Base station for mobile networks having integrated deep learning radiofrequency signal inference models

Abstract

Various base stations for use in a radio access network are disclosed herein. The base stations can include a radio unit for transmitting digital radio signals and a base station hardware device. The base station hardware device can include one or more processors configured to implement a software radio module in communication with the radio unit, receiving the digital radio signals and an inference engine operable to receive output signals from the software radio module and perform inferences on the output signals. The inference engine can include deep learning inference models that are trained to perform the inferences and the inferences can be transmitted to a server in communication with the base station.

Description

Title: BASE STATION FOR MOBILE NETWORKS HAVING INTEGRATED DEEP LEARNING RADIOFREQUENCY SIGNAL INFERENCE MODELSCROSS-REFERENCE TO RELATED APPLICATION

[0001] The present application claims priority to U.S. Provisional Patent Application No. 63 / 553,914 which was filed February 15, 2024, the content of which is incorporated herein by reference in its entirety.FIELD

[0002] The described embodiments relate to a base station having deep learning radiofrequency signal classification models integrated thereon for use in a mobile network and associated mobile networks.BACKGROUND

[0003] Radio access networks (RAN) are a major part of modern mobile communications networks and enable wireless enabled devices to connect to other parts of mobile networks, through radio connections, enabling communications over wide areas.

[0004] In RANs, many of the functionalities traditionally performed by physical, hardware components, are performed by software modules, which allows for increased flexibility and efficiency, and reduced costs. Typical RAN architectures, however, rely on fixed hardware configurations and require manual updates and optimizations when network demand and user requirements change, making them unable to, or slow to adapt to changes and new developments in mobile network technology. Further, in many cases, these RAN architectures do not have sufficient computational resources to implement more complex functionalities.

[0005] There is a need for improvements to RAN architectures.SUMMARY

[0006] The various embodiments described herein generally relate to a base station for use in a radio access network (RAN) having deep learning radio frequency signal inference models integrated thereon and related mobile networks.

[0007] In a first aspect, in at least embodiment, there is provided a base station for use in a radio access network (RAN). The base station includes: a radio unit for transmitting digital radio signals; and a hardware device comprising one or more processors configured to implement: a software radio module in communication with the radio unit, receiving the digital radio signals; and an inference engine operable to receive output signals from the software radio module and perform inferences on the output signals.

[0008] In at least one embodiment, the software radio module is configured to: receive the digital radio signals from the radio unit; duplicate the digital radio signals to obtain a first copy of the digital radio signals and a second copy of the digital radio signals; process the first copy of the digital radio signals; and transmit the second copy of the digital radio signals to the inference engine, and the inference engine is configured to perform the inferences on the second copy of the digital radio signals.

[0009] In at least one embodiment, the software radio module is configured to preprocess the digital radio signals to obtain a derivative representation of the digital radio signals and the inference engine is operable to perform the inferences on the derivative representation of the digital radio signals.

[0010] In at least one embodiment, the derivative representation of the digital radio signals is one of: a spectrogram, a bitstream or a packetized stream.

[0011] In at least one embodiment, the inference engine is configured to: preprocess the output signals; and perform the inferences on the preprocessed output signals.

[0012] In at least one embodiment, the hardware device comprises commodity computing hardware.

[0013] In at least one embodiment, the radio unit is a 5G radio unit.

[0014] In at least one embodiment, the radio unit is an LTE radio unit.

[0015] In at least one embodiment, the one or more processors are configured to transmit the inferences to one of a core server and an external server in communication with the base station.

[0016] In at least one embodiment, the inference engine is operable to transmit the inferences to the software radio module.

[0017] In at least one embodiment, the inference engine comprises one or more deep learning inference models trained to perform the inferences.

[0018] In at least one embodiment, the inference engine is configured to perform one or more of: radio signal classification, spectrum monitoring, radio frequency fingerprinting, signal-to-noise calculation and anomaly detection.

[0019] In at least one embodiment, the radio unit is a software defined radio unit.

[0020] In at least one embodiment, the base station comprises a radio hardware device comprising one or more radio unit processors configured to implement the radio unit; and a radio inference engine for performing inferences on the digital radio signals received from the radio unit.

[0021] In at least one embodiment, the software radio module is configured to receive configuration parameters from one of an external server and a core server based on the inferences.

[0022] In a another aspect, in at least embodiment, there is provided a base station for use in a radio access network (RAN) comprising: a radio hardware device comprising one or more processor configured to implement: a radio unit; and a first inference engine for performing inferences on digital radio signals received from the radio unit; a base station hardware device comprising one or more base station hardware device processors configured to implement: a software radio module in communication with the radio unit.

[0023] In at least one embodiment, the one or more radio processors are configured to transmit the inferences to the software radio module.

[0024] In at least one embodiment, the one or more base station hardware device processors are configured to transmit the received inferences to a core server in communication with the base station.

[0025] In at least one embodiment, the one or more radio processors are configured to transmit the inferences to one of a core server and an external server in communication with the base station.

[0026] In at least one embodiment, the one or more base station hardware processors are configured to implement a second inference engine operable to receive output signals from the software radio module and perform inferences on the output signals.

[0027] In at least one embodiment, at least one of: the one or more radio processors and the one or more base station hardware processors are configured to receive inferences from an external inference engine, in communication with the base station.

[0028] In another aspect, in at least embodiment, there is provided a base station for use in a radio access network (RAN) comprising: a radio unit for transmitting digital radio signals; and a hardware device comprising one or more processors configured to implement: a software radio module in communication with the radio unit, receiving the digital radio signals; and an inference engine operable to: receive data from a receiver; and perform inferences on the data received.

[0029] In another aspect, in at least embodiment, there is provided a radio access network comprising: a base station comprising: a radio unit for transmitting digital radio signals; and a hardware device comprising one or more processors configured to implement: a software radio module in communication with the radio unit, receiving the digital radio signals; and an inference engine operable to receive output signals from the software radio module and perform inferences on the output signals; and a server in communication with the base station configured to: receive the inferences from the base station.

[0030] In at least one embodiment, the server is configured to determine configuration parameters for the base station and transmit the configuration parameters to the base station.BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Several embodiments will be described in detail with reference to the drawings, in which:FIG. 1 is a block diagram of an example mobile network, in accordance with an embodiment;FIG. 2A is a block diagram of an example base station in communication with an example core server, in accordance with an embodiment;FIG. 2B is a block diagram of the example base station of FIG. 2A, in communication with an example external server, in accordance with an embodiment;FIG. 2C is a block diagram of the example base station of FIG. 2A, in communication with an example external server, in accordance with another embodiment;FIG. 3A is a block diagram of an example base station in communication with another example core server, in accordance with another embodiment;FIG. 3B is a block diagram of the example base station of FIG. 3A, in communication with an example external server, in accordance with an embodiment;FIG. 3C is a block diagram of the example base station of FIG. 3A, in communication with an example external server, in accordance with another embodiment;FIG. 4 is a block diagram of another example base station in communication with an example core server, in accordance with another embodiment;FIG. 5A is a block diagram of another example base station in communication with an example core server, in accordance with another embodiment;FIG. 5B is a block diagram of the example base station of FIG. 5A, in communication with an example external server, in accordance with an embodiment;FIG. 50 is a block diagram of the example base station of FIG. 5A, in communication with an example external server, in accordance with another embodiment;FIG. 6A is a block diagram of another example base station in communication with an example core server, in accordance with another embodiment;FIG. 6B is a block diagram of the example base station of FIG. 6A, in communication with an example external server, in accordance with an embodiment;FIG. 60 is a block diagram of the example base station of FIG. 6A, in communication with an example external server, in accordance with another embodiment;FIG. 7 A is a block diagram the base station of FIG. 5A, in communication with the example core server of FIG. 5A and in communication with an inference engine, in accordance with another embodiment;FIG. 7B is a block diagram the base station of FIG. 5A, in communication with an example external server and in communication with an inference engine, in accordance with an embodiment; andFIG. 70 is a block diagram the base station of FIG. 5A, in communication with an example external server and in communication with an inference engine, in accordance with another embodiment; andFIG. 8 is a block diagram showing components of a software radio module and an inference engine, in accordance with an embodiment.

[0032] The drawings, described below, are provided for purposes of illustration, and not of limitation, of the aspects and features of various examples of embodiments described herein. For simplicity and clarity of illustration, elements shown in the drawings have not necessarily been drawn to scale. The dimensions of some of the elements may be exaggerated relative to other elements for clarity. It will be appreciated that for simplicity and clarity of illustration, where considered appropriate, reference numerals may be repeated among the drawings to indicate corresponding or analogous elements or steps.DESCRIPTION OF EXAMPLE EMBODIMENTS

[0033] RANs enable wireless enabled devices such as user devices to connect to broader mobile networks, allowing user devices to transmit and receive data over small and large geographical areas and form a major part of modern mobile communication networks.

[0034] In some cases, it can be advantageous to analyze the signals received or transmitted by a receiver (e.g., a radio unit) of a RAN to derive inferences (e.g., signal classification) that can be used by downstream or upstream devices. For example, it may be advantageous to classify signals received, monitor signals to detect anomalies in received signals or analyze the quality of signals received to adjust the operation of upstream or downstream devices.

[0035] Artificial intelligence (Al) and machine learning (ML) have been used to develop models analyzing these signals. Al and ML algorithms can also be used for dynamically optimizing network conditions and resource allocation and helping manage signal interference in real-time. However, conventional RAN architectures, including RAN base stations are often not configured for implementing these complex algorithms, due to, amongst other factors, their insufficient computing resources and fixed hardware configurations.

[0036] Disclosed herein are various base stations for use in mobile networks, such as a RANs. The various base stations implement one or more inference engines for analyzing RF signals and performing inferences (e.g., signal classification) on the received signals. The inference engine(s) can run one or more deep learning inference models. The inferences performed by the inference engine(s) can be used by upstream and / or downstream devices and / or components of the RANs. For example, the inferences can be used for allocating and / or adjusting resources associated with various components of the RANs. The various embodiments disclosed herein implement the inference engine(s) on the same hardware devices as one or more other components of the network (e.g., the components receiving and / or processing the RF signals). The proximity of the inference engine and the components of the network receiving and / orprocessing the RF signals allow latency to be minimized (e.g., 10 to 20 ms) and efficiency to be increased.

[0037] At least some of the embodiments described herein implement a software radio module (e.g., a base station software) and an inference engine on the same physical hardware device. The described embodiments can be deployed onto existing, readily available computer computing components, which can allow for scalability.

[0038] At least some of the embodiments described herein can be used in nonterrestrial networks (NTN). In such embodiments, the inferences performed by the inference engine(s) can be used to modify parameters of satellites (e.g., beam angle, power level, frequency, waveform configuration, resource block usage, etc.) and / or for improving spectrum sharing between various satellites.

[0039] Reference is first made to FIG. 1 , which shows a block diagram of an example mobile network 100. The mobile network 100 includes user equipment devices 110, base stations 120, a core server 130 and an external network 140. Although only two base stations 120 are shown, the mobile network 100 can include a greater number of base stations 120. Similarly, although only two user equipment devices 110 are shown to be in communication with each base station 120, it will be understood that each base station 120 can be in communication with a greater number of user equipment devices 110. Similarly, although only one core server 130 is shown, it will be understood that the mobile network 100 can include a greater number of core servers. Similarly, although only one external network 140 is shown, it will be understood that the mobile network 100 can include a greater number of external networks 140. The mobile network 100 can include more components, depending on the implementation of the mobile network 100.

[0040] The mobile network 100 can be a conventional mobile network, for example, the mobile network 100 can be a 4G LTE network, a 5G network, a 6G network, or any type of radio network that includes software components, including but not limited to software-defined radio (SDR) systems.

[0041] The user equipment devices 110 can be any wireless enabled device, for example any user device that can be used for mobile communications, including but notlimited to smart phones, feature phones, smart watches, computer tablets, laptops, desktops, or any other SIM-enabled devices and any embedded device and / or modems.

[0042] The base station 120 enable the user equipment devices 110 to connect with the core server 130 and the network 140 and can provide a range of signal processing functions, as is commonly known to those skilled in the art of base stations. The base station 120 can be implemented using a combination of hardware devices and software modules, as will be described with reference to FIGS. 2A-4.

[0043] The core server 130 can enable base stations 120 to communicate with the network 1 0 when the base stations 120 do not communicate directly with the network 140. For example, in some cases the base stations 120 can communicate directly with the network 140 and in other cases, the base stations 120 may only communicate with the network 140 via a core server 130.

[0044] The external network 140 can be any type of network that can receive data from a core server 130. For example, the external network 140 can be the internet.

[0045] Reference is next made to FIG. 2A which shows a block diagram of an example base station 120A in communication with an example core server 130, in accordance with an embodiment. As shown, the base station 120A includes a radio hardware device 127A which implements a radio unit 128 and a base station hardware device 122A. Although only one radio hardware device 127A and only one radio unit 128 is shown, there can be more than one radio hardware device 127A and / or more than one radio unit 128 implemented on the base station 120A and in communication with the software radio module 124, depending on the implementation of the base station 120.

[0046] The radio hardware device 127A can be any type of radio hardware device that has the capabilities including the processing capabilities to implement a radio unit 128, including, for example, the universal software radio peripheral USRP from Ettus Research, the ADALM-Pluto from Analog Devices and the bladeRF 2.0 micro xA4 from Nuand. In some cases, the radio hardware device 127A can include an antenna for receiving radio signals. Alternatively, the radio hardware device 127A can be in communication with an external antenna receiving radio signals and transmitting received radio signals to the radio hardware device 127A. The radio unit 128 can be asoftware defined radio (SDR) unit configured to perform the functions of any radio unit used in radio access networks as is commonly known to those skilled in the art. For example, the radio unit 128 can be configured to convert radio signals received from the antenna into digital signals and / or process the digital signals into signals that can be processed by the software radio module 124 (e.g., IQ signals). The radio unit 128 can include a radio control interface for communicating with the software radio module 124, which can transmit the digital radio signals to the base station hardware device 122A as a stream. Alternatively, the signals transmitted by the radio unit 128 can be stored in a buffer in memory and retrieved from memory by the software radio module 124.

[0047] The base station hardware device 122A can be any type of general-purpose hardware device, including workstations and desktop computers, that is capable of performing a range of functions, that can communicate with external devices and that has sufficient memory to implement one or more inference engines 126A. The base station hardware device 122A can be implemented using commodity hardware. The base station hardware device 122A can include a central processing unit. In at least one embodiment, the base hardware device includes a graphics processing unit (GPU).

[0048] The base station hardware device 122A includes at least one processor (not shown) configured to implement the software radio module 124 and an inference engine 126A. The processor(s) can be any processor(s) of the base station hardware device 122A and can be processor(s) configured to implement other modules or functionalities related to the operation of the base station hardware device 122, as is commonly known to those skilled in the art. Alternatively, the processor(s) can be dedicated processor(s) for implementing the software radio module 124 and the inference engine 126A and there can be separate processor(s) for implementing the software radio module 124 and the inference engine 126A. The processor(s) can be any type of processors with sufficient processing capabilities to implement the inference engine 126A.

[0049] The software radio module 124 can be any type of software module that is configured to receive signals from the radio unit 128 and that is configured to implement the functionalities of a software base station. In a 5G network, for example, the software radio module 124 can implement a gNodeB base station. For example, the software radiomodule can be the srsRAN project developed by Software Radio Systems. The software radio module 124 can be configured to receive radio signals from the radio unit 128. For example, the software radio module 124 can receive radio signals in the form of IQ baseband signals.

[0050] In some embodiments, as shown in FIG. 8, the software radio module 124 is configured to duplicate the received signals 808 from the radio unit 128 and transmit the duplicated signals to the inference engine 126A, for example, via a user datagram protocol (UDP) socket 812 and / or a Zero MQ publisher socket 814 of the software radio module 124. The inference engine 126A can receive the duplicated signals at a socket such as a Zero MQ subscriber socket 816 and perform inferences on the signals received using an inference model 818. In some embodiments, the processor(s) of the base station hardware device 122A preprocesses the received signals to convert the signals into a format suitable for the inference engine 126A, depending on the implementation of the inference engine 126A. In some embodiments, the preprocessing is external to the software radio module 124, that is, the preprocessing is not a functionality of the software radio module 124 and the software radio module 124 receives preprocessed signals.

[0051] The software radio module 124 can be configured to transmit the digital radio signals or a subset of the digital radio signals to the core module 134, for example, via a communication interface (e.g., network interface) of the software radio module 124 and the core module 134 can transmit the digital radio signals to an external network, for example, network 140.

[0052] The software radio module 124 can also be configured to transmit inferences from the inference engine 126A to the core module 134. For example, the inference engine 126A can be configured to transmit inferences to the software radio module 124 and the software radio module 124 can be configured to transmit the inferences to the core module. Alternatively, in at least one embodiment, the inference engine 126A is configured to directly transmit the inferences to the core module 134. In embodiments where the base station 120A is in direct communication with an external network such as external network 140, the inference engine 126A can transmit inferences to the external network 140.

[0053] In at least one embodiment, as shown in FIG. 2B, the inference engine 126A transmits the inferences to an external server 150 running an application 160 and the inferences are optionally transmitted to the core module 134. The external server 150 can be a local server or a remote server. In at least one embodiment, as shown in FIG. 2C, the external server 150 can be a cloud-based server (i.e., the external server 150 can run on a cloud 242).

[0054] In at least one embodiment, a configuration of the software module 124 can be modified based on the inferences. For example, based on the inferences, the resources of the software module 124 can be reallocated. In the embodiments shown in FIGS. 2B-2C, the application 160 can be a real-time application (e.g., xApp) that is configured to transmit configuration parameters to the software radio module 124, based on the inferences received from the inference engine 126A.

[0055] As shown, the software radio module 124 and the inference engine 126A can be implemented on the same base station hardware device 122A. In some embodiments, the software radio module 124 and the inference engine 126A are deployed using containers, for example, Docker containers. In such embodiments, the preprocessing of signals transmitted to the inference engine 126A can be performed external to the container of the software radio module 124. The inference engine 126A can perform inferences on data received from the software radio module 124 without interrupting the functionalities of the software radio module 124 and without interrupting the transmission of the digital radio signals from the software radio module 124 to the core module 134. The implementation of the software radio module 124 and the inference engine 126A on the same base station hardware device 122A can reduce latency associated with the transmission of signals from the software radio module 124 to the inference engine 126A. For example, latency can be reduced to the order of 10 to 20 ms. The low latency enabled by the configuration can be particularly advantageous for physical layer signal processing tasks such as mission critical tasks and / or time constrained tasks (e.g., beamforming, beam selection, interference management, interference reduction, channel selection, transmitter recognition, modulationrecognition, modulation, demodulation, sensing, error detection, etc.) that require low latency.

[0056] The inference engine 126A can be any kind of inference engine that can analyze data (e.g., IQ signals, signals based on IQ signals) received to perform inferences on the data, for example, output signals from the software radio module 124. The inference engine 126A can include one or more inference models that can generate inferences from the output signals. The output signals can be the digital radio signals received at the software radio module 124. For example, as explained, the software radio module 124 can duplicate digital radio signals received from the radio unit and transmit a copy of the digital radio signals received to the inference engine 126A. Alternatively, in some embodiments, the radio unit 128 can be configured to duplicate digital radio signals and transmit a first copy to the software radio module 124 and a second copy to the inference engine 126A. The signals can be pre-processed prior to being transmitted to the inference engine 126A. For example, the signals can be pre-processed by one or more processors of the base station hardware device 122A. In at least one embodiment, the output signals are digital radio signals preprocessed by the software radio module 124. For example, the software radio module 124 can be configured to process the digital radio signals received from the radio unit 128, for example, to obtain a derivative representation of the digital radio signals (e.g., spectrograms, bitstreams, packets), and transmit the processed signals to the inference engine 126A for analysis.

[0057] The inference engine 126A can be configured to perform a range of inferences tasks. For example, the inference engine 126A can perform radio signal classification, spectrum monitoring, radio frequency (RF) fingerprinting, signal-to-noise (SNR) calculations and anomaly detection. The inferences performed by the inference engine 126A can vary, depending on the downstream application. For example, the inference engine 126A can be configured for performing emitter recognition for security and authentication applications. As another example, the inference engine 126A can be configured for spectrum forecasting for channel assignment applications. As a further example, the inference engine 126A can be configured for sensing, tracking and / or monitoring objects for defense and security applications, or for an intelligenttransportation system. The one or more models can be models trained using machinelearning techniques, including deep learning techniques. The model(s) can be trained by a separate process and prior to being deployed onto the base station hardware device 122A. In some embodiments, the software radio module 124 is configured to process the digital radio signals received from the radio unit 128 to obtain derivative representations of the digital radio signals (e.g., spectrograms, bitstreams, packets). In such cases, the inference engine 126A can be configured to perform inferences on the derivative representations of the digital radio signals. The inference engine 126A can be configured to continuously perform inferences on data received or to periodically perform inferences on data received.

[0058] In some embodiments, the inference engine 126A is configured to preprocess the data received prior to performing inferences on the data. For example, the inference engine 126A can be configured to preprocess the data to change the representation of the data (e.g., from IQ samples to spectrogram data).

[0059] The core server 130A can be implemented using a core server hardware device 132. The core server hardware device 132 includes at least one processor (not shown) configured to implement a core module 134. The processor(s) can be any type of processor(s) with sufficient processing capabilities to implement the functionalities of the core module 134. The core server hardware device 132 can be any type of hardware device capable of receiving data from a component of the base station hardware device 122A, for example, the software radio module 214. The core module 134 can be any type of software module for implementing core network functions (e.g., 5G Core network functions), for example, openly available core software modules such as the Open5GS core. The core module 134 can be deployed using a container such as the Docker container. In some embodiments, the core server 130 is implemented on the base station hardware device 122A.

[0060] The core server module 134 can be configured to transmit inferences received from the software radio module 124 to the network 140 via a communication interface (not shown).

[0061] Reference is next made to FIG. 3A, which shows a block diagram of the example base station 120A of FIGS. 2A-2C, in communication with another core server 130B, in accordance with another embodiment of the invention. Reference is simultaneously made to FIGS. 3B-3C, which show block diagrams of the base station 120A in communication with an external server 150 and optionally, the core server 130B. The core server 130B shown in FIGS. 3A-3C can be substantially similar to the core server 130A shown in FIGS. 2A-2C. However, the at least one processor of the core server hardware device 132 can be additionally configured to implement an inference engine 126B. The inference engine 126B can be generally similar to the inference engine 126A. However, the inference engine 126B can be configured to obtain data from the core module 134 and perform inferences on the data from the core module 134. The inferences generated by the inference engine 126B can be different from the inferences generated by the inference engine 126A, depending on the downstream application. For example, the inference engines 126A and 126B can perform inferences on different data representations (e.g., IQ signals, spectrograms). As another example, the inference engine 126A can be configured for performing inference tasks that require low latency while the inference engine 126B can be configured for performing inference tasks that do not have latency requirements or that have less stringent latency requirements.

[0062] In some embodiments, the inference engine 126B is configured to transmit the inferences to an external network, such as network 140. Alternatively, or in addition, the inference engine 126B can be configured to transmit the inferences to the core module 134 and the core module 134 is configured to transmit the inferences to an external network, such as network 140, for use by downstream devices.

[0063] In at least one embodiment, as shown in FIG. 3B, the inference engine 126A transmits the inferences to an external server 150 running an application 160 and the inferences are optionally transmitted to the core module 134. The external server 150 can be a local server or a remote server. In at least one embodiment, as shown in FIG. 3C, the external server 150 can be a cloud-based server (i.e., the external server 150 can run on a cloud 242).

[0064] In at least one embodiment, a configuration of the software module 124 can be modified based on the inferences. For example, based on the inferences, the resources of the software module 124 can be reallocated. In the embodiments shown in FIGS. 3B-3C, the application 160 can be a real-time application (e.g., xApp) that is configured to transmit configuration parameters to the software radio module 124, based on the inferences received from the inference engine 126A.

[0065] Reference is next made to FIG. 4, which shows a block diagram of another example base station 120B in communication with a core server such as the core server 130A described with reference to FIGS. 2A-2C, in accordance with another embodiment of the invention. The base station 120B includes radio hardware 127B and a base station hardware device 122B. The base station hardware device 122B can be substantially similar to the base station hardware device 122A described with reference to FIGS. 2A3C. However, as shown, the at least one processor of the base station hardware device 122B does not implement an inference engine. Instead, in the embodiment shown in FIG. 4, the radio hardware 127B includes a radio unit 128, as described with reference to FIGS. 2A-3C and at least one processor (not shown) for implementing an inference engine 126C.

[0066] The inference engine 126C can be generally similar to the inference engines described with reference to FIGS. 2A-3C. However, as shown, the inference engine 126C can be configured to perform inferences on data from the radio unit 128. As shown the inference engine 126C and the radio unit 128 can be implemented on the same radio hardware device 127B. As shown the processor(s) implementing the inference engine 126C can be configured to transmit inferences to the software radio module 124. In at least one embodiment, the software radio module 124 is configured to transmit the received inferences to the core module 134 and / or to an external network in embodiments where the base station 120B is in direct communication with an external network. Alternatively, or in addition, inferences can be transmitted from the inference engine 126C to the core module 134. Implementing the inference engine 126C on the radio hardware device 127B can further reduce latency when compared to the embodiment of FIGS. 2A-2C since the inference engine 126C can receive data directlyfrom the radio unit 128 and the processor(s) of the radio hardware device 127B can have higher processing capabilities dedicated to implementing the inference engine 126C (e.g., since the inference engine 126C competes with fewer software modules or does not compete with other software modules for processing resources) In the embodiment shown in FIG. 4, the radio unit 128 can be configured to duplicate the signals transmitted by the radio unit 128 and transmit a first copy to the inference engine 1260 and a second copy to the software radio module 124.

[0067] Reference is next made to FIG. 5A, which shows a block diagram of another example base station 120C, in communication with the core server 130A, in accordance with another embodiment of the invention. Reference is simultaneously made to FIGS. 5B-5C which show block diagrams of the base station 120C in communication with an external server 150 and optionally, with the core server 130A. The base station 120C can include the radio hardware 127B, as described with reference to FIG. 4 and the base station hardware device 122A, as described with reference to FIGS. 2A-3C. Accordingly, in the example of FIG. 5A, inferences can be performed on data from the radio unit 128 and from the software radio module 124. The inference engines 126A and 126C can be configured for generating different types of inferences and / or for generating inferences based on different types of data. For example, the inference engine 126C can be configured to generate inferences from digital radio signals received from the radio unit 128 while the inference engine 126A can be configured to generate inferences from a derivative representation of the digital radio signals transmitted by the radio unit 128. Alternatively, the inferences performed can be different. For example, the inference engine 126A can analyze the received signal to perform radio signal classification and the inference engine 126A analyze the received signal to perform anomaly detection.

[0068] The inferences performed by the inference engine 126C can be transmitted to the software radio module 124, directly to the core module 134 and / or to an external network in embodiments where the base station 120C is in direct communication with an external network.

[0069] In at least one embodiment, as shown in FIG. 5B, the inference engine 126A transmits the inferences to an external server 150 running an application 160 and theinferences are optionally transmitted to the core module 134. The external server 150 can be a local server or a remote server. In at least one embodiment, as shown in FIG. 3C, the external server 150 can be a cloud-based server (i.e., the external server 150 can run on a cloud 242).

[0070] In at least one embodiment, a configuration of the software module 124 can be modified based on the inferences. For example, based on the inferences, the resources of the software module 124 can be reallocated. In the embodiments shown in FIGS. 5B-5C, the application 160 can be a real-time application (e.g., xApp) that is configured to transmit configuration parameters to the software radio module 124, based on the inferences received from the inference engine 126A.

[0071] Reference is now briefly made to FIG. 6A, which shows another base station 120D in communication with a core server 130A, in accordance with another embodiment of the invention. Reference is simultaneously made to FIGS. 6B-6C, which shows the base station 120D in communication with the external server 150 and in optional communication with the core server 130A. The base station 120D can be substantially similar to the base station 120A described with reference to FIGS. 2A-2C. However, as shown, the inference engine 126A can receive signals from a receiver 150 instead of the software radio module 124 and / or the radio unit 128. The receiver 150 can be configured to receive radio signals and preprocess and / or convert the radio signals into a format that is suitable for the inference engine 126A. Though the base station 120D shows a configuration including a radio hardware device 127A, a radio unit 128, a base station hardware device 122C, a software radio module 124 and an inference engine 126A, it will be understood that a receiver can be implemented in any of the configurations shown in FIGS. 3A-5. Additionally, the receiver 150 can be in communication with the inference engine 126C of FIG. 4 and the inference engine 126C can receive signals from the receiver 150. In some embodiments, the inference engine 126A receives signals from the receiver 150 and from the software radio module 124 (e.g., signals from the radio unit 128). In such embodiments, the receiver 150 can have a different capture configuration than the radio unit 128 and / or the receiver 150 caninclude antenna(s) that are different from the antenna(s) from which the radio unit 128 receives radio signals. For example, the use of multiple receivers (i.e. , the receiver 150 and the radio unit 128) can enable the base station 122D to receive a wider frequency range of signals and / or signals from a wider geographic region or from different directions, monitor a higher number of channels or adjacent channels, monitor a selection of antennas, and / or can increase the bandwidth of the signals analyzed by the inference engine 126A, or the settings for the same bandwidth .

[0072] Reference is now briefly made to FIG. 7A, which shows another base station 120E in communication with a core server 130A and an external inference engine 126D. Reference is simultaneously made to FIGS. 7B-7C, which shows the base station 120E in communication with the external server 150 and in optional communication with the core server 130A. The base station 120E can be substantially similar to the base station 120C described with reference to FIGS. 5A-5C. The external inference engine 126D can be implemented in addition to the inference engine 126A and / or the inference engine 126C and / or can be implemented as an alternative to one of inference engine 126A and inference engine 126C. The inference engine 126D can be configured to perform inferences on RF signals external to the base station 120D. In some cases, the inference engine 126D is configured to perform inferences on RF signals from an external network, such as the external network 140. The inference engine 126D can be configured to transmit inferences to the software radio module 124, the inference engine 126A and / or the core module 134.

[0073] It will be appreciated that numerous specific details are set forth in order to provide a thorough understanding of the example embodiments described herein. However, it will be understood by those of ordinary skill in the art that the embodiments described herein may be practiced without these specific details. In other instances, well- known methods, procedures and components have not been described in detail so as not to obscure the embodiments described herein. Furthermore, this description and the drawings are not to be considered as limiting the scope of the embodiments described herein in any way, but rather as merely describing the implementation of the various embodiments described herein.

[0074] It should be noted that terms of degree such as "substantially", "about" and "approximately" when used herein mean a reasonable amount of deviation of the modified term such that the end result is not significantly changed. These terms of degree should be construed as including a deviation of the modified term if this deviation would not negate the meaning of the term it modifies.

[0075] In addition, as used herein, the wording “and / or” is intended to represent an inclusive-or. That is, “X and / or Y” is intended to mean X or Y or both, for example. As a further example, “X, Y, and / or Z” is intended to mean X or Y or Z or any combination thereof.

[0076] The embodiments of the systems and methods described herein may be implemented in hardware or software, ora combination of both. These embodiments may be implemented in computer programs executing on programmable computers, each computer including at least one processor, a data storage system (including volatile memory or non-volatile memory or other data storage elements or a combination thereof), and at least one communication interface. For example and without limitation, the programmable computers (referred to below as computing devices) may be a server, network appliance, embedded device, computer expansion module, a personal computer, laptop, personal data assistant, cellular telephone, smart-phone device, tablet computer, a wireless device or any other computing device capable of being configured to carry out the methods described herein.

[0077] In some embodiments, the communication interface may be a network communication interface. In embodiments in which elements are combined, the communication interface may be a software communication interface, such as those for inter-process communication (IPC). In still other embodiments, there may be a combination of communication interfaces implemented as hardware, software, and combination thereof.

[0078] Program code may be applied to input data to perform the functions described herein and to generate output information. The output information is applied to one or more output devices, in known fashion.

[0079] Each program may be implemented in a high-level procedural or object- oriented programming and / or scripting language, or both, to communicate with a computer system. However, the programs may be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Each such computer program may be stored on a storage media or a device (e.g. ROM, magnetic disk, optical disc) readable by a general or special purpose programmable computer, for configuring and operating the computer when the storage media or device is read by the computer to perform the procedures described herein. Embodiments of the system may also be considered to be implemented as a non- transitory computer-readable storage medium, configured with a computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner to perform the functions described herein.

[0080] Furthermore, the system, processes and methods of the described embodiments are capable of being distributed in a computer program product comprising a computer readable medium that bears computer usable instructions for one or more processors. The medium may be provided in various forms, including one or more diskettes, compact disks, tapes, chips, wireline transmissions, satellite transmissions, internet transmission or downloadings, magnetic and electronic storage media, digital and analog signals, and the like. The computer useable instructions may also be in various forms, including compiled and non-compiled code.

[0081] Various embodiments have been described herein by way of example only. Various modification and variations may be made to these example embodiments without departing from the spirit and scope of the invention, which is limited only by the appended claims.

Claims

WE CLAIM:1 . A base station for use in a radio access network (RAN) comprising: a radio unit for transmitting digital radio signals; and a hardware device comprising one or more processors configured to implement: a software radio module in communication with the radio unit, receiving the digital radio signals; and an inference engine operable to receive output signals from the software radio module and perform inferences on the output signals.

2. The base station of claim 1 , wherein the software radio module is configured to: receive the digital radio signals from the radio unit; duplicate the digital radio signals to obtain a first copy of the digital radio signals and a second copy of the digital radio signals; process the first copy of the digital radio signals; and transmit the second copy of the digital radio signals to the inference engine, and wherein the inference engine is configured to perform the inferences on the second copy of the digital radio signals.

3. The base station of claim 1 , wherein the software radio module is configured to preprocess the digital radio signals to obtain a derivative representation of the digital radio signals and wherein the inference engine is operable to perform the inferences on the derivative representation of the digital radio signals.

4. The base station of claim 3, wherein the derivative representation of the digital radio signals is one of: a spectrogram, a bitstream or a packetized stream.

5. The base station of any one of claims 1 to 4, wherein the inference engine is configured to: preprocess the output signals; and perform the inferences on the preprocessed output signals.

6. The base station of any one of claims 1 to 5, wherein the hardware device comprises commodity computing hardware.

7. The base station of any one of claims 1 to 6, wherein the radio unit is a 5G radio unit.

8. The base station of any one of claims 1 to 7, wherein the radio unit is an LTE radio unit.

9. The base station of any one of claims 1 to 8, wherein the one or more processors are configured to transmit the inferences to one of a core server and an external server in communication with the base station.

10. The base station of any one of claims 1 to 9, wherein the inference engine is operable to transmit the inferences to the software radio module.11 . The base station of any one of claims 1 to 9, wherein the inference engine comprises one or more deep learning inference models trained to perform the inferences.

12. The base station of any one of claims 1 to 11 , wherein the inference engine is configured to perform one or more of: radio signal classification, spectrum monitoring, radio frequency fingerprinting, signal-to-noise calculation and anomaly detection.

13. The base station of any one of claims 1 to 12, wherein the radio unit is a software defined radio unit.

14. The base station of any one of claims 1 to 13, further comprising: a radio hardware device comprising one or more radio unit processors configured to implement the radio unit; and a radio inference engine for performing inferences on the digital radio signals received from the radio unit.

15. The base station of any one of claims 1 to 14, wherein the software radio module is configured to receive configuration parameters from one of an external server and a core server based on the inferences.

16. A base station for use in a radio access network (RAN) comprising: a radio hardware device comprising one or more processor configured to implement: a radio unit; and a first inference engine for performing inferences on digital radio signals received from the radio unit; a base station hardware device comprising one or more base station hardware device processors configured to implement: a software radio module in communication with the radio unit.

17. The base station of claim 16, wherein the one or more radio processors are configured to transmit the inferences to the software radio module.

18. The base station of claim 17, wherein the one or more base station hardware device processors are configured to transmit the received inferences to a core server in communication with the base station.

19. The base station of any one of claims 16 to 18, wherein the one or more radio processors are configured to transmit the inferences to one of a core server and an external server in communication with the base station.

20. The base station of any one of claims 16 to 19, wherein the one or more base station hardware processors are configured to implement a second inference engine operable to receive output signals from the software radio module and perform inferences on the output signals.21 . The base station of any one of claims 16 to 20, wherein at least one of: the one or more radio processors and the one or more base station hardware processors are configured to receive inferences from an external inference engine, in communication with the base station.

22. A base station for use in a radio access network (RAN) comprising: a radio unit for transmitting digital radio signals; and a hardware device comprising one or more processors configured to implement: a software radio module in communication with the radio unit, receiving the digital radio signals; and an inference engine operable to: receive data from a receiver; and perform inferences on the data received.

23. A radio access network comprising: a base station comprising: a radio unit for transmitting digital radio signals; anda hardware device comprising one or more processors configured to implement: a software radio module in communication with the radio unit, receiving the digital radio signals; and an inference engine operable to receive output signals from the software radio module and perform inferences on the output signals; and a server in communication with the base station configured to: receive the inferences from the base station.

24. The radio access network of claim 23, wherein the server is configured to determine configuration parameters for the base station and transmit the configuration parameters to the base station.

Images