Machine learning based receiver performance improvement using peak-reducing tones

Abstract

Various embodiments include a method performed within receiver circuitry of a wireless communication device for demodulating a wireless transmit waveform to recover a data tone, the method may include receiving from a transmitter a wireless transmit waveform that includes a peak-reduced tone (PRT) inserted by a PRT neural network in the transmitter, and demodulating the received wireless transmit waveform using a decoder neural network trained based on the output of the transmitter to output a recovery of the data tone. Further embodiments include exchanging information between the transmitter circuitry and the receiver circuitry to coordinate the PRT neural network used to insert the PRT in the transmitting wireless communication device and the decoder neural network used in the receiving wireless communication device to demodulate the transmit waveform received from the transmitting wireless communication device.

Description

[Technical field]

[0001] Related Applications This application claims the benefit of priority to U.S. Provisional Patent Application No. 62 / 980,776, entitled "Machine Learning Based Receiver Performance Improvement Using Peak Reduction Tones," filed February 24, 2020, and U.S. Provisional Patent Application No. 63 / 009,369, entitled "Machine Learning Based Uplink Coverage Enhancement Using Peak Reduction Tones," filed April 13, 2020. The entire contents of both provisional applications are incorporated herein by reference for all purposes. [Background technology]

[0002] In wireless communication systems, such as those defined under standards for fifth generation (5G) new radio (NR), conventional systems using orthogonal frequency division multiplexing (OFDM) waveforms suffer from large peak-to-average power ratios (PAPRs), which can require a large amount of power amplifier backoff at the expense of reduced power amplifier efficiency. Summary of the Invention [Means for solving the problem]

[0003] Various aspects include systems and methods of wireless communication performed by wireless communication devices, such as base stations and mobile wireless devices.

[0004] Various aspects may include performing operations in receiver circuitry of a receiver wireless communications device to demodulate the wireless transmit waveform to recover a data tone, including receiving from a transmitter a wireless transmit waveform including a peak reduced tone (PRT) generated by a PRT neural network in the transmitter, and demodulating the received wireless transmit waveform using the receiver neural network trained based on the output of the transmitter to output a recovery of the data tone.

[0005] Some aspects may further include selecting a receiver neural network trained on the PRT neural network used by the transmitter of the received transmit waveform from among a plurality of trained receiver neural networks stored within the receiver wireless communications device. Some aspects may further include receiving an indicator from the transmitter of the received transmit waveform of the receiver neural network trained on the PRT neural network used by the transmitter of the received transmit waveform, and selecting a receiver neural network from among a plurality of trained receiver neural networks stored within the receiver wireless communications device includes selecting the receiver neural network based on the indicator received from the transmitter. Some aspects may further include receiving weights for the receiver neural network from the transmitter, and demodulating the received transmit waveform using the trained receiver neural network based on an output of the transmitter includes demodulating the received transmit waveform using the selected receiver neural network and the weights for the receiver neural network received from the transmitter.

[0006] Some aspects may further include transmitting to the transmitter an indicator of a PRT neural network used by the transmitter to insert a PRT into the transmit waveform, where the PRT neural network associated with the indicator was used to train a receiver neural network used to demodulate the transmit waveform received from the transmitter. Some aspects may further include transmitting weights for the PRT neural network to the transmitter.

[0007] Various aspects may include performing operations within a transmitter circuitry of a transmitter wireless communications device to transmit the data tone via a wireless transmit waveform to a receiver, including processing the data tone through a PRT neural network trained to generate a PRT that when combined with the data tone results in a transmit waveform exhibiting a peak-to-average power ratio (PAPR) within a PAPR limit threshold, and combining the PRT output by the PRT neural network with the data tone to output a transmit waveform for transmission to a receiving wireless communications device configured with a receiver neural network trained based on the output of the transmitter circuitry to demodulate the wireless transmit waveform from the transmitter circuitry.

[0008] Some aspects may further include transmitting a neural network indicator configured to indicate a receiver neural network for use by the receiving wireless communication device in demodulating the transmitted wireless transmit waveform, the indicated receiver neural network trained based on the output of the transmitter circuit. Some aspects may further include transmitting weights for the indicated receiver neural network to the receiving wireless communication device. Some aspects may further include selecting a PRT neural network from among a plurality of PRT neural networks stored on the receiving wireless communication device for use in processing the data tones to generate a PRT that was used to train the receiver neural network used in the receiving wireless communication device. Some aspects may further include receiving an indicator of a PRT neural network to use from the receiving wireless communication device, where selecting a PRT neural network from among a plurality of PRT neural networks stored on the wireless communication device for use in processing the data tones to generate a PRT includes selecting a PRT neural network corresponding to the received indicator. Some aspects may further include receiving weights for the receiving wireless communication device from the receiving wireless communication device.

[0009] Further aspects may include methods of training a PRT neural network used in a transmit circuit to insert a PRT into a transmitted transmit waveform and a receiver neural network used in a receiver circuit to demodulate a received transmit waveform to recover the data tones. Such aspects may include training the PRT neural network using data tones to generate a PRT that, when inserted into a transmitted transmit waveform having the data tones, results in a time domain signal for amplification that meets a PAPR threshold limit, and training the receiver neural network to demodulate a transmit waveform received from the transmit circuit by comparing the data tone output of the receiver neural network to the data tone applied to the PRT neural network.

[0010] In some aspects, training the PRT neural network may include determining an error of the PRT neural network and updating the PRT neural network using the error. In some aspects, determining an error of the PRT neural network may include determining an error of the PRT neural network based on an amount by which the PAPR of the transmitted transmit waveform exceeds a PAPR threshold.

[0011] In some aspects, training the PRT neural network and training the receiver neural network may include training the PRT neural network and training the receiver neural network for the receiver circuitry.

[0012] Some aspects may include presenting the PRT neural network to a transmit circuit for use in transmitting a transmit waveform to a receiver circuit.

[0013] In some aspects, training the PRT neural network and training the receiver neural network may include training the PRT neural network and training the receiver neural network for a transmit circuit.

[0014] Some aspects may include indicating to the receiver circuitry a receiver neural network to use on a transmit waveform received from the transmit circuitry.

[0015] Further aspects may include wireless communication devices having transmitting and / or receiving circuitry configured to perform the operations of any of the methods summarized above. Further aspects may include wireless communication devices having means for performing the functions of any of the methods summarized above.

[0016] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the claims and, together with the general description given above and the detailed description below, serve to explain the features of the claims. [Brief description of the drawings]

[0017] [Figure 1] FIG. 1 is a system block diagram illustrating an example of a communication system suitable for implementing any of the various embodiments. [Diagram 2] FIG. 1 is a component block diagram illustrating an example computing system and wireless modem system suitable for implementing any of the various embodiments. [Diagram 3] FIG. 1 is a component block diagram illustrating an example of a software architecture including radio protocol stacks for user and control planes in wireless communications suitable for implementing any of the various embodiments. [Figure 4A]FIG. 1 is a component block diagram illustrating an example of a system configured to manage information transmission over a wireless communication to implement any of the various embodiments. [Figure 4B] FIG. 1 is a component block diagram illustrating an example of a system configured to manage information transmission over a wireless communication to implement any of the various embodiments. [Diagram 5] FIG. 1 is a component block diagram illustrating an example of an encoder circuit configured to implement machine learning and / or machine learning based encoding for improved receiver performance using peak reduced tones (PRTs) in accordance with various embodiments. [Figure 6] FIG. 1 is a component block diagram illustrating an example of a neural network decoder configured to implement machine learning and / or machine learning based decoding for improved receiver performance using PRT, in accordance with various embodiments. [Figure 7] FIG. 1 is a component signaling diagram illustrating an example of a system configured to implement machine learning based encoding and decoding for improved receiver performance using PRT, in accordance with various embodiments. [Figure 8] FIG. 1 is a component signaling diagram illustrating an example of a system configured to implement machine learning based encoding and decoding for improved receiver performance using PRT, in accordance with various embodiments. [Figure 9] FIG. 1 is a process flow diagram illustrating an example of a method for transmitting machine learning based data for receiver performance improvement using a PRT, according to various embodiments. [Figure 10] FIG. 1 is a process flow diagram illustrating an example of a method for implementing machine learning based encoding and decoding for improved receiver performance using PRT, according to various embodiments. [Figure 11A]FIG. 1 is a process flow diagram illustrating an example of a method for training a neural network to generate a PRT based on an input data signal and to extract the PRT at the transmitter side to recover the data signal on the receiver, according to various embodiments. [Figure 11B] FIG. 1 is a process flow diagram illustrating an example of a method for training a neural network to generate a PRT based on an input data signal and to extract the PRT at the transmitter side to recover the data signal on the receiver, according to various embodiments. [Figure 12] FIG. 2 is a carrier signal block diagram illustrating an example of a transmit waveform having data tones and PRTs on orthogonal subcarriers in accordance with various embodiments. [Figure 13] FIG. 1 is a component block diagram illustrating an example networked computing device. [Figure 14] FIG. 1 is a component block diagram illustrating an example wireless device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] Like reference numbers and designations in the various drawings indicate like elements.

[0019] The following description is directed to several embodiments for the purposes of describing the inventive aspects of the disclosure, however, those skilled in the art will readily recognize that the teachings herein may be applied in many different ways.

[0020] Various embodiments provide a method that may be implemented within a wireless communication device (e.g., a base station, a wireless access point, a mobile device, etc.) that utilizes a trained PRT neural network module on a transmitter side of a wireless communication link to determine an appropriate PRT for a given transmission, and an associated trained receiver neural network module on a receiver side of the wireless communication link to remove the PRT from a received signal and extract information regarding a target data tone carried by the PRT.

[0021] The described embodiments may be implemented using any of the Institute of Electrical and Electronics Engineers (IEEE) 16.11 standards, any of the IEEE 802.11 standards, the Bluetooth standard, Code Division Multiple Access (CDMA), Frequency Division Multiple Access (FDMA), Time Division Multiple Access (TDMA), Global System for Mobile Communications (GSM), GSM / General Packet Radio Service (GPRS), Enhanced Data GSM Environment (EDGE), Terrestrial Telecommunications (TETRA), Wideband-CDMA (W-CDMA), Evolution Data Optimized (EV-DO), 1xEV-DO, EV-DO Rev A, EV-DO Rev B, EV-DO Rev C, EV-DO Rev D, EV-DO Rev E, EV-DO Rev F, EV-DO Rev G, EV-DO Rev H, EV-DO Rev I ... B. may be implemented in any device, system, or network capable of transmitting and receiving radio frequency (RF) signals in accordance with High Speed ​​Packet Access (HSPA), High Speed ​​Downlink Packet Access (HSDPA), High Speed ​​Uplink Packet Access (HSUPA), Evolved High Speed ​​Packet Access (HSPA+), Long Term Evolution (LTE), AMPS, Fifth Generation (5G) New Radio (NR), or other signals used to communicate within a wireless, cellular, or Internet of Things (IoT) network, such as a system utilizing 3G, 4G, or 5G technologies or further embodiments thereof.

[0022] The term "wireless device" is used herein to refer to any one or all of wireless router devices, wireless appliances, cellular telephones, smartphones, portable computing devices, personal or mobile multimedia players, laptop computers, tablet computers, smartbooks, ultrabooks, palmtop computers, wireless email receivers, multimedia Internet-enabled cellular telephones, medical devices and equipment, biometric sensors / devices, wearable devices including smart watches, smart clothing, smart glasses, smart wristbands, and smart jewelry (smart rings, smart bracelets, etc.), entertainment devices (wireless game controllers, music and video players, satellite radio, etc.), wireless network-enabled Internet of Things (IoT) devices including smart meters / sensors, industrial manufacturing equipment, and large and small machines and appliances for home and business use, wireless communication elements in autonomous and semi-autonomous vehicles, wireless devices fixed or embedded in various mobile platforms, global positioning system devices, and similar electronic devices including memory, wireless communication components, and programmable processors.

[0023] The term "system on chip" (SOC) is used herein to refer to a single integrated circuit (IC) chip that contains multiple resources or processors integrated on a single substrate. A single SOC may include circuits for digital, analog, mixed signal, and radio frequency functions. A single SOC may also include any number of general purpose or special purpose processors (digital signal processors, modem processors, video processors, etc.), memory blocks (ROM, RAM, Flash, etc.), and resources (timers, voltage regulators, oscillators, etc.). A SOC may also include software for controlling the integrated resources and processors, as well as for controlling peripheral devices.

[0024] The term "system in package" (SIP) may be used herein to refer to a single module or package that contains multiple resources, computing units, cores, or processors on two or more IC chips, substrates, or SOCs. For example, a SIP may include a single substrate on which multiple IC chips or semiconductor dies are stacked in a vertical configuration. Similarly, a SIP may include one or more multi-chip modules (MCMs) on which multiple ICs or semiconductor dies are packaged in a unifying substrate. A SIP may also include multiple independent SOCs packaged in close proximity, coupled to each other via high-speed communication circuits, such as on a single motherboard or within a single wireless device. The proximity of the SOCs facilitates high-speed communication and sharing of memory and resources.

[0025] Traditional orthogonal frequency division multiplexing (OFDM) waveforms adopted in the 5GNR standard suffer from a large peak-to-average power ratio (PAPR). Without other mitigation of the waveform's PAPR, the transmitter may need to reduce power amplification (i.e., perform power amplifier backoff) at the expense of reduced power amplifier efficiency to avoid distortion caused by power amplifier nonlinearity. Signal processing approaches to PAPR reduction include tone reservation schemes, in which peak reduction tones (PRTs) orthogonal to data tones are used to shape the transmit waveform, i.e., the OFDM symbol containing the PRTs and data tones, in the time domain. The PRTs are designed to reduce peaks in the amplitude of the transmit waveform in the time domain. However, the relationship between the data tones and the PRTs is not known when the PRTs are found by traditional signal processing algorithms. In other words, the mapping between the data tones and the PRTs may be arbitrary. Therefore, the receiver is agnostic to the content of the PRTs, and the PRTs only present overhead to the receiver. When a PRT is paired with a data tone, the peak power of the transmit waveform is reduced at the expense of increased average transmit power and the error vector magnitude (EVM) is maintained for the data tone.

[0026] The embodiments described herein use a machine learning trained neural network configured to map data tones to PRTs for transmission from a particular transmitter and / or for transmission to a particular receiver. Machine learning may be used to train a pair of PRT neural networks and receiver neural networks. In some embodiments, the PRT neural network may be trained to pair data tones with PRTs based on input of the data tones. The receiver neural network may be trained to demodulate a transmitted waveform of an OFDM symbol having data tones and a PRT to generate a reconstruction of the data tones based on input of the transmitted waveform. In some embodiments, the PRT neural network and the receiver neural network may be trained for a particular transmitter based, for example, on the hardware configuration of the transmitter. In some embodiments, the PRT neural network and the receiver neural network may be trained for a particular receiver based, for example, on the hardware configuration of the receiver.

[0027] In some embodiments, the transmitter and receiver may share the configuration of the PRT neural network and / or the receiver neural network. For example, a transmitter, such as a wireless device, may share the configuration of the receiver neural network with a receiver, such as a base station or node, and the receiver neural network may be a receiver neural network trained in conjunction with a PRT neural network implemented by the transmitter for the receiver to implement. As a further example, the receiver may share the configuration of the PRT neural network with the transmitter, and the PRT neural network may be a PRT neural network trained in conjunction with a receiver neural network implemented by the receiver for the transmitter to implement.

[0028] In some embodiments, the transmitters and receivers may be preconfigured with multiple PRT neural networks and / or receiver neural networks, and sharing the configuration of the PRT neural networks and / or receiver neural networks may include sharing an indicator of the configuration of the PRT neural networks and / or receiver neural networks. The transmitters and / or receivers may use the indicator to select the configuration of the PRT neural networks and / or receiver neural networks.

[0029] In some embodiments, sharing the configuration of the PRT neural network and / or the receiver neural network may include sharing weights resulting from training of the PRT neural network and / or the receiver neural network. Sharing the configuration of the PRT neural network and / or the receiver neural network may include transmitting indicators and / or weights to the transmitter and / or the receiver.

[0030] In some embodiments, the transmitter may use a default PRT neural network and / or a PRT neural network selected in response to receiving an indicator of a configuration of the PRT neural network to generate a PRT for the data tones. The transmitter may generate a transmit waveform by combining the data tones with the PRT. The transmitter may transmit the transmit waveform to a receiver. The receiver may receive the transmit waveform and demodulate the transmit waveform to recover the data tones using a default receiver neural network and / or a receiver neural network selected in response to receiving an indicator of a configuration of the receiver neural network. For a PRT neural network and a receiver neural network trained together, the PRT neural network used to generate a PRT for combining with the data tones to generate a transmit waveform may be known to the receiver implementing the receiver neural network. As such, the PRT may contain information to facilitate generating a recovery of the data tones. The information contained in the PRT may improve demodulation performance, such as improving the accuracy of the recovery of the data tones, compared to transmitting a PRT generated by a traditional signal processing algorithm.

[0031] 1 is a system block diagram illustrating an example communication system 100 suitable for implementing any of the various embodiments. The communication system 100 may be a 5G New Radio (NR) network, or any other suitable network, such as a Long Term Evolution (LTE) network.

[0032] The communication system 100 may include a heterogeneous network architecture including a core network 140 and various mobile devices (shown in FIG. 1 as wireless devices 120a-120e). The communication system 100 may also include several base stations (shown as BS 110a, BS 110b, BS 110c, and BS 110d) and other network entities. A base station is an entity that communicates with wireless devices (mobile devices) and may also be referred to as a Node B, Node B, LTE evolved Node B (eNB), Access Point (AP), radio head, Transmit Receive Point (TRP), New Radio Base Station (NR BS), 5G Node B (NB), Next Generation Node B (gNB), etc. Each base station may provide communication coverage for a particular geographic area. In 3GPP, the term "cell" may refer to a base station's coverage area, a base station subsystem serving this coverage area, or a combination thereof, depending on the context in which the term is used.

[0033] The base stations 110a-110d may provide communication coverage for a macro cell, a pico cell, a femto cell, another type of cell, or a combination thereof. A macro cell may cover a relatively large geographic area (e.g., a few kilometers in radius) and may allow unrestricted access by mobile devices with service subscriptions. A pico cell may cover a relatively small geographic area and may allow unrestricted access by mobile devices with service subscriptions. A femto cell may cover a relatively small geographic area (e.g., a home) and may allow restricted access by mobile devices that have an association with the femto cell (e.g., mobile devices in a Closed Subscriber Group (CSG)). A base station for a macro cell may be referred to as a Macro BS. A base station for a pico cell may be referred to as a Pico BS. A base station for a femto cell may be referred to as a Femto BS or a Home BS. In the example shown in Figure 1, base station 110a may be a macro BS for macro cell 102a, base station 110b may be a pico BS for pico cell 102b, and base station 110c may be a femto BS for femto cell 102c. Base stations 110a-110d may support one or more (e.g., three) cells. The terms "eNB", "base station", "NR BS", "gNB", "TRP", "AP", "Node B", "5G NB", and "cell" may be used interchangeably herein.

[0034] In some examples, the cells may not be stationary and the geographic areas of the cells may move according to the location of the mobile base station. In some examples, the base stations 110a-110d may be interconnected to each other and to one or more other base stations or network nodes (not shown) in the communication system 100 through various types of backhaul interfaces, such as direct physical connections, virtual networks, or combinations thereof, using any suitable transport network.

[0035] The base stations 110a-110d may communicate with the core network 140 over wired or wireless communications links 126. The wireless devices 120a-120e may communicate with the base stations 110a-110d over wireless communications links 122.

[0036] The wired communications link 126 may use one or more wired communications protocols, such as Ethernet, Point-to-Point Protocol, High-Level Data Link Control (HDLC), Advanced Data Communication Control Protocol (ADCCP), and Transmission Control Protocol / Internet Protocol (TCP / IP), and may use a variety of wired networks (e.g., Ethernet, TV cable, telephony, fiber optics, and other forms of physical network connections).

[0037] The communication system 100 may also include a relay station (e.g., relay BS 110d). A relay station is an entity that can receive a transmission of data from an upstream station (e.g., a base station or a mobile device) and transmit the data to a downstream station (e.g., a wireless device or a base station). A relay station may also be a mobile device that can relay transmissions for other wireless devices. In the example shown in FIG. 1, relay station 110d may communicate with macro base station 110a and wireless device 120d to facilitate communication between base station 110a and wireless device 120d. A relay station may also be referred to as a relay base station, a relay base station, a repeater, etc.

[0038] The communications system 100 may be a heterogeneous network including different types of base stations, e.g., macro base stations, pico base stations, femto base stations, relay base stations, etc. These different types of base stations may have different transmit power levels, different coverage areas, and may have different impacts on interference in the communications system 100. For example, macro base stations may have high transmit power levels (e.g., 5-40 watts), whereas pico base stations, femto base stations, and relay base stations may have lower transmit power levels (e.g., 0.1-2 watts).

[0039] A network controller 130 may couple to a set of base stations and provide coordination and control for these base stations. The network controller 130 may communicate with the base stations via a backhaul. The base stations may also communicate with each other directly or indirectly, e.g., via wireless or wireline backhaul.

[0040] The wireless devices 120a, 120b, 120c may be dispersed throughout the communication system 100, and each wireless device may be fixed or mobile. A wireless device may also be referred to as an access terminal, a terminal, a mobile station, a subscriber unit, a station, etc.

[0041] The macro base station 110a may communicate with the communications network 140 over a wired or wireless communications link 126. The wireless devices 120a, 120b, 120c may communicate with the base stations 110a-110d over wireless communications links 122.

[0042] The wireless communication links 122, 124 may include multiple carrier signals, frequencies, or frequency bands, each of which may include multiple logical channels. The wireless communication links 122 and 124 may utilize one or more radio access technologies (RATs). Examples of RATs that may be used in the wireless communication links include 3GPP LTE, 3G, 4G, 5G (e.g., NR), GSM, Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Worldwide Interoperability for Microwave Access (WiMAX), Time Division Multiple Access (TDMA), and other mobile telephony communication technology cellular RATs. Further examples of RATs that may be used in one or more of the various wireless communication links 122, 124 in the communication system 100 include medium-range protocols such as Wi-Fi, LTE-U, LTE-Direct, LAA, MuLTEfire, and relatively short-range RATs such as ZigBee, Bluetooth, and Bluetooth Low Energy (LE).

[0043] Some wireless networks (e.g., LTE) utilize orthogonal frequency division multiplexing (OFDM) on the downlink and single-carrier frequency division multiplexing (SC-FDM) on the uplink. OFDM and SC-FDM partition the system bandwidth into multiple (K) orthogonal subcarriers, also commonly referred to as tones, bins, etc. Each subcarrier may be modulated with data. Generally, modulation symbols are sent in the frequency domain with OFDM and in the time domain with SC-FDM. The spacing between adjacent subcarriers may be fixed, and the total number of subcarriers (K) may be dependent on the system bandwidth. For example, the subcarrier spacing may be 15 kHz, and the minimum resource allocation (called a "resource block") may be 12 subcarriers (or 180 kHz). Thus, the nominal fast file transfer (FFT) size may be equal to 128, 256, 512, 1024, or 2048 for a system bandwidth of 1.25, 2.5, 5, 10, or 20 megahertz (MHz), respectively. The system bandwidth may also be partitioned into subbands. For example, a subband may cover 1.08 MHz (i.e., 6 resource blocks), and there may be 1, 2, 4, 8, or 16 subbands for a system bandwidth of 1.25, 2.5, 5, 10, or 20 MHz, respectively.

[0044] Although the description of some embodiments may use terminology and examples related to LTE technology, various embodiments may be applicable to other wireless communication systems, such as New Radio (NR) or 5G networks. NR may utilize OFDM with cyclic prefix (CP) on the uplink (UL) and downlink (DL) and may include support for half-duplex operation using time division duplex (TDD). A single component carrier bandwidth of 100 MHz may be supported. An NR resource block may span 12 subcarriers with a subcarrier bandwidth of 75 kHz for a duration of 0.1 milliseconds (ms). Each radio frame may be composed of 50 subframes having a length of 10 ms. Consequently, each subframe may have a length of 0.2 ms. Each subframe may indicate a link direction (i.e., DL or UL) for data transmission, and the link direction for each subframe may be dynamically switched. Each subframe may include DL / UL data as well as DL / UL control data. Beamforming may be supported and the beam direction may be dynamically configured. Multiple-input multiple-output (MIMO) transmission with precoding may also be supported. MIMO configurations in the DL may support up to eight transmit antennas with multi-layer DL transmission of up to eight streams and up to two streams per wireless device. Multi-layer transmission with up to two streams per wireless device may be supported. Aggregation of multiple cells may be supported with up to eight serving cells. Alternatively, NR may support an air interface different from the OFDM-based air interface.

[0045] Some mobile devices may be considered as machine type communication (MTC) or evolved or extended machine type communication (eMTC) mobile devices. MTC and eMTC mobile devices include, for example, a robot, a drone, a remote device, a sensor, a meter, a monitor, a location tag, etc. that may communicate with a base station, another device (e.g., a remote device), or some other entity. A wireless node may provide, for example, connectivity for or to a network (e.g., a wide area network such as the Internet or a cellular network) via a wired or wireless communication link. Some mobile devices may be considered as Internet of Things (IoT) devices or may be implemented as NB-IoT (narrowband Internet of Things) devices. Wireless devices 120a-120e may be included within a housing that houses components of the wireless device, such as a processor component, a memory component, similar components, or a combination thereof.

[0046] In general, any number of communication systems and any number of wireless networks may be deployed in a given geographic area. Each communication system and wireless network may support a particular radio access technology (RAT) and may operate on one or more frequencies. The RAT may also be referred to as a radio technology, an air interface, etc. The frequencies may also be referred to as a carrier, a frequency channel, etc. Each frequency may support a single RAT in a given geographic area to avoid interference between communication systems of different RATs. In some cases, NR or 5G RAT networks may be deployed.

[0047] 2 is a component block diagram illustrating an exemplary computing system and wireless modem system 200 suitable for implementing any of the various embodiments. The various embodiments may be implemented on a number of single-processor and multi-processor computer systems, including systems on a chip (SOC) or systems in a package (SIP).

[0048] 1 and 2, the illustrated exemplary computing system 200 (which may be a SIP in some embodiments) includes two SOCs 202, 204 coupled to a clock 206, a voltage regulator 208, and a wireless transceiver 266 configured to transmit and receive wireless communications to and from a wireless device, such as a base station 110a, via an antenna (not shown). In some implementations, the first SOC 202 operates as a central processing unit (CPU) of the wireless device, executing instructions of a software application program by performing arithmetic, logic, control, and input / output (I / O) operations specified by the instructions. In some embodiments, the second SOC 204 may operate as a dedicated processing unit. For example, the second SOC 204 may operate as a dedicated 5G processing unit responsible for managing high-capacity, high-speed (e.g., 5 Gbps, etc.), and / or very short wavelength (e.g., 28 GHz mmWave spectrum, etc.) communications.

[0049] The first SOC 202 may include a digital signal processor (DSP) 210, a modem processor 212, a graphics processor 214, an application processor 216, one or more coprocessors 218 (e.g., vector coprocessors) connected to one or more of the processors, memory 220, custom circuitry 222, system components and resources 224, an interconnect / bus module 226, one or more temperature sensors 230, a thermal management unit 232, and a thermal power envelope (TPE) component 234. The second SOC 204 may include a 5G modem processor 252, a power management unit 254, an interconnect / bus module 264, multiple mmWave transceivers 256, memory 258, and various additional processors 260, such as application processors, packet processors, etc.

[0050] Each processor 210, 212, 214, 216, 218, 252, 260 may include one or more cores, and each processor / core may perform operations independent of the other processors / cores. For example, the first SOC 202 may include a processor that executes a first type of operating system (e.g., FreeBSD, LINUX, OS X, etc.) and a processor that executes a second type of operating system (e.g., Microsoft Windows 10). Additionally, any or all of the processors 210, 212, 214, 216, 218, 252, 260 may be included as part of a processor cluster architecture (e.g., a synchronous processor cluster architecture, an asynchronous or heterogeneous processor cluster architecture, etc.).

[0051] The first SOC 202 and the second SOC 204 may include various system components, resources, and custom circuitry for managing sensor data, analog-to-digital conversion, wireless data transmission, and performing other specialized operations, such as processing encoded audio and video signals for decoding data packets and rendering in a web browser. For example, the system components and resources 224 of the first SOC 202 may include power amplifiers, voltage regulators, oscillators, phase-locked loops, peripheral bridges, data controllers, memory controllers, system controllers, access ports, timers, and other similar components used to support processors and software clients running on the wireless devices. The system components and resources 224 and / or custom circuitry 222 may also include circuitry for interfacing with peripheral devices, such as cameras, electronic displays, wireless communication devices, external memory chips, etc.

[0052] The first SOC 202 and the second SOC 204 may communicate via an interconnect / bus module 250. The various processors 210, 212, 214, 216, 218 may be interconnected to one or more memory elements 220, system components and resources 224, and custom circuitry 222, as well as a thermal management unit 232, via an interconnect / bus module 226. Similarly, the processor 252 may be interconnected to a power management unit 254, an mmWave transceiver 256, a memory 258, and various additional processors 260, via an interconnect / bus module 264. The interconnect / bus modules 226, 250, 264 may include an array of reconfigurable logic gates and / or implement a bus architecture (e.g., CoreConnect, AMBA, etc.). Communication may occur through advanced interconnects such as high-performance networks on chips (NoCs).

[0053] The first and / or second SOC 202, 204 may further include input / output modules (not shown) for communicating with resources external to the SOC, such as a clock 206 and a voltage regulator 208. Resources external to the SOC (e.g., clock 206, voltage regulator 208) may be shared by two or more of the internal SOC processors / cores.

[0054] In addition to the exemplary SIP 200 described above, various embodiments may be implemented in a wide variety of computing systems, which may include a single processor, multiple processors, multi-core processors, or any combination thereof.

[0055] FIG. 3 is a component block diagram illustrating a software architecture 300 including radio protocol stacks for user and control planes in wireless communications suitable for implementing any of the various embodiments. With reference to FIGS. 1-3, a wireless device 320 may implement the software architecture 300 to facilitate communication between the wireless device 320 (e.g., wireless devices 120a-120e, 200) and a base station 350 (e.g., base station 110a) of a communication system (e.g., 100). In various embodiments, layers in the software architecture 300 may form logical connections with corresponding layers in the software of the base station 350. The software architecture 300 may be distributed among one or more processors (e.g., processors 212, 214, 216, 218, 252, 260). Although illustrated with respect to one radio protocol stack in a multi-SIM (Subscriber Identity Module) wireless device, software architecture 300 may include multiple protocol stacks, each associated with a different SIM (e.g., two protocol stacks associated with two SIMs, respectively, in a dual-SIM wireless communication device). Although described below with respect to an LTE communication layer, software architecture 300 may support any of a variety of standards and protocols for wireless communications and / or may include additional protocol stacks supporting any of a variety of standards and protocols for wireless communications.

[0056] The software architecture 300 may include a non-access stratum (NAS) 302 and an access stratum (AS) 304. The NAS 302 may include functions and protocols for supporting packet filtering, security management, mobility control, session management, and traffic and signaling between a wireless device's SIM (e.g., SIM 204) and its core network 140. The AS 304 may include functions and protocols for supporting communications between a SIM (e.g., SIM 204) and supported access network entities (e.g., base stations). In particular, the AS 304 may include at least three layers (Layer 1, Layer 2, and Layer 3), each of which may include various sublayers.

[0057] In the user and control plane, Layer 1 (L1) of the AS 304 may be a physical layer (PHY) 306, which may oversee functions that enable transmission and / or reception over the air interface via a wireless transceiver (e.g., 256). Examples of such physical layer 306 functions may include cyclic redundancy check (CRC) attachment, coding blocks, scrambling and descrambling, modulation and demodulation, signal measurement, MIMO, etc. The physical layer may include various logical channels, including a physical downlink control channel (PDCCH) and a physical downlink shared channel (PDSCH).

[0058] In the user and control plane, Layer 2 (L2) of the AS 304 may carry the link between the wireless device 320 and the base station 350 on the physical layer 306. In various embodiments, Layer 2 may include a Medium Access Control (MAC) sublayer 308, a Radio Link Control (RLC) sublayer 310, and a Packet Data Convergence Protocol (PDCP) sublayer 312, each of which forms a logical connection that terminates at the base station 350.

[0059] In the control plane, Layer 3 (L3) of the AS 304 may include a radio resource control (RRC) sublayer 313. Although not shown, the software architecture 300 may include additional Layer 3 sublayers, as well as various upper layers above Layer 3. In various embodiments, the RRC sublayer 313 may provide functions including broadcasting system information, paging, and establishing and releasing RRC signaling connections between the wireless device 320 and the base station 350.

[0060] In various embodiments, the PDCP sublayer 312 may provide uplink functions including multiplexing between different radio bearers and logical channels, sequence numbering, handover data processing, integrity protection, ciphering, and header compression. In the downlink, the PDCP sublayer 312 may provide functions including in-order delivery of data packets, duplicate data packet detection, integrity verification, decryption, and header recovery.

[0061] In the uplink, the RLC sublayer 310 may provide segmentation and concatenation of upper layer data packets, retransmission of lost data packets, and automatic repeat request (ARQ). In the downlink, the RLC sublayer 310 functions may include reordering of data packets to compensate for out-of-order reception, reassembly of upper layer data packets, and ARQ.

[0062] In the uplink, the MAC sublayer 308 may provide functions including multiplexing between logical and transport channels, random access procedures, logical channel priorities, and Hybrid ARQ (HARQ) operations. In the downlink, the MAC layer functions may include channel mapping within a cell, demultiplexing, discontinuous reception (DRX), and HARQ operations.

[0063] While the software architecture 300 may provide functionality for transmitting data over a physical medium, the software architecture 300 may further include at least one host layer 314 for providing data transfer services to various applications in the wireless device 320. In some embodiments, the application-specific functionality provided by the at least one host layer 314 may provide an interface between the software architecture and the general-purpose processor 206.

[0064] In other embodiments, software architecture 300 may include one or more upper logical layers (e.g., transport, session, presentation, application, etc.) that provide host layer functionality. For example, in some embodiments, software architecture 300 may include a network layer (e.g., Internet Protocol (IP) layer) whose logical connections terminate at a Packet Data Network (PDN) Gateway (PGW). In some embodiments, software architecture 300 may include an application layer whose logical connections terminate at another device (e.g., an end user device, a server, etc.). In some embodiments, software architecture 300 may further include a hardware interface 316 between physical layer 306 and communications hardware (e.g., one or more radio frequency (RF) transceivers) within AS 304.

[0065] 4A and 4B are component block diagrams illustrating a system 400 configured to manage information transmission for wireless communication in accordance with various embodiments. With reference to FIGS. 1-4B, the system 400 may include a base station 402 (e.g., 120a-120e, 200, 320) and a wireless device 404 (e.g., 120a-120e, 200, 320). The base station 402 and the wireless device 404 may communicate over a wireless communication network 424 (aspects of which are shown in FIG. 1).

[0066] The base station 402 and / or the wireless device 404 may include one or more processors 428, 432 (e.g., 210, 212, 214, 216, 218, 252, 260) coupled to electronic storage 426, 430 and a wireless transceiver 266. The wireless transceiver 266 may be configured to receive messages to be transmitted in uplink transmissions from the processors 428, 432 and transmit such messages via an antenna (not shown) to the wireless communications network 424 for relay to the base station 402 and / or the wireless device 404. Similarly, the wireless transceiver 266 may be configured to receive messages from the base station 402 and / or the wireless device 404 in downlink transmissions from the wireless communications network 424 and pass the messages to the one or more processors 428, 432 (e.g., via a modem (e.g., 252) that demodulates the messages).

[0067] The processors 428, 432 may be configured with machine-readable instructions 406, 434. The machine-readable instructions 406 may include one or more instruction modules. The instruction modules may include computer program modules. The instruction modules may include one or more of the neural network training modules 410, 436, the PRT neural network module 412, the receiver neural network module 438, the wireless communication modules 414, 440, or other instruction modules.

[0068] The neural network training modules 410, 436 may be configured to train the PRT neural network and / or the receiver neural network on a data set of data tones for accurate recovery of the data tones within a threshold of error.

[0069] The PRT neural network module 412 may be configured to apply a PRT neural network to the data tones to generate a PRT for combining with the data tones to reduce the PAPR of the transmitted waveform. The PRT neural network may be trained to generate a PRT such that a trained receiver neural network can accurately generate a reconstruction of the data tones from the transmitted waveform made by combining the data tones and the PRT.

[0070] The receiver neural network module 438 may be configured to apply a receiver neural network to the transmitted waveform to accurately generate a reconstruction of the data tone. The receiver neural network may be trained to provide information to the trained receiver neural network to help the PRT generated from the trained PRT neural network accurately generate a reconstruction of the data tone from the transmitted waveform made up of a combination of the data tone and the PRT.

[0071] The wireless communication modules 414, 440 may be configured to transmit indicators of the PRT neural network and / or the receiver neural network, weights of the PRT neural network and / or the receiver neural network, and / or transmit waveforms between the wireless device 404 and the base station 402.

[0072] In some embodiments, the base station 402 and the wireless device 404 may be operatively linked via one or more electronic communications links. For example, such electronic communications links may be established at least in part via a network, such as the Internet and / or other networks. However, this example is not intended to be limiting, and the scope of the present disclosure includes embodiments in which the base station 402 and the wireless device 404 may be operatively linked via some other communications medium.

[0073] The electronic storage devices 426, 430 may include non-transitory storage media that electronically store information. The electronic storage media of the electronic storage devices 426, 430 may include one or both of a system storage device that is integral (i.e., substantially non-removable) with the base station 402 or the wireless device 404, and / or a removable storage device that is removably connectable to the base station 402 or the wireless device 404, for example, via a port (e.g., a Universal Serial Bus (USB) port, a Firewire port, etc.) or a drive (e.g., a disk drive, etc.). The electronic storage devices 426, 430 may include one or more of an optically readable storage medium (e.g., an optical disk, etc.), a magnetically readable storage medium (e.g., a magnetic tape, a magnetic hard drive, a floppy drive, etc.), a charge-based storage medium (e.g., an EEPROM, a RAM, etc.), a solid-state storage medium (e.g., a flash drive, etc.), and / or other electronically readable storage medium. The electronic storage 426, 430 may include one or more virtual storage resources (e.g., cloud storage, virtual private networks, and / or other virtual storage resources). The electronic storage 426, 430 may store software algorithms, information determined by the processors 428, 432, information received from the base station 402 or wireless device 404, or other information that enables the base station 402 or wireless device 404 to function as described herein.

[0074] The processors 428, 432 may be configured to provide information processing capabilities within the base station 402. As such, the processors 428, 432 may include one or more of a digital processor, an analog processor, a digital circuit designed to process information, an analog circuit designed to process information, a state machine, and / or other mechanisms for electronically processing information. Although the processors 428, 432 are shown as a single entity, this is for illustrative purposes only. In some embodiments, the processors 428, 432 may include multiple processing units and / or processor cores. The processing units may be physically located within the same device, or the processors 428, 432 may represent the processing functions of multiple devices operating in cooperation. The processors 428, 432 may be configured to execute the modules 408-414 and modules 436-440 and / or other modules as software, hardware, firmware, some combination of software, hardware and / or firmware, and / or other mechanisms for configuring the processing capabilities on the processors 428, 432. As used herein, the term "module" may refer to any component or set of components that performs the functions attributed to the module, which may include one or more physical processors during execution of processor-readable instructions, processor-readable instructions, circuitry, hardware, storage media, or any other component.

[0075] The description of the functionality provided by the various modules 410-414 and modules 436-440 described herein is for purposes of illustration and not intended to be limiting, as any of the modules 410-414 and modules 436-440 may provide more or less functionality than that described. For example, one or more of the modules 410-414 and modules 436-440 may be omitted, and some or all of its functionality may be provided by the other modules 410-414 and modules 436-440. As another example, the processors 428, 432 may be configured to execute one or more additional modules that may perform some or all of the functionality attributed to one of the following modules 410-414 and modules 436-440:

[0076] FIG. 5 illustrates a functional block diagram including an example encoder circuit 500 configured to implement machine learning and / or machine learning-based encoding to insert PRT into a transmit waveform, according to various embodiments. With reference to FIGS. 1-5, a transmitter (e.g., wireless device 120a-120e, 200, 320, 404) may include an encoder 500. The encoder 500 may be implemented in hardware, software executing on a processor, and / or a combination of hardware and software executing on a processor. The encoder 500 may be a standalone component of a transmitter, an integrated component of a SoC (e.g., SoC 200, 204), and / or an integrated hardware and / or software component of a processor (e.g., processors 210, 212, 214, 216, 218, 252, 260, 428). The encoder 500 may include a PRT neural network 502, a signal combiner 504, and an inverse fast Fourier transform component 506. The encoder 500 may use a machine learning trained neural network that is configured to map data tones and PRTs for transmission from a particular transmitter and / or to a particular receiver.

[0077] FIG. 6 illustrates an example decoder 600 configured to implement machine learning and / or machine learning based decoding to extract the PRT from the received transmitted waveform and use information from this process to improve the receiver performance in estimating the original data tones, according to various embodiments. With reference to FIGS. 1-6, a receiver (e.g., base station 110a-110d, 350, 402) may include the decoder 600. The decoder 600 may be implemented in hardware, software executing on a processor, and / or a combination of hardware and software executing on a processor. The decoder 600 may be a standalone component of a receiver, an integrated component of a SoC (e.g., SoC 200, 204), and / or an integrated hardware and / or software component of a processor (e.g., processors 210, 212, 214, 216, 218, 252, 256, 260, 432, etc.). The decoder 600 may include a fast Fourier transform component 602 and a receiver neural network 604. The decoder 600 may use a machine learning trained neural network configured to demodulate a transmitted waveform, including data tones and PRTs, received from a particular transmitter and / or by a particular receiver.

[0078] 5 and 6 together, machine learning may be used to train a pair of PRT neural network 502 and receiver neural network 604. The machine learning method may be implemented separately at the transmitter side and / or on the receiver, during which the transmitter and / or receiver may implement the PRT neural network 502 and the receiver neural network 604. The PRT neural network 502 may be trained to receive data tones as input and output a PRT combined with the data tones in a signal combiner 504. When the combined signal is transformed to the time domain by an inverse fast Fourier transform 506, it results in a transmit waveform with reduced peaks that reduces the PAPR and is suitable for amplification for transmission. At the receiver side, the receiver neural network 604 may be trained to demodulate the transmit waveform, which is a combination of the data tones and the PRT, to generate a reconstruction of the data tones based on the input of the transmit waveform.

[0079] In a non-limiting example, the PRT neural network 502 and receiver neural network 604 pair may be implemented as an autoencoder using unsupervised machine learning. The PRT neural network 502 and receiver neural network 604 pair may be trained such that the PRT neural network 502 generates a PRT for combining with the data tones such that the receiver neural network 604 accurately generates a reconstruction of the data tones, such as within a threshold of error. Errors in the recovered data tones may be determined by comparing the reconstruction of the data tones to the data tones. In some embodiments, the PRT neural network 502 and receiver neural network 604 may be trained for a particular transmitter, for example by being trained using the hardware configuration of the transmitter.

[0080] In some embodiments, the PRT neural network 502 and the receiver neural network 604 may be trained for a particular receiver, for example, by being trained using the hardware configuration of the receiver. For example, the PRT neural network 502 and the receiver neural network 604 pair may be trained using a particular transmitter and / or a transmitter similar to the particular transmitter, such as a transmitter using the same hardware and / or software configuration, such as a benchmark transmitter. As a further example, the PRT neural network 502 and the receiver neural network 604 pair may be trained using a particular receiver and / or a receiver similar to the particular receiver, such as a receiver using the same hardware and / or software configuration, such as a benchmark receiver.

[0081] The transmitter and receiver may share the configuration of the PRT neural network 502 and / or the receiver neural network 604. The transmitter may share the configuration of the receiver neural network 604 by wirelessly transmitting the configuration of the receiver neural network 604 to the receiver. For example, the transmitter may share the configuration of the receiver neural network 604 with the receiver for the receiver to implement, and the receiver neural network 604 may be the receiver neural network 604 trained in conjunction with the PRT neural network 502 implemented by the transmitter.

[0082] In some embodiments, the receiver may be pre-configured with multiple receiver neural networks 604. Sharing the configuration of the receiver neural network 604 may include sharing an indicator of the configuration of the receiver neural network 604. The receiver may use the indicator to select a configuration of the receiver neural network 604 from the multiple receiver neural networks 604.

[0083] In some embodiments, sharing the configuration of the receiver neural network 604 may include sharing weights resulting from training of the PRT neural network 502 and / or the receiver neural network 604. Sharing the configuration of the receiver neural network 604 may include transmitting an indicator and / or a weight to the receiver. In some embodiments, the weight may be an indicator.

[0084] The receiver may share the configuration of the PRT neural network 502 by wirelessly transmitting the configuration of the PRT neural network 502 to the transmitter. For example, the receiver may share the configuration of the PRT neural network 502 with the transmitter for implementation by the transmitter, and the PRT neural network 502 may be a trained PRT neural network 502 in conjunction with the receiver neural network 604 implemented by the receiver. In some embodiments, the transmitter may be pre-configured with multiple PRT neural networks 502. Sharing the configuration of the PRT neural network 502 may include sharing an indicator of the configuration of the PRT neural network 502. The transmitter may use the indicator to select the configuration of the PRT neural network 502 from the multiple PRT neural networks 502. In some embodiments, sharing the configuration of the PRT neural network 502 may include sharing weights resulting from training of the PRT neural network 502 and / or the receiver neural network 604. Sharing the configuration of the PRT neural network 502 may include transmitting the indicators and / or weights to a transmitter. In some embodiments, the weights may be the indicators.

[0085] The transmitter may use the PRT neural network 502 to generate a PRT for reducing the PAPR in the transmitter, and the receiver may then use a corresponding trained receiver neural network 604 in demodulating (i.e., decoding) the transmitted waveform received from the transmitter. In various embodiments, the transmitter may use the trained PRT neural network 502 to generate a PRT as the data tones are received for transmission (i.e., generate the PRT "on the fly"). By using the trained receiver neural network 604 in conjunction with (or using output from) the transmitter's PRT neural network 502, the receiver may recover the data tones from the received waveform more accurately than could be achieved using conventional demodulation circuitry. To enable this advantage, some embodiments include operations for coordination between the transmitter and receiver such that the trained PRT neural network used in the transmitter corresponds to the trained receiver neural network used in the receiver, and vice versa. In some embodiments, the transmitter may use a selected PRT neural network 502 in response to receiving an indicator of the configuration of the PRT neural network 502. In some embodiments, the transmitter may use a PRT neural network 502 that is preconfigured on the transmitter. A signal combiner 504 of the transmitter may receive the data tones and the PRT and generate a transmit waveform by combining the data tones and the PRT. The signal combiner 504 may combine the data tones and the PRT by allocating the PRT to subcarriers not allocated to the data tones. An inverse fast Fourier transform component 506 of the transmitter may receive the transmit waveform and transform the transmit waveform from the frequency domain to the time domain. The transmitter may transmit the transmit waveform in the time domain to a receiver.

[0086] The receiver may receive the transmit waveform in the time domain, and a fast Fourier transform component 602 of the receiver may convert the transmit waveform from the time domain to the frequency domain. The receiver may use a receiver neural network 604 to demodulate the transmit waveform to generate a reconstruction of the data signal. The receiver neural network 604 may receive the transmit waveform in the frequency domain, and demodulate the transmit waveform to generate a reconstruction of the data tones. In some embodiments, the receiver may use a receiver neural network 604 that is preconfigured on the receiver. In some embodiments, the receiver may use a receiver neural network 604 that is selected in response to receiving an indicator of the configuration of the receiver neural network 604.

[0087] For the PRT neural network 502 and receiver neural network 604 trained together, the means used to generate a PRT by the PRT neural network 502 for combining with the data tones may be known to a receiver implementing the receiver neural network 604. As such, the PRT may contain information for demodulating the transmit waveform that the trained receiver neural network 604 can use to generate a reconstruction of the data tones. The information contained in the PRT may enable a receiver using the trained receiver neural network 604 to improve demodulation performance, such as improving the accuracy of the reconstruction of the data tones, compared to transmitting a transmit waveform with a PRT generated by demodulation using traditional signal processing algorithms and conventional demodulation circuitry.

[0088] 7 illustrates an example of a system configured to implement machine learning-based encoding and decoding for improved receiver performance using PRT, according to various embodiments. With reference to FIGS. 1-7, a transmitter 700 (e.g., wireless device 120a-120e, 200, 320, 404) may present a receiver neural network (e.g., receiver neural network 604) to a receiver 702 (e.g., base station 110a-110d, 350, 402) for use in decoding a transmitted waveform to generate a recovery of the data tones.

[0089] The transmitter 700 and the receiver 702 may establish an uplink 704. The transmitter 700 may transmit a receiver neural network indicator 706 to the receiver 702. The receiver neural network indicator may be configured to select from a plurality of receiver neural networks to indicate to the receiver 702 a receiver neural network for decoding the transmit waveform from the transmitter 700. In some embodiments, the receiver neural network indicator may be a reference symbol, such as a flag bit in a signal, a value content in a signal, a quality of a signal, etc., configured to indicate to the receiver 702 a receiver neural network to be selected. In some embodiments, the receiver neural network indicator may be an indicator of a receiver neural network to be selected. In some embodiments, the receiver neural network indicator may be an indicator of a PRT neural network (e.g., PRT neural network 502) with which the transmitter 700 is configured to generate a transmit waveform from which the receiver 702 may determine a receiver neural network to be selected. In some embodiments, the receiver neural network indicator may be a weight for use with the receiver neural network from which the receiver 702 may determine which receiver neural network to select. In some embodiments, the receiver 702 may use the receiver neural network indicator as a value for a hash function, lookup table, data structure location, etc. to select the receiver neural network.

[0090] The transmitter 700 may transmit the weights for the receiver neural network 708 to the receiver 702. The weights may be weights for the receiver 702 for use in implementing a selected receiver neural network to demodulate the transmitted waveform to generate a recovery of the data tones. In some embodiments, the weights may be receiver neural network indicators, and transmitting the weights for the receiver neural network 708 may be combined with transmitting the receiver neural network indicator 706.

[0091] The transmitter 700 may encode the data tones into PRTs using a PRT neural network 710. The PRT neural network of the transmitter 700 used to encode the data tones may be a PRT neural network trained in conjunction with the receiver neural network. The PRT neural network may receive the data tones and generate the PRTs based on the training of the PRT neural network and the receiver neural network, which may result in the receiver neural network accurately generating a reconstruction of the data tones. A signal combiner (e.g., signal combiner 504) may receive and combine the data tones and the PRTs to generate a transmit waveform. An inverse fast Fourier transform component (e.g., inverse fast Fourier transform component 506) may receive the transmit waveform in the frequency domain and transform the transmit waveform to the time domain. The transmitter 700 may transmit 712 the transmit waveform in the time domain to the receiver 702.

[0092] The receiver 702 may receive the transmit waveform and decode the transmit waveform using a receiver neural network and weights 714. A fast Fourier transform component (e.g., fast Fourier transform component 602) may receive the transmit waveform in the time domain and transform the transmit waveform to the frequency domain. The receiver neural network used to demodulate the transmit waveform may be a receiver neural network selected based on an indicator received from the transmitter 700. The weights that the receiver 702 may use with the receiver neural network may be the weights received from the transmitter 700. The receiver neural network may receive the transmit waveform in the frequency domain and demodulate the transmit waveform using the received weights. Demodulating the transmit waveform may produce a recovery of the data tones.

[0093] 8 illustrates an example of a system configured to implement machine learning-based encoding and decoding for improved receiver performance using PRTs, according to various embodiments. With reference to FIGS. 1-8, a receiver 702 (e.g., base station 110a-110d, 350, 402) may present a PRT neural network (e.g., PRT neural network 502) to a transmitter 700 (e.g., wireless device 120a-120e, 200, 320, 404) for use in encoding data tones into PRTs to generate a transmit waveform.

[0094] The transmitter 700 and the receiver 702 may establish an uplink 704. The receiver 702 may transmit 800 a PRT neural network indicator to the transmitter 700. The PRT neural network indicator may be configured to select from a plurality of PRT neural networks and indicate to the transmitter 700 a PRT neural network for encoding data tones into PRTs by the transmitter 700. In some embodiments, the PRT neural network indicator may be a reference symbol, such as a flag bit in a signal, content of a value in a signal, quality of a signal, etc., configured to indicate to the transmitter 700 a PRT neural network to be selected. In some embodiments, the PRT neural network indicator may be an indicator of a PRT neural network to be selected. In some embodiments, the PRT neural network indicator may be an indicator of a receiver neural network (e.g., receiver neural network 604) with which the receiver 702 is configured to decode a transmit waveform from which the transmitter 700 may determine a PRT neural network to be selected. In some embodiments, the PRT neural network indicator may be a weight for use with a PRT neural network from which the transmitter 700 may determine which PRT neural network to select. In some embodiments, the transmitter 700 may use the PRT neural network indicator as a value for a hash function, lookup table, data structure location, etc. to select a PRT neural network.

[0095] The receiver 702 may transmit 802 the weights for the PRT neural network to the transmitter 700. The weights may be weights for the transmitter 700 for use in implementing the selected PRT neural network to generate the PRT and generate the transmit waveform. In some embodiments, the weights may be PRT neural network indicators and transmitting 802 the weights for the PRT neural network may be combined with transmitting 800 the PRT neural network indicator.

[0096] The transmitter 700 may encode the data tones into a PRT using a PRT neural network and weights 804. The PRT neural network that the transmitter 700 may use to encode the data tones into a PRT may be a PRT neural network selected based on an indicator received from the receiver 702. The weights that the transmitter 700 may use with the PRT neural network to encode the data tones into a PRT may be weights received from the receiver 702. The PRT neural network may receive the data tones and generate a PRT using the received weights based on training of the PRT neural network and the receiver neural network, and the PRT may result in the receiver neural network accurately generating a reconstruction of the data tones. A signal combiner (e.g., signal combiner 504) may receive and combine the data tones and the PRT to generate a transmit waveform. An inverse fast Fourier transform component (e.g., inverse fast Fourier transform component 506) may receive the transmit waveform in the frequency domain and transform the transmit waveform to the time domain. The transmitter 700 may transmit 712 a transmit waveform in the time domain to the receiver 702 .

[0097] The receiver 702 may receive the transmitted waveform and decode the transmitted waveform using a receiver neural network 806. A fast Fourier transform component (e.g., fast Fourier transform component 602) may receive the transmitted waveform in the time domain and transform the transmitted waveform to the frequency domain. The receiver neural network that the receiver 702 may use to decode the transmitted waveform may be a receiver neural network trained in conjunction with a PRT neural network. The receiver neural network may receive the transmitted waveform in the frequency domain and demodulate the transmitted waveform to generate a reconstruction of the data tones. As part of demodulating the received signal, the receiver neural network not only extracts the PRT to obtain the data tones, but also uses information specific to how the PRT was generated by the PRT neural network to improve the accuracy of the recovered data tones.

[0098] FIG. 9 illustrates a method for transmitting machine learning based data for receiver performance improvement using a PRT according to various embodiments. 1-9, the method 900 may be implemented in a computing device (e.g., base stations 110a-110d, 350, 402, wireless devices 120a-120e, 200, 320, 404, transmitter 700, receiver 702), in general purpose hardware, in dedicated hardware, in software executing on a processor (e.g., processors 210, 212, 214, 216, 218, 252, 256, 260, 428, 432), or in a combination of processors configured in software and dedicated hardware, such as processors executing software within a peak reduction encoder and / or decoder system (e.g., processors 210, 212, 214, 216, 218, 252, 256, 260, 428, 432, encoder 500, decoder 600), including other individual components and various memory / cache controllers. To encompass the alternative configurations possible in the various embodiments, the hardware that implements method 900 is referred to herein as a "wireless communication device."

[0099] At block 902, a wireless communication device may establish an uplink with another wireless communication device. Establishing the uplink may be initiated by the wireless communication device and established based on communication signals and data transmitted between the wireless communication devices. The uplink may be established via various known means, including means for establishing an uplink in a 5G NR network. In some embodiments, establishing the uplink at block 902 may occur between a transmitter and a receiver. In some embodiments, establishing the uplink at block 902 may occur between the wireless device and a base station.

[0100] The wireless communication device may transmit an indicator of the PRT neural network and / or the receiver neural network at block 904. The receiver neural network may be a receiver neural network trained in conjunction with the PRT neural network used by the wireless communication device. In some embodiments, the transmitter and / or the wireless device may transmit the receiver neural network at block 904.

[0101] In some embodiments, transmitting the PRT neural network in block 904 may be implemented by transmitting an indicator of a PRT neural network configured to indicate to the other wireless communication device which PRT neural network (e.g., PRT neural network 502) to select from among a plurality of PRT neural networks preconfigured on the other wireless communication device. In some embodiments, transmitting the receiver neural network may be implemented by transmitting an indicator of a receiver neural network (e.g., receiver neural network 604) configured to indicate to the other wireless communication device which receiver neural network to select from among a plurality of receiver neural networks preconfigured on the other wireless communication device. In some embodiments, the indicator may be a reference symbol, such as a flag bit in a signal, content of a value in a signal, quality of a signal, etc., configured to indicate to the other wireless communication device which PRT neural network and / or receiver neural network to select. In some embodiments, the indicator may be an indicator of the PRT neural network and / or receiver neural network to select. In some embodiments, the indicator may be an indicator of a PRT neural network with which the transmitting wireless communication device is configured to generate a transmit waveform from which the receiving wireless communication device may determine a receiver neural network to select. In some embodiments, the indicator may be an indicator of a receiver neural network with which the receiving wireless communication device is configured to decode a transmit waveform from which the transmitting wireless communication device may determine a PRT neural network to use to generate a PRT.In some embodiments, the indicator may be a weight for use with the PRT neural network and / or receiver neural network from which the other wireless communication device may determine which PRT neural network and / or receiver neural network to select. In some embodiments, the other wireless communication device may use the indicator as a value for a hash function, lookup table, data structure location, etc. to select the PRT neural network and / or receiver neural network.

[0102] In some embodiments, the wireless communication device may transmit the PRT neural network to other wireless communication devices. The PRT neural network may be a PRT neural network trained in conjunction with a receiver neural network used by the wireless communication device. In some embodiments, the receiver and / or the base station may transmit the PRT neural network in block 904. In some embodiments, the wireless communication device may transmit the receiver neural network to other wireless communication devices.

[0103] In block 906, the wireless communication device may transmit the weights of the PRT neural network and / or the receiver neural network. The wireless communication device may transmit the weights to other wireless communication devices. The weights may be weights for the transmitting wireless communication device to use in implementing the selected PRT neural network to generate the PRT and generate the transmit waveform. In some embodiments, the weights may be PRT neural network indicators and transmitting the weights for the PRT neural network 906 in block 906 may be combined with transmitting the PRT neural network indicator in block 904. In some embodiments, the weights may be weights for the receiving wireless communication device to use in implementing the selected receiver neural network to demodulate the transmit waveform and generate a recovery of the data tones. In some embodiments, the weights may be receiver neural network indicators and transmitting the weights for the receiver neural network in block 906 may be combined with transmitting the receiver neural network indicator in block 904. In some embodiments, the wireless communication device may transmit the PRT neural network to other wireless communication devices. In some embodiments, the receiver and / or the base station may transmit the weights of the PRT neural network in block 906. In some embodiments, the transmitter and / or the wireless device may transmit the weights of the receiver neural network in block 906.

[0104] In some embodiments, method 900 may be performed for each uplink established between a mobile wireless communication device and a base station wireless communication device. In some embodiments, method 900 may be repeatedly performed for each interval of several uplinks established between a wireless communication device and a base station wireless communication device. In some embodiments, method 900 may be repeatedly performed for uplinks established between a wireless communication device and a base station wireless communication device after a specified period of time.

[0105] FIG. 10 illustrates an example of a method for implementing machine learning based encoding and decoding for improved receiver performance using PRT, according to various embodiments. 1-10, the method 1000 may be implemented in a computing device (e.g., base stations 110a-110d, 350, 402, wireless devices 120a-120e, 200, 320, 404, transmitter 700, receiver 702), in general purpose hardware, in dedicated hardware, in software executing on a processor (e.g., processors 210, 212, 214, 216, 218, 252, 256, 260, 428, 432), or in a combination of processors configured in software and dedicated hardware, such as processors executing software within a peak reduction encoder and / or decoder system (e.g., processors 210, 212, 214, 216, 218, 252, 256, 260, 428, 432, encoder 500, decoder 600), including other individual components and various memory / cache controllers. To encompass the alternative configurations possible in the various embodiments, the hardware that performs method 1000 is referred to herein as a "wireless communication device."

[0106] At block 1002, a transmitter, a wireless device, an encoder, and / or a PRT neural network (e.g., PRT neural network 502) of a wireless communication device may receive data tones at block 1002. In some embodiments, a transmitter, a wireless device, an encoder, and / or a signal combiner (e.g., signal combiner 504) may receive data tones at block 1002. The data tones may be a stream of data, such as data packets, mapped to a sequence of modulation symbols, such as quadrature amplitude modulation (QAM) symbols (e.g., 16QAM symbols).

[0107] At block 1004, a trained PRT neural network (e.g., 502) in the wireless communication device may receive data tones as input and output one or more PRTs configured to reduce the PAPR of a signal output by an inverse fast Fourier transform circuit (e.g., 506). As described herein, the PRT neural network (e.g., 502) may be trained to generate a PRT based on the input data tones that compensate for peaks in the signal output from the inverse fast Fourier transform circuit. The PRTs may be allocated to subcarriers that are orthogonal to the subcarriers of the data tones such that the combination of the PRT and the data tones reduces the amplitude of peaks in the waveform corresponding to the data tones.

[0108] In block 1006, a signal combiner (e.g., 504) in the wireless communication device may combine the data tones and the PRT output by the trained neural network and output a combined signal directed to an inverse fast Fourier transform circuit. The PRT may be combined with the data tones to reduce the PAPR of the data tones. The PRT and data signal may be combined via frequency division multiplexing of the PRT and the data signal. The combined data signal and PRT may be referred to as a transmit waveform.

[0109] At block 1008, an inverse fast Fourier transform component (e.g., 506) within the wireless communication device may convert the processed data signal from the frequency domain to the time domain and generate a transmit waveform for amplification and transmission.

[0110] At block 1010, a transmitter of the wireless communication device may transmit the transmit waveform to another wireless communication device. In some embodiments, the wireless communication device may transmit the transmit waveform to the wireless communication device over a 5G NR network.

[0111] At block 1012, an antenna of the receiver wireless communication device may receive the transmit waveform from the transmitting wireless communication device.

[0112] At block 1014, a Fast Fourier Transform component (eg, 602) of the receiver wireless communications device may transform the transmit waveform from the time domain to the frequency domain.

[0113] At optional block 1016, the receiving wireless communication device may select a receiver neural network for demodulating the wireless signal received from the transmitter wireless communication device. The operation at block 1016 is optional since, in some embodiments, the receiver neural network may not change. Also, the operation at block 1016 may be performed prior to receiving the transmit waveform, such as during a process for establishing a wireless communication link with the transmitter. In some embodiments, the receiver wireless communication device may have already received and / or may receive a receiver neural network indicator as part of a transmitter wireless communication device transmission, as described herein in block 904 (FIG. 9) of method 900. In some embodiments, another wireless communication device may use the receiver neural network indicator as a value for a hash function, lookup table, data structure location, etc. to select a receiver neural network. In some embodiments, the receiving wireless communication device may be communicatively linked with multiple wireless communication devices and may use metadata of the transmit waveform identifying the wireless communication device to associate a receiver neural network indicator from the wireless communication device with the transmit waveform. In some embodiments, the receiver, base station, decoder, and / or receiver neural network may select a receiver neural network for demodulating the transmitted waveform in block 1016.

[0114] At block 1018, the receiver wireless communications device may process the frequency domain signals output from the fast Fourier transform component at block 1014 via a trained receiver neural network to generate a reconstruction of the data tones. As described herein, the receiver neural network in the receiver may be trained based on the output from the transmitter, including a PRT neural network trained to reduce the PAPR in a signal for transmission by the transmitting wireless communications device (i.e., the PRT neural network used on block 1004). The receiver neural network in the receiver wireless communications device may be configured to demodulate the frequency domain signals to recover the data tones. By being trained based on the output from the transmitter, including a PRT neural network, the receiver neural network can use information inherent in the PRT to recover the data tones more accurately than would be possible using conventional demodulation circuitry.

[0115] 11A and 11B show examples of methods 1100a, 1100b for training a machine learning neural network at the transmitter side to generate a PRT based on input of a data tone and to extract the PRT to recover the data tone at the receiver side according to various embodiments. 1-11B, the methods 1100a, 1100b may be implemented in a wireless communications device (e.g., base stations 110a-110d, 350, 402, wireless devices 120a-120e, 200, 320, 404, transmitter 700, receiver 702) having a neural network (e.g., 502, 602) within or coupled to a processor (e.g., processor 210, 212, 214, 216, 218, 252, 256, 260, 428, 432), or in a combination of software configured processors and dedicated hardware, such as a processor running software within a peak reduction encoder and / or decoder system (e.g., processor 210, 212, 214, 216, 218, 252, 256, 260, 428, 432, encoder 500, decoder 600) including other individual components and various memory / cache controllers. To encompass the alternative configurations possible in various embodiments, the hardware that performs the methods 1100a, 1100b is referred to herein as a "training device."

[0116] At block 1102, the training device may receive a data signal representative of a data signal that a transmitter may transmit to a receiver during wireless communication. In some embodiments, an encoder and / or a PRT neural network (e.g., PRT neural network 502) may receive the data tones at block 1102. In some embodiments, an encoder and / or a signal combiner (e.g., signal combiner 504) may receive the data tones at block 1102.

[0117] In block 1104, the training device may determine a PRT configured to reduce the PAPR of a transmit waveform to be amplified for transmission. The PRT may be a tone orthogonal to the data tones, where the combination of the PRT and the data tone is configured to reduce the amplitude of peaks in the waveform of the output signal to be amplified for transmission. In some embodiments, multiple PRTs may be determined using a conventional recursive algorithm. In some embodiments, the encoder and / or a PRT neural network may determine the PRT in block 1104.

[0118] In optional block 1106, the training device may combine the data tones with the generated PRT to generate a combined signal. The PRT and data tones may be combined via frequency division multiplexing of the PRT and the data tones. The combined data tones and PRT may be referred to as a transmit waveform. In some embodiments, an encoder and / or signal combiner may combine the data tones and PRT in block 1106. In some embodiments, the transmit waveform may be provided to a decoder and / or receiver neural network for block 1116 of method 1100b of FIG. 11B.

[0119] In optional block 1108, the training device may convert the transmit waveform from the frequency domain to the time domain. In some embodiments, in block 1108, an encoder and / or an inverse fast Fourier transform component (e.g., the inverse fast Fourier transform component 506) may convert the transmit waveform from the frequency domain to the time domain.

[0120] At block 1110, the training device may determine an error to the PRT neural network output. In some embodiments, the error to the PRT neural network output may be determined by the training device comparing the PAPR of the transmit waveform to a PAPR threshold (e.g., a maximum allowable peak-to-average power for the transmitter device) and providing the amount that exceeds the PAPR threshold as a correction error to the PRT neural network. In some embodiments, the PAPR threshold may be a general PAPR threshold for any number of data signals. In some embodiments, the PAPR threshold may be a PAPR threshold for a particular data signal. In some embodiments, the PAPR threshold may be a value that represents a threshold for the difference in the PAPR value of the data signal and the PAPR of the transmit waveform. The PAPR threshold may represent an acceptable PAPR value for amplification and transmission of the transmit waveform.

[0121] At block 1112, the training device may train the PRT neural network using the errors on the PRT neural network output. The training device may be configured to update weight values ​​of the PRT neural network used to determine the PRT to reduce one or more errors determined at block 1110. The training device may use an algorithm configured to use the error values ​​and weights as inputs and output updated weights. In some embodiments, the encoder and / or processor may train the PRT neural network at block 1112 using the errors on the PRT neural network output.

[0122] Once the transmitter PRT neural network is trained to have the transmit waveform fall within the PAPR threshold, the receiver neural network may be trained. Referring to FIG. 11B, in optional block 1114, the training device may convert the transmit waveform output by the transmitter from the time domain to the frequency domain. In some embodiments, in block 1114, a decoder and / or a fast Fourier transform component (e.g., fast Fourier transform component 602) may convert the transmit waveform from the time domain to the frequency domain. In some embodiments, the training device may receive the transmit waveform output by the training device in block 1108 of method 1100a of FIG. 11A.

[0123] In block 1116, the training device may apply the frequency domain transmit waveform to a receiver neural network to generate a recovery of the data tones that were input to the transmitter (i.e., in block 1102). The receiver neural network, with the aid of the received PRT inserted by the transmitter, demodulates the transmit waveform to recover the data tones.

[0124] At block 1118, the training device may determine an error to the data tone reconstruction. The training device may be configured to compare the data tones and the reconstruction of the data tones by various known means to determine an error value for the data tone reconstruction. For example, the reconstruction of the data tones may be compared to the quality and / or content of the data tones. As a further example, a result of processing the reconstruction of the data tones may indicate an error value to the reconstruction of the data tones and / or a comparison of the results to an expected result may indicate an error value for the data tone reconstruction. In some embodiments, the encoder, decoder, and / or processor may determine an error to the data tone reconstruction at block 1118.

[0125] In block 1120, the training device may train the receiver neural network using the errors for the recovery of the data tones. The training device may be configured to update weight values ​​of the PRT neural network used to generate the PRT and / or weight values ​​of the receiver neural network used to demodulate the transmit waveform to recover the data tones to reduce the errors. The training device may use an algorithm configured to use the error values ​​and weights as inputs and output updated weights. In some embodiments, the encoder, decoder, and / or processor may train the receiver neural network in block 1120 using the errors for the recovery of the data signal.

[0126] In some embodiments, the methods 1100a, 1100b may be repeated training both the PRT neural network and the receiver neural network together until the difference between the data tones input to the transmitter with its PRT neural network and the recovered data tones output by the receiver neural network falls within an acceptable error threshold. In some embodiments, the methods 1100a, 1100b may be repeated using multiple different data signals for multiple iterations of the methods 1100a, 1100b. The methods 1100a, 1100b may be repeated using multiple data signals until some of the multiple data signals, all of the multiple data signals, result in error values ​​that do not exceed the error threshold. Subsequent iterations of the methods 1100a, 1100b may use updated weights resulting from training the PRT neural network and / or the receiver neural network using the error values ​​at blocks 1112, 1120.

[0127] Various embodiments improve the transmitter and receiver capabilities of a wireless communication device compared to a conventional transmitter / receiver pair. By training a receiver neural network based on the output of a transmitter using a PRT neural network trained to generate the PRT, the trained receiver neural network is able to utilize information related to the PRT embedded in the transmit waveform output by the transmitter to recover the data tones more accurately (i.e., with less error) than would be possible using conventional demodulation circuitry that ignores the PRT. By training the PRT neural network and the receiver neural network together, a close connection between the encoder and the decoder may be coupled via the PRT inserted in the transmitted signal, allowing both neural networks to reduce the PAPR to within acceptable levels at the transmitter while simultaneously allowing better recovery of the data tones at the receiver than would be possible using conventional PRT generation and demodulation circuits.

[0128] FIG. 12 illustrates an example of a transmit waveform 1200 having data tones and PRTs on orthogonal subcarriers 1202, 1204, according to various embodiments. Referring to FIGS. 1-12, an encoder (e.g., encoder 500) on a transmitter (e.g., transmitter 700) may generate the transmit waveform 1200 from the data tones and PRTs generated from the data tones by a PRT neural network (e.g., PRT neural network 502). The data tones on the data tone subcarriers 1202 and the PRTs on the orthogonal subcarriers 1204 may be combined by a signal combiner (e.g., 504) of the transmitter (e.g., 700) via frequency division multiplexing. The data tones and PRTs may be transmitted by the transmitter on the orthogonal subcarriers 1202, 1204. The PRTs may be transmitted on the subcarriers 1204 that are orthogonal to the subcarriers 1202 reserved for data tones.

[0129] Various embodiments including methods 900, 1000 and 1100 may be performed in a variety of network computing devices (e.g., in base stations 110a-110d, 350, 402), an example of which is shown in FIG. 13, which illustrates an example of a network computing device 1300. Referring to FIGS. 1-12, the network computing device 1300 may function as a network element of a communication network, such as a base station. The network computing device 1300 may include a processor 1301 coupled to a volatile memory 1302 and a large capacity non-volatile memory, such as a disk drive 1303. The network computing device 1300 may also include a peripheral memory access device, such as a floppy disk drive, compact disk (CD), or digital video disk (DVD) drive 1306, coupled to the processor 1301. The network computing device 1300 may also include a network access port 1304 (or interface) coupled to the processor 1301 for establishing a data connection with a network, such as the Internet or a local area network coupled to other system computers and servers. The networked computing device 1300 may include one or more antennas 1307 for sending and receiving electromagnetic radiation, which may be connected to a wireless communications link. The networked computing device 1300 may include additional access ports, such as USB, Firewire, Thunderbolt, etc., for coupling to peripherals, external memory, or other devices.

[0130] Various embodiments, including methods 900, 1000, and 1100, may be performed in a variety of wireless devices (e.g., wireless devices 120a-120e, 200, 320, 404), an example of which is shown in FIG. 14, which illustrates an example of a wireless device 1400 suitable for use with various embodiments. With reference to FIGS. 1-12, the wireless device 1400 may include a first SOC 202 (e.g., SOC-CPU) coupled to a second SOC 204 (e.g., a 5G-enabled SOC). The first SOC 202 and the second SOC 204 may be coupled to an internal memory 426, 1416, a display 1412, and a speaker 1414. Additionally, the wireless device 1400 may include an antenna 1404 for transmitting and receiving electromagnetic radiation, which may be connected to a wireless data link and / or cellular telephone transceiver 266 coupled to one or more processors in the first SOC 202 and / or second SOC 204. The wireless device 1400 may also include menu selection buttons or rocker switches 1420 for receiving user input.

[0131] The wireless device 1400 may also include a voice encoding / decoding (codec) circuit 1410 that digitizes sound received from the microphone into data packets suitable for wireless transmission and decodes the received sound data packets to generate analog signals that are provided to a speaker to generate sound. One or more of the processors, the wireless transceiver 266, and the codec 1410 in the first SOC 202 and the second SOC 204 may also include digital signal processor (DSP) circuitry (not separately shown).

[0132] The processors of the network computing device 1300 and the wireless device 1400 may be any programmable microprocessor, microcomputer, or one or more multi-processor chips that may be configured by software instructions (applications) to perform various functions, including those of the various embodiments described below. In some mobile devices, multiple processors may be provided, such as one processor in the SOC 204 dedicated to wireless communication functions and one processor in the SOC 202 dedicated to running other applications. Software applications may be stored in memory 426, 430, 1416 before being accessed and loaded into the processor. The processors may include sufficient internal memory to store application software instructions.

[0133] As used in this application, terms such as "component," "module," and "system" are intended to include computer-related entities, such as, but not limited to, hardware, firmware, a combination of hardware and software, software, or software in execution, configured to perform certain operations or functions. For example, a component may be, but is not limited to, a process running on a processor, a processor, an object, an executable, a thread of execution, a program, or a computer. As an example, both an application running on a wireless device and the wireless device may be referred to as a component. One or more components may reside within a process or thread of execution, and a component may be localized on one processor or core or distributed among two or more processors or cores. In addition, these components may execute from various non-transitory computer-readable media having various instructions or data structures stored thereon. Components may communicate by way of local or remote processes, function or procedure calls, electronic signals, data packets, memory reads / writes, and other known network, computer, processor, or process related communication methods.

[0134] A number of different cellular and mobile communication services and standards are available or are contemplated in the future, all of which may implement and benefit from the various embodiments. Such services and standards include, but are not limited to, Third Generation Partnership Project (3GPP), Long Term Evolution (LTE) systems, third generation wireless mobile communication technologies (3G), fourth generation wireless mobile communication technologies (4G), fifth generation wireless mobile communication technologies (5G), Global System for Mobile Communications (GSM), Universal Mobile Telecommunications System (UMTS), 3GSM, General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA) systems (cdmaOne, CDMA1020™, etc.), Enhanced Data Rates for GSM Evolution (EDGE), Advanced Mobile Phone System (AMPS), Digital AMPS (IS-136 / TDMA), Evolution Data Optimized (EV-DO), Digital Enhanced Cordless Telecommunications (DECT), and the like. Examples of common standards include wireless telecommunications, Worldwide Interoperability for Microwave Access (WiMAX), Wireless Local Area Network (WLAN), Wi-Fi Protected Access I & II (WPA, WPA2), and Integrated Digital Enhanced Network (iDEN). Each of these technologies involves, for example, the transmission and reception of voice, data, signaling, or content messages. Any reference to terminology or technical details regarding a particular telecommunications standard or technology is for illustrative purposes only and is not intended to limit the scope of the claims to any particular communications system or technology unless specifically recited in the claim language.

[0135] The various embodiments shown and described are provided merely as examples to illustrate various features of the claims. However, features shown and described with respect to any given implementation are not necessarily limited to the associated implementation and may be used with or combined with other embodiments shown and described. Furthermore, the claims are not intended to be limited by any one exemplary implementation. For example, one or more of the operations of the methods disclosed herein may be substituted for or combined with one or more operations of the methods disclosed herein.

[0136] As used herein, a phrase referring to "at least one of" a list of items refers to any combination of those items, including single members. As an example, "at least one of a, b, or c" is intended to encompass a, b, c, ab, ac, bc, and abc.

[0137] The various example logic, logic blocks, modules, components, circuits, and algorithmic operations described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or a combination of both. Interchangeability of hardware and software has been described generally in terms of functionality and illustrated in the various example components, blocks, modules, circuits, and processes described above. Whether such functionality is implemented as hardware or software depends on the particular application and design constraints imposed on the overall system.

[0138] The hardware and data processing devices used to implement the various example logic, logic blocks, modules, and circuits described in connection with the aspects disclosed herein may be implemented or performed with general purpose single-chip or multi-chip processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, or any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. In some embodiments, particular processes and methods may be performed by circuitry specific to a given function.

[0139] In one or more aspects, the functions described may be implemented in hardware, digital electronic circuitry, computer software, firmware, or any combination thereof, including the structures disclosed herein and their structural equivalents. Embodiments of the subject matter described herein may also be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a computer storage medium for execution by or to control the operation of a data processing apparatus.

[0140] If implemented in software, the functions may be stored on or transmitted over a computer-readable medium as one or more instructions or code. The processes of the methods or algorithms disclosed herein may be implemented in a processor-executable software module that may reside on a computer-readable medium. Computer-readable media includes both computer storage media and communication media, including any medium that may enable a computer program to be transferred from one place to another. A storage medium may be any available medium that may be accessed by a computer. By way of example, and not limitation, such computer-readable media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that may be used to store desired program code in the form of instructions or data structures and that may be accessed by a computer. Also, any connection may be properly referred to as a computer-readable medium. As used herein, disk and disc include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk, and Blu-ray disc, where disks typically reproduce data magnetically and discs reproduce data optically using lasers. Combinations of the above should also be included within the scope of computer-readable media. In addition, operations of a method or algorithm may reside on machine-readable and computer-readable media that may be incorporated into a computer program product as one or any combination or set of code and instructions.

[0141] Various modifications to the embodiments described in this disclosure may be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the spirit or scope of the disclosure. Thus, the claims are not intended to be limited to the embodiments shown herein, but are to be accorded the widest scope consistent with this disclosure, the principles and novel features disclosed herein.

[0142] In addition, those skilled in the art will readily appreciate that the terms "upper" and "lower" may be used to facilitate description of the figures, and refer to relative positions that correspond to the orientation of the figure on a properly oriented page, and may not reflect the proper orientation of any implemented device.

[0143] Some features described herein in the context of separate embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented in multiple embodiments separately or in any suitable subcombination. Furthermore, although features may be described above as working in some combinations and may even initially be claimed as such, one or more features from a claimed combination may, in some cases, be deleted from the combination, and the claimed combination may be directed to a subcombination or a variation of the subcombination.

[0144] Similarly, although operations are shown in the figures in a particular order, this should not be understood as requiring such operations to be performed in the particular order or sequentially shown, or that all of the illustrated operations be performed, to achieve desirable results. Additionally, the figures may generally depict many exemplary processes in the form of flow diagrams. However, other operations not shown may be incorporated into the exemplary processes depicted in the schematic. For example, one or more additional operations may be performed before, after, simultaneously with, or between any of the depicted operations. In some circumstances, multitasking and parallel processing may be advantageous. Additionally, the separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems may generally be integrated together in a single software product or packaged in multiple software products. Additionally, other implementations are within the scope of the following claims. In some cases, the actions recited in the claims may be performed in a different order and still achieve desirable results. [Explanation of symbols]

[0145] 100 Communication Systems 102a Macro Cell 102b Picocell 102c Femtocell 110a Base station (BS) 110b BS 110c BS 110d BS 120a Wireless Device 120b Wireless Devices 120c Wireless Device 120d Wireless Device 120e Wireless Device 122 Wireless Communication Links 124 Wireless Communication Links 126 Wired Communication Links 130 Network Controller 140 Core Network 200 Wireless Modem System 202 System on Chip (SOC) 204 SOC 206 Clock 208 Voltage Regulator 210 Digital Signal Processor (DSP 212 Modem Processor 214 Graphics Processor 216 Application Processor 218 Coprocessor 220 Memory 222 Custom Circuit 224 System Components and Resources 226 Interconnect / Bus Modules 230 Temperature Sensor 232 Thermal Management Unit 234 Thermal Power Envelope Components 250 Interconnect / Bus Modules 252 5G modem processor 254 Power Management Unit 256 mmWave transceiver 258 memory 260 additional processors 264 Interconnect / Bus Modules 266 Wireless Transceiver 300 Software Architecture 302 Non-Access Layer (NAS) 304 Access Layer (AS) 306 Physical Layer (PHY) 308 Medium Access Control (MAC) Sublayer 310 Radio Link Control (RLC) Sublayer 312 Packet Data Convergence Protocol (PDCP) Sublayer 313 Radio Resource Control (RRC) Sublayer 314 Host Layer 316 Hardware Interface 320 Wireless Devices 350 base station 400 Systems 402 Base Station 404 Wireless Devices 406 Machine-Readable Instructions 410 Neural Network Training Module 412 PRT Neural Network Module 414 Wireless Communication Module 424 Wireless Communication Network 426 Electronic storage devices, internal memory 428 processor 430 Electronic storage devices 432 processor 434 Machine-Readable Instructions 436 Neural Network Training Module 438 Receiver Neural Network Module 440 Wireless Communication Module 500 Encoder Circuit 502 PRT Neural Network 504 Signal Combiner 506 Inverse Fast Fourier Transform Component 600 Decoder 602 Fast Fourier Transform Component 604 Receiver Neural Network 700 Transmitter 702 Receiver 704 Uplink 706 Receiver Neural Network Indicator 708 Receiver Neural Network 710 PRT Neural Network 1200 Transmit waveform with data tone and PRT 1202 Data Tone Subcarrier 1204 Orthogonal Subcarriers 1300 Network Computing Device 1301 Processor 1302 Volatile Memory 1303 Disk Drive 1304 Network Access Port 1306 Digital Video Disc (DVD) Drive 1307 Antenna 1400 Wireless Devices 1404 Antenna 1410 Voice encoding / decoding (codec) circuit 1412 Display 1414 Speaker 1416 Internal Memory 1420 Rocker Switch

Claims

1. 1. A method performed within a receiver circuit of a wireless communication device for demodulating a wireless transmission waveform to recover data tones, comprising: receiving, from a transmitter, a wireless transmit waveform including a peak reduced tone (PRT) generated by a PRT neural network within the transmitter; demodulating the received wireless transmission waveform using a receiver neural network trained based on the output of the transmitter to output a reconstruction of the data tones; selecting a receiver neural network trained on the PRT neural network used by the transmitter of the received transmit waveform from among a plurality of trained receiver neural networks stored within the wireless communication device; A method comprising:

2. The method of claim 1, further comprising the step of receiving weights for the receiver neural network from the transmitter, and wherein the step of demodulating the received transmit waveform using the receiver neural network trained based on the output of the transmitter comprises the step of demodulating the received transmit waveform using the selected receiver neural network and the weights for the receiver neural network received from the transmitter.

3. 2. The method of claim 1, further comprising receiving, from the transmitter of the received transmit waveform, an indicator of the receiver neural network trained on the PRT neural network being used by the transmitter of the received transmit waveform, and wherein selecting a receiver neural network from among a plurality of trained receiver neural networks stored within the wireless communication device comprises selecting the receiver neural network based on the indicator received from the transmitter.

4. 4. The method of claim 3, further comprising receiving weights for the receiver neural network from the transmitter, and wherein demodulating the received transmit waveform using the receiver neural network trained based on the output of the transmitter comprises demodulating the received transmit waveform using the selected receiver neural network and the weights for the receiver neural network received from the transmitter.

5. 2. The method of claim 1, further comprising transmitting to the transmitter an indicator of a PRT neural network used by the transmitter to insert a PRT into a transmit waveform, the PRT neural network associated with the indicator being used to train the receiver neural network used to demodulate a transmit waveform received from the transmitter.

6. The method of claim 5 further comprising the step of transmitting weights for the PRT neural network to the transmitter.

7. 1. A method performed within a transmitter circuit of a wireless communication device for transmitting a data tone via a wireless transmit waveform to a receiver, the method comprising: processing the data tones through a peak-to-average power ratio (PRT) neural network trained to generate peak-reduced tones (PRTs) that when combined with the data tones result in a transmit waveform exhibiting a PAPR within a PAPR limit threshold; combining the PRT output by the PRT neural network with the data tones to output a transmit waveform for transmission to a receiving wireless communications device configured with a receiver neural network trained based on the output of the transmitter circuitry to demodulate a wireless transmit waveform from the transmitter circuitry; transmitting a neural network indicator configured to indicate a receiver neural network for use by the receiving wireless communication device in demodulating the transmitted wireless transmit waveform; Including, The illustrated receiver neural network is trained based on the output of the transmitter circuitry. method.

8. The method of claim 7, further comprising transmitting weights for the indicated receiver neural network to the receiving wireless communication device.

9. 8. The method of claim 7, further comprising the step of selecting a PRT neural network from among a plurality of PRT neural networks stored on the wireless communication device for use in processing the data tones to generate a PRT used to train the receiver neural network used in the receiving wireless communication device.

10. 10. The method of claim 9, further comprising receiving a weight for the receiving wireless communication device from the receiving wireless communication device.

11. 10. The method of claim 9, further comprising receiving an indicator of the PRT neural network to use from the receiving wireless communication device, and wherein selecting a PRT neural network from among a plurality of PRT neural networks stored on the wireless communication device for use in processing the data tones to generate a PRT comprises selecting the PRT neural network that corresponds to the received indicator.

12. 12. The method of claim 11, further comprising receiving a weight for the receiving wireless communication device from the receiving wireless communication device.

13. A wireless communication device, comprising:

13. A wireless communication device comprising: a processing device configured with processor-executable instructions for performing operations configured to perform the method of any one of claims 1 to 6 and claims 7 to 12.

Images