output communication signal of the current TSa will not meet the white interference noise level when transmitted on the PDSCH, as referenced. Figures 1 to 15 As described. BS 210 selects the updated NNSI for reconfiguring BS TX DL DNN 206 to generate DL output communication signals for TSa and any subsequent time slots, which are predicted to meet white interference noise levels, as referenced. Figures 1 to 15As described. The selected updated NNSI includes a random permutation iteration number (e.g., RPermIt #) associated with a sequence of random permutations used to generate the ordering of the current specific neural network layer defined in operation 1623a. In operation 1614a, BS DNNC 214 transmits RRC control plane signaling (e.g., RRC / MAC control message (RPermIt #, TS {a, b, c, d})) on PDCCH to UE 220 with the selected updated NNSI and corresponding specific timing information (e.g., TSa, TSb, TSc, and TSd, etc.). The specific timing information (e.g., TS {a, b, c, d}) describes when UE DNNC 224 should apply the updated NNSI to UE DL RX DNN 228. In operation 1614b, when UEDNNC 224 receives an RRC / MAC control message including the selected updated NNSI and specific timing information, UEDNNC 224 stores the updated NNSI and the corresponding specific timing information (e.g., TS {a, b, c, d}) in the UE NNSI storage device / lookup table / buffer 225a. In operation 1614b, UE DNNC 224 uses uplink control plane signaling to send an acknowledgment (e.g., ACK) to BS 210 on the PUCCH to confirm receipt of the updated NNSI and specific timing information. Upon receiving the acknowledgment in operation 1614b, in operation 1610a, BS DNNC 214 sends a configuration command (e.g., Cfg(RPermIt#)) with scrambling configuration information associated with the selected random permutation iteration number of the selected updated NNSI.

[0253] In operation 1602a, BS 210 inputs or applies input communication data (e.g., I_Data Za) for time slot a to a reconfigured BS DL TX DNN 206, which generates a DL output communication signal (e.g., OC_Data Za) for TS a corresponding to I_Data Za. In operation 1604a, OC_Data Za is provided to BSDNNC 214 to be transmitted to UE 220 as a transmission waveform signal on the PDSCH in a specific time slot TSA. (See reference...) Figure 5As described, upon receiving OC_Data Za, the BS DNNC 214 detects whether the transmission of OC_Data Za meets the white interference noise level. If so, the OC_Data Za is buffered and transmitted as a transmission waveform signal in a specific time slot TSa. If the transmission of OC_Data Za does not meet the white interference noise level, operation 1606a is performed again.

[0254] In operation 1616a, BS DNNC 214 transmits OC_Data Za as a transmit waveform signal in TSa on the PDSCH (e.g., PDSCH RF TX WAVEFORM (OC_DataZa, TSa)). UE 220 receives the transmit signal waveform for TSa on the PDSCH and down-converts it to the received OC_Data Za (e.g., Rx OC Za)). See reference... Figure 5 As described, UE DNNC 224 can buffer Rx OC Za for TSa until UE DL RX DNN 228 is ready to process TSa. Before processing TSa, in operation 1648a, UE DNNC 224 retrieves the NNSI associated with TSa and generates configuration data for reconfiguring UE DL RX DNN 228 based on the retrieved NNSI, as described in reference [reference missing]. Figures 3a to 3c and / or Figures 4c to 4dAs described, and sends a configuration instruction (e.g., Cfg(RPermit#)) with a scrambling configuration for reconfiguring UE DL RX DNN 228. In operation 1650a-1, UE DNNC 224 applies Rx OC Za for TSa to the reconfigured UE DL RX DNN 228 for processing. In operation 1650a-2, UE DL RX DNN 228 processes Rx OCZa and generates reconstructed communication data for TSa corresponding to I_Data Za (e.g., R_IData Za). In operation 1652a, after UE DL RX DNN 228 has successfully generated R_IData Za, UE DNNC 224 sends an acknowledgment (e.g., ACK) to BS 210 on the PUCCH using uplink control plane signaling. Upon receiving an ACK from UE 220, when BS 210 transmits subsequent input communication data (e.g., I_Data Zb, I_Data Zc, I_Data Zd, etc.) to UE 220, operations 1602a to 1652a are repeated for at least TSb, TSC, and TSd, and other subsequent time slots. As shown, operations 1602a-1652a are repeated for TSd, wherein BS 210 transmits input communication data for TSd (e.g., I_Data Zd), as described in operations 1602d to 1652d.

[0255] During DL DNN communication between BS 210 and UE 220, UE 220 and BS 210 are also performing UL DNN communication, where UE 220 uses UL transmit DNN structure 226 (e.g., UE UL TX DNN) to transmit to BS 210 on the PUSCH, and the BS uses UL receive DNN structure 208 (e.g., BS UL RX DNN) to receive and process the transmission. Assume that UE UL TX DNN 226 and BS UL RX DNN 208 are as described in the reference. Figures 12 to 15 The configuration described is as follows. In Operation 1606b, the UE DNNC224 detects that the current UL output communication signal will not meet the white noise level during transmission, as referenced. Figures 1 to 15As described. UE 220 reconfigures its UE UL TX DNN 226. In this example, UE 220 does not have the ability to select the updated NNSI, whereby instead, in operation 1607, UE DNNC 224 uses uplink control plane signaling on the PUCCH to request communication resources (e.g., UL timeslots / frequency) for uplink scrambled transmissions including the updated NNSI (e.g., PUCCH request (UL NNSI for scrambled transmissions)). Upon receiving the request for communication resources and the updated NNSI, BS 210 selects the updated NNSI for reconfiguring UE TX UL DNN 226 to generate a UL output communication signal that will satisfy the white interference noise level of the PUSCH, as referenced. Figures 1 to 15 As described (for example, see Figure 2a and Figure 15 (Operation 1508 in the above). The selected updated NNSI includes a random permutation iteration number (e.g., RPermIt#) associated with a random permutation sequence that generates the ordering of the current specific neural network layer for permuting the UE UL TX DNN 226. In operation 1614e, the BS DNNC 214 transmits the selected updated NNSI and the corresponding specific timing information (e.g., TSe) for transmission (e.g., DCI control message (RPermlt #, TSe)) to the UE 220 on the PDCCH using DCI control plane signaling, where TSe indicates the time slot from which the updated NNSI should be applied. When the UE DNNC 224 receives the DCI message including the selected updated NNSI and the specific timing information TSe, the UE DNNC 224 stores the updated NNSI and the corresponding specific timing information TSe in the UE NNSI storage device 225b. In Operation 1614f, UE DNNC 224 of UE 220 uses uplink control plane signaling to send an acknowledgment (e.g., ACK) to BS 210 on PUCCH for receiving the updated NNSI and specific timing information.

[0256] In operation 1648b, for UL transmission in TSe, UE DNNC 224 sends a configuration instruction (e.g., Cfg(RPermIt#)) with scrambling configuration information based on the number of random permutation iterations of the selected updated NNSI for TSe to reconfigure UE UL TX DNN 226. In operation 1602e, the reconfigured UE UL TX DNN 226 processes input communication data (e.g., I_DataZe) for TSe, and the UE UL TX DNN outputs a corresponding UL communication signal (e.g., OC_DataZe) for TSe. In operation 1604e, UE DNNC 224 of UE 220 processes OC_DataZe for transmission as a transmission waveform signal on PUSCH to BS 210 in TSe. (See reference...) Figure 5 As described, upon receiving OC_Data Ze, UE DNNC 224 detects whether the transmission of OC_Data Ze meets the white interference noise level. If so, OC_Data Ze is buffered and transmitted as a transmission waveform signal on PUSCH to BS 210 via TSe. If the transmission of OC_Data Ze does not meet the white interference noise level, operation 1606b is repeated.

[0257] In operation 1616e, UE DNNC 224 transmits OC_DataZe as a transmit waveform signal in TSe on the PUSCH to BS 210 (e.g., PUSCH RF TX WAVEFORM (OC_DataZe, TSe)). BS 210 receives the transmit signal waveform for TSe on the PUSCH and processes (e.g., down-converts to baseband) it into the received OC_DataZe (e.g., Rx OCZe). See reference... Figure 5 As described, BS DNNC 214 can buffer Rx OC Ze for TSe until BS UL RX DNN 208 is ready to process TSe. Before processing TSe, in operation 1648e, BS DNNC 214 retrieves the NNSI associated with TSe and generates configuration data for reconfiguring BS UL RX DNN 208 based on the retrieved NNSI, as described in reference [reference missing]. Figures 3a to 3c and / or Figures 4c to 4dAs described, a configuration instruction (e.g., Cfg(RPermit#)) with a scrambling configuration for reconfiguring the BS UL RX DNN 208 is sent. In Operation 1650e-1, the BS DNNC 214 applies the Rx OC Ze for TSe to the reconfigured BS UL RX DNN 208 for DNN processing. In Operation 1650e-2, the BS DLRX DNN 208 processes the Rx OC Ze and generates reconstructed communication data for TSe corresponding to the I_Data Ze (e.g., R_IData Ze). In Operation 1652e, after the BS DL RX DNN 208 has successfully generated the R_IData Ze, the BS DNNC 214 sends an acknowledgment (e.g., ACK) to the UE 220 on the PDCCH using downlink control plane signaling. Upon receiving an ACK from BS 210, when UE 220 has further input communication data to transmit to BS 210, the signal flow repeats operation 1602e to 1652e for subsequent time slots.

[0258] In Operation 1627a, in this example, BS 210 determines that scrambling DNN communication should be terminated (e.g., the UE requests the end of the communication session, the UE switches to RRC_IDLE state or RRC_INACTIVE state, the UE / BS connection fails, or the communication session reverts to using regular communication, etc.). In this case, BS 210 uses RRC control plane signaling to indicate that scrambling DNN communication will be disabled (e.g., RRC Scrambling DNN Disable Request()). In Operation 1628a, BS DNNC 214 sends a configuration message to the corresponding BSTX / RX DNNs 206 / 208 to restore them to their original configuration for standard DNN communication, and / or release the associated DNN computing resources used for performing DNN communication for regular communication and / or terminate the communication session. In operation 1628b, UE DNNC 224 sends a configuration message to the corresponding UE TX / RX DNN 226 / 228 to restore them to their original configuration for standard DNN communication, and / or release the associated DNN computing resources used to perform DNN communication for regular communication and / or terminate the communication session.

[0259] Figure 17A non-transitory computer-readable medium 1700 according to some embodiments is illustrated. The non-transitory computer-readable medium 1700 may include a computer-readable storage medium 1702 and / or an input / output mechanism 1704 for enabling a computing system to access the computer-readable medium 1702. Although in this example, the non-transitory computer-readable medium 1700 is a USB stick, this is merely an example and not a limitation, and those skilled in the art will understand that the non-transitory computer-readable medium 1700 can be any other type of computer-readable medium or computer program product, such as, for example, a CD, DVD, USB stick, Blu-ray disc, flash drive, etc., and / or any other computer-readable medium as required by the application. The non-transitory computer-readable medium 1700 stores a computer program, computer program code, and / or instructions that, when executed by one or more processors of a device or system, cause one or more processors of that device or system to perform one or more of the methods, operations, or processes described herein, such as any signal flow, flowchart, method, and / or procedure, for example, regarding... Figures 1 to 16 The signal flow graphs, flowcharts, and schematic diagrams, along with their related features, are disclosed.

[0260] The methods or processes described herein can be implemented in digital electronic circuit systems, integrated circuit systems, specially designed ASICs (Application-Specific Integrated Circuits), computer hardware, firmware, software, and / or combinations thereof. These may include computer program products (such as software stored, for example, on a disk, optical disk, memory, or programmable logic device) comprising computer-readable instructions that, when executed by a processor, cause the processor to perform one or more of the methods and / or processes described herein.

[0261] Any system feature described herein can also be provided as a method or process feature, and vice versa. As used herein, apparatus plus functional features can be expressed alternatively according to their corresponding structures. In particular, method aspects can be applied to system aspects, and vice versa.

[0262] Furthermore, any, some, and / or all features of one aspect may be applied in any suitable combination to any, some, and / or all features of any other aspect. It should also be understood that specific combinations of the various features described and defined in any aspect of the invention may be implemented and / or provided and / or used independently.

[0263] Although several embodiments have been shown and described, those skilled in the art will understand that changes may be made to these embodiments without departing from the principles of this disclosure, the scope of which is defined in the claims and their equivalents.

Claims

1. A method performed by a first apparatus (104a) in communication with a second apparatus (104b), the method comprising: processing (404) input communication data with a transmit deep neural network, DNN (106), to generate an output communication signal (118) for transmission to the second apparatus; in response to a predicted transmission of the generated output communication signal not satisfying a white noise jamming level, performing (406) a jamming DNN operation, the jamming DNN operation comprising: selecting (408) neural network jamming information, NNSI, for reconfiguring the transmit DNN to process the input communication data to generate a jammed output communication signal that satisfies the white noise jamming level when transmitted; transmitting (414) a control message to the second apparatus, the control message including an indication of the NNSI and jamming timing information for directing the second apparatus when to reconfigure a receive DNN (109); and transmitting (416) the jammed output communication signal that satisfies the white noise jamming level to the second apparatus based on the jamming timing information.

2. The method of claim 1, wherein, the NNSI comprises data representing randomization of an ordering of a set of neural network nodes of an input to the transmit DNN, an output of the transmit DNN, and / or one or more neural network layers of the transmit DNN to be randomized, and the performing the jamming DNN operation further comprises: reconfiguring (410) the transmit DNN based on the NNSI by randomizing the ordering of the set of neural network nodes in the input, the output, or the one or more neural network layers.

3. The method of any one of claims 1 or 2, the performing the jamming DNN operation further comprising: reconfiguring the transmit DNN by performing a random permutation of a set of neural network nodes in one or more neural network layers of the transmit DNN of the first apparatus using one or more random permutation parameters, wherein the NNSI specifies data representing the one or more random permutation parameters.

4. The method of claim 3, wherein, the one or more random permutation parameters comprise one or more of: seed data, a random permutation iteration or sequence number, or an identification of a seed generation and pseudo-randomization function for use in performing the random permutation of the set of neural network nodes, and the performing the random permutation of the set of neural network nodes further comprises: generating, for each neural network layer, a random permutation sequence corresponding to the random permutation iteration or sequence number, wherein the random permutation sequence has a length equal to a number of neural network nodes in the set of neural network nodes of the each neural network layer; and randomizing the set of neural network nodes of the each neural network layer by applying the generated random permutation sequence to the set of neural network nodes.

5. The method of claim 4, wherein, The seed data further includes one or more of an identity of the first device, an identity of the second device, an identity of a cell in which the second device is located, an identity of a cell in which the first device is located, time slot information, a frame identification number, and / or any other information associated with the first device or the second device, the method further including generating a seed from the seed data for performing the random permutation of the set of neural network nodes of the one or more neural network layers.

6. The method of any preceding claim, wherein, The performing the scrambling DNN operation further includes: reconfiguring the transmission DNN based on performing a randomization operation on the neural network nodes of a selected 1th neural network layer of the transmission DNN, where 1 < l < L, and L is a number of neural network layers, where the NNSI specifies that the selected 1th neural network layer of the transmission DNN is to be randomized.

7. The method of claim 6, wherein, The selected 1th neural network layer includes a set of N neural network nodes arranged in a particular order, where N > 1, and performing the randomization operation on the selected 1th neural network layer of the transmission DNN further includes randomizing the set of N neural network nodes of the selected 1th neural network layer based on: generating an N-dimensional permutation matrix using a random permutation sequence of length N for randomizing the set of N neural network nodes of the selected 1th neural network layer; and randomizing the selected 1th neural network layer by multiplying the particular ordering of the set of N neural network nodes of the selected 1th neural network layer by the N-dimensional permutation matrix to form a randomized ordering of the set of N neural network nodes.

8. The method of any preceding claim, wherein, Selecting the NNSI further includes: iteratively selecting an NNSI based on: for an i-th iteration i > 0, generating (672) an i-th randomization operation related to the input, the output, or a set of neural network nodes of the one or more neural network layers of the transmission DNN; reconfiguring (674) the transmission DNN based on the i-th randomization operation; processing (676) the input communication data by inputting to the reconfigured transmission DNN to generate a scrambled output communication signal; analyzing (678) the scrambled output communication signal to determine whether transmission of the scrambled output communication signal will satisfy the white noise jamming level; in response to the analysis indicating that transmission of the scrambled output communication signal of the reconfigured transmission DNN will satisfy the white noise jamming level, indicating (682) an NNSI including the i-th randomization operation for reconfiguring the transmission DNN; and in response to the analysis indicating that transmission of the output communication signal of the reconfigured transmission DNN will not satisfy the white noise jamming level, updating (680) to a next i-th iteration, and repeating the generating (672), reconfiguring (674), processing (676), and analyzing (678) steps.

9. The method of claim 8, wherein, The ith randomization operation is a random permutation of an ordering of the set of neural network nodes, and the method further comprises iteratively selecting NNSIs using the transmit DNN of the first device to process the input communication data in each ith random permutation iteration.

10. The method of any one of claims 8 or 9, further comprising iteratively selecting NNSIs by the first device based on a simulation of transmission of the scrambled output communication signal over an analog communication channel.

11. The method of any preceding claim, further comprising storing any selected NNSI and corresponding whitening characteristics in a NNSI lookup table at the first device, and wherein, The selecting the NNSI further comprises retrieving an NNSI from the NNSI lookup table based on the corresponding whitening characteristic indicating a likelihood that the selected NNSI will result in transmission of the output communication signal of the transmit DNN that satisfies the white noise interference level.

12. The method of any preceding claim, further comprising: establishing a DNN connection with the second device for defining and configuring a transmit DNN and a receive DNN for performing end-to-end communication therebetween; performing DNN communication with the second device for one or more time slots using the configured transmit DNN and receive DNN of the first device; and enabling performance of the scrambling DNN operation for one or more time slots in accordance with the scrambling timing information when the output communication signal from the transmit DNN does not satisfy the white noise interference level when transmitted.

13. The method of any preceding claim, the method further comprising, prior to performing the scrambling DNN operation, identifying whether an estimated spectral density of a predicted transmission of the output communication signal satisfies the white noise interference level.

14. The method of claim 13, wherein, Identifying whether the estimated spectral density satisfies the white noise interference level further comprises: identifying the estimated spectral density of the predicted transmission as satisfying the white noise interference level when a power spectral density of the predicted transmission over a bandwidth of interest is below the white noise interference level or within a predetermined threshold region of the white noise interference level.

15. The method of claim 13 or 14, wherein, The identifying whether the estimated spectral density satisfies the white noise interference level further comprises: identifying the estimated spectral density of the predicted transmission as not satisfying the white noise interference level when the estimated spectral density of the predicted transmission forms one or more interference spikes that are higher than the white noise interference level.

16. The method of claim 13 or 14, wherein, The estimated spectral density of the predicted transmission is estimated for each antenna output of the first device.

17. The method of any one of claims 13 to 16, wherein, The identifying whether the estimated spectral density of the predicted transmission satisfies the white noise interference level comprises: performing (406a) a spectral density estimation on the predicted transmission representing the output communication signal; and analyzing (406b) a relationship of the estimated spectral density to a white noise power spectral density corresponding to the white noise interference level.

18. The method of claim 17, further comprising: comparing (406b) the estimated spectral density of the predicted transmission to a white noise spectral density associated with the white noise interference level; in response to the estimated spectral density of the predicted transmission being less than or substantially matching the white noise spectral density associated with the white noise interference level, indicating (406d) that the output communication signal, when transmitted, satisfies the white noise interference level; and in response to the estimated spectral density of the predicted transmission being greater than or substantially deviating from the white noise spectral density associated with the white noise interference level, indicating (406c) that the output communication signal, when transmitted, does not satisfy the white noise interference level.

19. The method of any one of the preceding claims, wherein, the white noise interference level is defined by: a constant amplitude of a flat white noise power spectral density over a bandwidth of interest; or a total power of a flat white noise power spectral density over a bandwidth of interest.

20. The method of any preceding claim, the method further comprising: receiving (1105a) a notification from a third device (220b) indicating that a transmission of an output communication signal to the second device exceeds an acceptable white noise interference level associated with the third device; adjusting (1105b) the white noise interference noise level to satisfy the acceptable white noise interference level; and the performing the scrambling DNN operation comprises analyzing whether the output communication signal, when transmitted, satisfies the adjusted white noise interference level.

21. The method of any preceding claim, wherein, the NNSI comprises data specifying one or more selected random permutation sequences used for randomizing one or more neural network layers of the transmission DNN, wherein the transmitting the control message comprises sending a further control message specifying the selected random permutation sequences used.

22. The method of claim 21, wherein, the transmitting the control message comprises transmitting the selected random permutation iteration or order and seed information and corresponding time slot information using control plane signaling.

23. The method of any one of claims 21 or 22, wherein, the transmitting the control message comprises transmitting each control message as a radio resource control, RRC, message, or a medium access control, MAC, message, or a downlink control information, DCI, message.

24. The method of any preceding claim, further comprising transmitting a further control message from the first device to the second device using downlink control information, DCI, the DCI being used to indicate a selected random permutation sequence iteration or order for reconfiguring the transmission DNN in accordance with the scrambling timing information.

25. A method performed by a second device in communication with a first device, the method comprising: receiving (432) a control message from the first device indicating neural network scrambling information, NNSI, and scrambling timing information; receiving (442) a communication signal from the first device transmitted in accordance with the scrambling timing information; reconfiguring (448) a receiving deep neural network, DNN, of the second device in accordance with the received NNSI and the scrambling timing information; processing (450) the received communication signal with the receiving DNN to generate reconstructed communication data represented by the received communication signal; and and sending (452) the reconstructed communication data to a data sink of the second device, or sending the reconstructed communication data to one or more upper protocol layers of a protocol stack of the second device.

26. The method of claim 25, receiving (432) the control message further comprises: receiving (432), from the first device, one or more control messages, each control message indicating an NNSI and corresponding scrambling timing information; and storing (434) the received NNSI in storage for use in relation to the scrambling timing information.

27. The method of claim 26, wherein, the scrambling timing information includes one or more time slots for receiving a transmission from the first device, the method further comprising: receiving (442) the communication signal from the first device further comprises receiving the communication signal from the first device in a particular time slot of the one or more time slots; reconfiguring (448) the receive DNN of the second device using the NNSI retrieved from storage; and processing (450) the received communication signal with the receive DNN to produce reconstructed communication data represented by the received communication signal for the particular time slot.

28. The method of any one of claims 25 to 27, wherein, the NNSI includes data that randomizes an ordering of inputs, outputs, or a set of neural network nodes of one or more neural network layers of a transmitting DNN of the first device, and reconfiguring the receive DNN of the second device further comprises using the NNSI to reconfigure the receive DNN of the second device according to the scrambling timing information to reverse the randomization applied to the inputs, the outputs, and / or the neural network layers of the transmitting DNN.

29. The method of claim 28, wherein, the NNSI specifies a random permutation of the ordering of a set of neural network nodes in the one or more neural network layers of the transmitting DNN.

30. The method of claim 29, wherein, the NNSI includes one or more random permutation parameters including: seed data, a random permutation iteration or sequence number, or an identification of a seed generation and pseudo-randomization function for use in performing the random permutation of the set of neural network nodes of one or more neural network layers, and reconfiguring the receive DNN of the second device includes: for each neural network layer of the receive DNN corresponding to each neural network layer of the transmitting DNN, generating an inverse random permutation sequence corresponding to the random permutation iteration or sequence number, wherein the inverse random permutation sequence has a length equal to a number of neural network nodes in the set of neural network nodes of the each neural network layer of the receive DNN, and for each neural network layer of the receive DNN corresponding to each neural network layer of the transmitting DNN, de-randomizing the set of neural network nodes of the each neural network layer of the receive DNN by applying the generated inverse random permutation sequence to the set of neural network nodes.

31. The method of claim 30, wherein, The seed data further comprises one or more of an identity of the first device, an identity of the second device, an identity of a cell in which the second device is located, an identity of a cell in which the first device is located, time slot information, a frame identification number, and / or any other information associated with the first or second device, and the method further comprises generating a seed from the seed data for performing an inverse random permutation of the set of neural network nodes of the neural network layer of the receive DNN corresponding to each neural network layer of the transmit DNN.

32. The method of any one of claims 25 to 31, wherein, The NNSI specifies that an l-th neural network layer of the transmit DNN is randomized, where 1 < l < L, and L is a number of neural network layers of the transmit DNN including an input layer and an output layer, and reconfiguring the receive DNN further comprises performing an inverse randomization operation on the set of neural network nodes of an (L-l+1)-th neural network layer of the receive DNN.

33. The method of claim 32, wherein, The (L-l+1)-th neural network layer of the receive DNN has a set of N neural network nodes with a particular ordering, N > 1, and performing the inverse randomization operation on the set of neural network nodes of the (L-l+1)-th neural network layer of the receive DNN comprises: generating an N-dimensional permutation matrix based on a random permutation of the (L-l+1)-th neural network layer, N > 1; inverting the N-dimensional permutation matrix to generate an inverse N-dimensional permutation matrix; and multiplying the particular ordering of the set of N neural network nodes of the (L-l+1)-th neural network layer by the inverse N-dimensional permutation matrix to form an inverse randomized ordering of the set of N neural network nodes.

34. The method of any one of claims 25 to 33, wherein, The NNSI specifies a random permutation iteration or order, seed information, and corresponding time slot information, and the control message is received using control plane signaling, where reconfiguring the receive DNN of the second device further comprises reconfiguring the receive DNN of the second device using the random permutation iteration or order and seed to process the received communication signal transmitted from the first device according to the time slot information.

35. The method of claim 34, wherein, The control message is a radio resource control, RRC, message, or a medium access control, MAC, message, or a downlink control information, DCI, message.

36. The method of any one of claims 25 to 35, further comprising: one or more further control messages are received from the first device using downlink control information, DCI, the DCI being used to indicate a selected random permutation sequence iteration or order for reconfiguring the receive DNN according to scrambling timing information and the scrambling timing information.

37. The method of any preceding claim, wherein, The first device (104a) is a base station (210), and the second device (104b) is a user equipment (220).

38. The method of any one of claims 1 to 36, wherein, The first device (104a) is a user equipment (220), and the second device (104b) is a base station (210).

39. A computer readable medium (1700) comprising computer readable instructions stored thereon, which when executed by a computer, cause the computer to perform the method of any preceding claim.

40. A first apparatus (210) comprising one or more processors (212) and memory (213), the memory storing computer-readable instructions that, when executed by the one or more processors, cause the first apparatus (210) to perform the method of any of claims 1-24 and 37-38.

41. A second apparatus (220) comprising one or more processors (222) and memory (223), the memory (223) storing computer-readable instructions that, when executed by the one or more processors (222), cause the second apparatus (210) to perform the method of any of claims 25-38.

42. An apparatus (210, 220) comprising: one or more antennas (203a / b); one or more processors (212, 222); and memory (213, 223), wherein the one or more processors (212, 222) are connected to the memory (213, 223) and the one or more antennas (203a / b), and the memory (213, 223) further stores computer- readable instructions that, when executed by the one or more processors (212, 222), cause the apparatus (210, 220) to perform the method of any of claims 1-38.

43. A communication system (200) comprising: a first apparatus (210) configured according to claim 40; and a second apparatus (220) configured according to claim 41; wherein the first apparatus (210) and the second apparatus (220) establish a deep neural network communication session.

Images