Multistage impedance tuning in radio frequency transmitters

Abstract

Certain aspects of the present disclosure provide techniques for multi-stage impedance tuning for radio frequency (RF) transmitters. An example apparatus includes RF circuitry comprising: an amplifier; an impedance tuning circuit coupled to an output of the amplifier; and an antenna tuner coupled between an antenna feed and an output of the impedance tuning circuit. The apparatus also includes one or more memories and one or more processors coupled to the one or more memories. The one or more processors are configured to cause the apparatus to: configure at least one parameter of the impedance tuning circuit or the amplifier based at least in part on at least one load characteristic at the antenna feed; and communicate one or more signals via the RF circuitry while the at least one parameter is configured based at least in part on the at least one load characteristic at the antenna feed.

Description

Cross-references to related applications

[0001] This patent application claims priority and benefit to U.S. Patent Application No. 18 / 611,139, filed March 20, 2024, which is expressly incorporated herein by reference in its entirety. Technical Field

[0002] Various aspects of this disclosure relate to wireless communication, and more specifically, to impedance tuning of radio frequency (RF) transmitters. Background Technology

[0003] Wireless communication systems are widely deployed to provide a variety of telecommunications services, such as telephone, video, data, messaging, and broadcasting. These wireless communication systems may employ multiple access technologies that enable communication with multiple users by sharing available wireless communication system resources. Wireless communication devices may transmit RF signals via any of a variety of suitable radio access technologies (RATs) (including but not limited to 5G New Radio (NR), Evolved Universal Terrestrial Radio Access (E-UTRA), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Wideband CDMA (WCDMA), Global System for Mobile Communications (GSM), Bluetooth, Bluetooth Low Energy (BLE), ZigBee, Wireless Local Area Network (WLAN) RATs (e.g., the IEEE 802.11 specification), and any future RATs).

[0004] In some cases, wireless communication devices are equipped with an RF transceiver (also known as an RF front-end) for transmitting received radio frequency (RF) signals. Typically, modulation techniques such as phase shift keying (PSK) or any other suitable modulation technique are used to modulate the baseband signal to transmit information. In transmit mode, the RF transceiver is responsible for multiplexing the baseband signal with the RF carrier signal transmitted over the air (e.g., via a wireless communication channel). This operation is called up-conversion. In receive mode, the RF transceiver converts the received RF signal back into a baseband signal. This operation is called down-conversion. The received baseband signal can then be demodulated into the information encoded at the transmitter. An RF transceiver may consist of cascaded components in the transmit and receive chains, respectively. The cascaded components may include one or more of, for example, attenuators, switches, couplers, filters, mixers, amplifiers, frequency synthesizers, oscillators, antenna tuners, duplexers, double-ended converters, detectors, etc. Summary of the Invention

[0005] Some aspects provide an apparatus configured for wireless communication. The apparatus includes radio frequency (RF) circuitry comprising: one or more amplifiers; one or more impedance tuning circuits coupled to one or more outputs of the one or more amplifiers; and an antenna tuner coupled between an antenna feed and one or more outputs of the one or more impedance tuning circuits. The apparatus also includes one or more memories. The apparatus further includes one or more processors coupled to the one or more memories. The one or more processors are configured to cause the apparatus to: configure at least one parameter of at least one of the one or more impedance tuning circuits or one or more amplifiers, at least in part, based on at least one load characteristic at the antenna feed; and transmit one or more signals via the RF circuitry when the at least one parameter is configured at least in part based on the at least one load characteristic at the antenna feed.

[0006] Some aspects provide a method for wireless communication by a device. The method includes configuring at least one parameter of at least one of one or more impedance-tuning circuits or one or more amplifiers, at least in part based on at least one load characteristic at an antenna feed, wherein the device includes radio frequency (RF) circuitry comprising: one or more amplifiers; one or more impedance-tuning circuits coupled to one or more outputs of the one or more amplifiers; and an antenna tuner coupled between the antenna feed and one or more outputs of the one or more impedance-tuning circuits. The method also includes transmitting one or more signals via the RF circuitry when at least one parameter is configured at least in part based on at least one load characteristic at the antenna feed.

[0007] Some aspects provide a radio frequency (RF) transmitter. The RF transmitter includes: one or more amplifiers; one or more impedance tuning circuits coupled to one or more outputs of the one or more amplifiers; an antenna tuner coupled between an antenna feed and one or more outputs of the one or more impedance tuning circuits; one or more memories; and one or more processors coupled to the one or more memories, the one or more processors being configured to: configure the tuning index of at least one of the one or more impedance tuning circuits or one or more amplifiers at least partially based on at least one load characteristic at the antenna feed; and transmit one or more signals when the tuning index is configured at least partially based on at least one load characteristic at the antenna feed.

[0008] Other aspects provide: an apparatus operable to, configured to, or otherwise adapted to perform one or more of the foregoing methods and / or those methods described elsewhere herein; a non-transitory computer-readable medium comprising instructions which, when executed by a processor of the apparatus, cause the apparatus to perform the foregoing methods and those methods described elsewhere herein; a computer program product embodied on a computer-readable storage medium comprising: code for performing the foregoing methods and those methods described elsewhere herein; and / or an apparatus comprising components for performing the foregoing methods and those methods described elsewhere herein. By way of example, an apparatus may include a processing system, a device having a processing system, or a processing system cooperating via one or more networks.

[0009] To achieve the foregoing and related objectives, one or more aspects include the features fully described below and specifically pointed out in the claims. The following description and drawings illustrate certain exemplary features of these one or more aspects in detail. However, these features indicate only a few of the various ways in which the principles of the various aspects may be employed. Attached Figure Description

[0010] To gain a more detailed understanding of the foregoing features of this disclosure, a more specific description of the brief overview can be obtained by referring to some aspects illustrated in the accompanying drawings. However, it should be noted that the drawings illustrate only certain typical aspects of this disclosure and are therefore not to be considered as limiting its scope, as other equally valid aspects may be acknowledged in this description.

[0011] Figure 1 An example wireless communication system is shown.

[0012] Figure 2 An example wireless communication device is shown that communicates with another device.

[0013] Figure 3 An example RF transmitter with multi-stage impedance tuning is shown.

[0014] Figure 4 An example mapping is shown for the combination of the error vector magnitude and the tuning index of the antenna load and impedance tuning circuit.

[0015] Figure 5 An example mapping is shown for the power amplifier output power in combination with antenna load and tuning index.

[0016] Figure 6 An example AI architecture is shown that can be used for artificial intelligence (AI) enhanced wireless communication.

[0017] Figure 7An example artificial neural network is shown.

[0018] Figure 8 An example architecture for training an ML model to determine a multi-level tuning configuration is illustrated.

[0019] Figure 9 An example operation of wireless communication by the device is illustrated.

[0020] Figure 10 Examples are illustrated of communication devices that may include various components configured to perform the operations of the techniques disclosed herein.

[0021] For ease of understanding, the same reference numerals have been used where possible to denote common elements in the figures. It is conceivable that elements disclosed in one aspect may be usefully applied to other aspects without specific description. Detailed Implementation

[0022] This disclosure provides apparatus, methods, processing systems, and computer-readable media for multi-stage impedance tuning of radio frequency (RF) transmitters.

[0023] In some respects, an RF transmitter may use a power amplifier (PA) to amplify a signal for transmission via an antenna. For example, a PA can convert a low-power RF signal into a higher-power RF signal, and the output of the power amplifier can drive the antenna to transmit RF energy. The performance of the PA (e.g., power output, efficiency, etc.) can be configured using certain resonant circuits coupled to the output of the PA. In some cases, the resonant circuit may have an adjustable impedance, which can be adjusted to match the impedance of the load coupled to the PA as the output impedance of the PA changes over time (e.g., due to variations in RF carrier frequency, output power, compression, gain, etc.). In some cases, the impedance of the resonant circuit can be adjusted assuming that the load impedance at the output of the PA remains constant (e.g., 50 ohms). The load may represent certain RF circuitry and / or conditions, including, for example, an antenna switching module, an antenna tuner, the antenna itself, and any power reflections.

[0024] Technical challenges of RF transmitters include, for example, achieving efficient performance of the PA (Power Amplifier) ​​within the RF transmitter. For certain wireless communication devices (e.g., portable devices such as cellular phones or smartphones), the antenna load impedance and load phase (e.g., complex antenna impedance) can vary depending on various factors, such as the RF environment (e.g., RF reflections, interference, and / or noise), the RF carrier frequency, output power, etc. An antenna tuner can be coupled between the RF transmitter and the antenna, and the antenna tuner can be tuned to match the impedance of the RF transmitter to the complex antenna impedance, for example, to maximize power delivery to the antenna. As an example, the antenna tuner may include an impedance matching network tuned to match the complex antenna impedance at an antenna feed, which may be or includes an output terminal and / or transmit line coupling the antenna to the RF transmitter. The complex impedance seen at the PA's output can be a function of the losses between the PA and the antenna, as well as the complex reflection coefficient at the antenna feed. A high voltage standing wave ratio (VSWR) at the antenna feed affects the linearization of the power amplifier (PA) and thus influences certain performance metrics, including adjacent channel leakage ratio (ACLR), error vector magnitude (EVM), power output, etc. Consequently, the complex impedance seen at the PA's output can vary depending on factors such as the antenna tuner stage of the antenna tuner (e.g., the impedance of the matching network) and the RF environment (e.g., RF reflections, interference, and / or noise). Therefore, the antenna impedance load and load phase seen at the antenna can affect PA performance, for example, in terms of digital predistortion (DPD), power output, ACLR, EVM, and / or power output efficiency, especially when the impedance matching at the PA is configured based on a fixed load rather than a varying load.

[0025] The aspects described herein overcome the aforementioned technical problems by providing multi-stage impedance tuning for the RF transmitter. In some aspects, in the RF transmitter, the impedance tuning circuitry can be arranged between the output of an amplifier (e.g., PA) and the antenna tuner, and the impedance tuning circuitry, together with the antenna tuner, can be tuned based on the antenna impedance (e.g., a complex antenna impedance including the impedance load and the load phase), as described herein. Figure 3 Further described. In some respects, the tuning of an impedance tuning circuit can be characterized via a mapping (e.g., heatmaps and / or lookup tables (LUTs)) between a performance metric of the RF circuit (e.g., ACLR, EVM, power output efficiency, etc.) and a combination of antenna impedance and the tuning index (or tuning state) of the impedance tuning circuit, for example, as described herein. Figure 4 and Figure 5Further description. As an example, the tuning index may correspond to and / or represent the reactance of a reactive component (e.g., a variable capacitor) of the impedance tuning circuit and / or the bias current or voltage applied to the PA. The controller may determine the tuning of the impedance tuning circuit based on this mapping, which maps to a target performance metric and the current antenna impedance. In some cases, a machine learning (ML) model may be used to identify the tuning of the impedance tuning circuit given at least the current antenna impedance (and in some cases, the target performance metric), as described in this paper regarding... Figure 3 Further described. For example, an ML model can be efficiently trained to learn the mappings discussed above and find the tuning index of the impedance-tuned circuit. This multi-stage impedance tuning effectively allows the PA's output impedance to account for varying loads from the antenna and / or RF circuitry coupled to the PA's output, such as due to impedance matching applied at the antenna tuner and / or the RF environment (e.g., reflections from surrounding objects and / or internal circuitry).

[0026] The techniques for multi-stage impedance tuning described herein offer a variety of beneficial effects and / or advantages. Techniques for multi-stage impedance tuning can improve the performance of the power amplifier (PA) in an RF transmitter, for example, in terms of linearization, ACLR, EVM, power output efficiency, etc. In some cases, multi-stage impedance tuning can help restore PA linearity without requiring online DPD (e.g., recalibrating and training the DPD weights associated with the PA), which can consume significant time, processing resources, and / or power. Improvements in PA performance can be attributed to impedance tuning at the PA output, which takes into account varying loads on the antenna and / or RF circuitry coupled to the PA output. In some aspects, the mappings discussed herein can allow the RF transmitter to meet one or more target performance metrics, such as linearization, ACLR, EVM, power output efficiency, etc. In some aspects, the ML-based multi-stage tuning described herein can improve PA performance, for example, because the ML model is trained to identify settings of certain impedance tuning circuit parameters that can achieve one or more PA performance metrics. Example wireless communication system

[0027] Figure 1Example wireless communication system 100 is illustrated, in which aspects of this disclosure may be performed. For example, wireless communication system 100 may include a wireless wide area network (WWAN) and / or a wireless local area network (WLAN). WWAN may include new radio (NR) systems (e.g., fifth-generation (5G) NR networks), evolved universal terrestrial radio access (E-UTRA) systems (e.g., fourth-generation (4G) networks), universal mobile telecommunications systems (UMTS) (e.g., second-generation (2G) or third-generation (3G) networks), code division multiple access (CDMA) systems (e.g., 2G / 3G networks), any future WWAN systems, or any combination thereof. WLAN may include a wireless network configured to communicate according to Institute of Electrical and Electronics Engineers (IEEE) standards (such as one or more standards in the 802.11 standard). In some cases, wireless communication system 100 may include a device-to-device (D2D) communication network or a short-range communication system, such as Bluetooth communication and / or near-field communication (NFC).

[0028] like Figure 1 As illustrated, the wireless communication system 100 may include a first wireless device 102 that communicates with any of a variety of second wireless devices 104a to 104d (hereinafter referred to as "second wireless device 104") via any of a variety of radio access technologies (RATs), wherein a wireless device may refer to a wireless communication device. RATs may include, for example, WWAN communication (e.g., E-UTRA and / or 5G NR), WLAN communication (e.g., IEEE 802.11), vehicle-to-everything (V2X) communication, non-terrestrial network (NTN) communication, short-range communication (e.g., Bluetooth and / or NFC), etc.

[0029] The first wireless device 102 may include any of a variety of wireless communication devices, including user equipment (UE), base station, radio station, access point, customer premises equipment (CPE), etc. In some aspects, according to various aspects of this disclosure, the first wireless device 102 includes a multi-level tuning manager 106 configured in the RF transmitter to adjust the tuning index of an impedance tuning circuit arranged between the output of an amplifier and an antenna tuner.

[0030] The second wireless device 104 may include, for example, a base station 104a, a vehicle 104b, an access point (AP) 104c, and / or a user equipment (UE) 104d. Furthermore, the wireless communication system 100 may include terrestrial aspects (such as ground-based network entities (e.g., base station 104a and / or access point 104e)) and / or non-terrestrial aspects (such as spaceborne platforms and / or airborne platforms). These non-terrestrial aspects may include onboard network entities (e.g., one or more base stations) capable of communicating with other network elements (e.g., ground base stations) and / or user equipment.

[0031] Base station 104a may typically include: Node B, Enhanced Node B (eNB), Next Generation Enhanced Node B (ng-eNB), Next Generation Node B (gNB or gNodeB), access point, transceiver base station, radio base station, radio transceiver, transceiver functionality, transmit / receive point, and / or others. Base station 104a can provide communication coverage for a corresponding geographic coverage area, which may sometimes be referred to as a cell, and in some cases may overlap (e.g., a small cell may have a coverage area that overlaps with the coverage area of ​​a macro cell). For example, the base station may provide communication coverage for macro cells (covering a relatively large geographic area), pico cells (covering a relatively small geographic area, such as a stadium), femtocells (covering a relatively small geographic area (e.g., a home)), and / or other types of cells.

[0032] The first wireless device 102 and / or UE 104d may generally include: cellular phones, smartphones, Session Initiation Protocol (SIP) phones, laptops, personal digital assistants (PDAs), satellite radios, global positioning systems, multimedia devices, video devices, digital audio players, cameras, game consoles, tablet computers, smart devices, wearable devices, vehicles, electricity meters, air pumps, large or small kitchen appliances, healthcare devices, implants, sensors / actuators, displays, Internet of Things (IoT) devices, always-on (AON) devices, edge processing devices, or other similar devices. The UE may also be more generally referred to as a mobile device, wireless device, wireless communication device, radio station (STA), mobile station, subscriber station, mobile subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, remote device, access terminal, mobile terminal, wireless terminal, remote terminal, mobile phone, and other terms.

[0033] Figure 2 Example components of a first wireless device 102 are illustrated, which can be used to communicate with any of the second wireless devices in the second wireless device 104.

[0034] The first wireless device 102 may be or may include a chip, a system-on-a-chip (SoC), a system-in-package (SiP), a chipset, a package, or a device including one or more modems 210 (hereinafter “modem 210”). In some cases, modem 210 may include, for example, any of the following: a WWAN modem (e.g., a modem configured to communicate via E-UTRA 5G NR and / or any future WWAN communication standard), a WLAN modem (e.g., a modem configured to communicate via the IEEE 802.11 standard), a Bluetooth modem, an NTN modem, etc. In some aspects, the first wireless device 102 also includes one or more RF transceivers (hereinafter “RF transceiver 250”). In some examples, RF transceiver 250 may be referred to as an RF front end (RFFE). In some aspects, modem 210 also includes one or more processors, processing blocks, or processing elements (hereinafter “processor 212”) and one or more storage blocks or elements (hereinafter “memory 214”). In some cases, processor 212 may implement and / or include Figure 1 Multi-level tuning manager 106.

[0035] In some respects, processor 212 may process any of the protocol stack layers associated with radio access technology (RAT). For example, processor 212 may process any of the application layer, packet layer, WLAN protocol stack layer (e.g., link or media access control (MAC) layer) and / or WWAN protocol stack layer (e.g., radio resource control (RRC) layer, packet data convergence protocol (PDCP) layer, radio link control (RLC) layer, and MAC layer).

[0036] Modem 210 may typically be configured to implement the physical (PHY) layer. For example, modem 210 may be configured to modulate packets and output the modulated packets to RF transceiver 250 for transmission over a wireless medium. Modem 210 is similarly configured to receive modulated packets received by RF transceiver 250 and demodulate these packets to provide demodulated packets. In addition to modulators and demodulators, modem 210 may also include digital signal processing (DSP) circuitry, automatic gain control (AGC), decoders, decoders, multiplexers, and / or demultiplexers (not shown).

[0037] As an example, when in transmit mode, modem 210 may obtain data from a data source, such as an application processor. The data may be provided to a decoder, which encodes the data to provide coded bits. The coded bits may be mapped (e.g., using a selected modulation and decoding scheme) to points in a modulation constellation to provide modulated symbols. The modulated symbols may be mapped to, for example, a spatial stream or a space-time stream. The modulated symbols may be multiplexed, transformed via an inverse fast Fourier transform (IFFT) block, and subsequently provided to DSP circuitry for transmit windowing and filtering. The digital signal may be provided to digital-to-analog converter (DAC) 216. In some aspects involving beamforming, the modulated symbols in the corresponding spatial stream may be pre-decoded via a steering matrix before being provided to the IFFT block.

[0038] Modem 210 can be coupled to RF transceiver 250 via a transmit (TX) path 218 (also referred to as the transmit chain) for transmitting signals via one or more antennas 220 (hereinafter referred to as "antenna 220") and a receive (RX) path 222 (also referred to as the receive chain) for receiving signals via antenna 220. When TX path 218 and RX path 222 share antenna 220, these paths can be coupled to antenna 220 via interface 224, which can include any of a variety of suitable RF devices, such as antenna tuners, switches, duplexers, dual-signalers, multiplexers, etc. As an example, modem 210 can output digital in-phase (I) baseband signals and / or quadrature (Q) baseband signals representing corresponding symbols to DAC 216. In some examples, all or most of the components included in RF transceiver 250 are illustrated in a single chip, die, or package (such as an RFFE integrated circuit). For example, in some configurations, all components of the RF transceiver except antenna 220 are implemented on a single chip. In some other configurations, interface 224 or a portion thereof is also omitted from a single chip.

[0039] Receiving either an I-band or Q-band analog signal from DAC 216, TX path 218 may include a baseband filter (BBF) 226, a mixer 228 (which may include one or more mixers), and a power amplifier (PA) 230. BBF 226 filters the baseband signal received from DAC 216, and mixer 228 mixes the filtered baseband signal with a transmit local oscillator (LO) signal to convert the baseband signal to a different frequency (e.g., up-converting from baseband to radio frequency). In some aspects, the frequency conversion process produces a sum and difference frequency between the LO frequency and the frequency of the baseband signal. This sum and difference frequency is called a beat frequency. Some beat frequencies are in the RF range, such that the signal output from mixer 228 is typically an RF signal, which may be amplified by PA 230 before being transmitted via antenna 220. Antenna 220 may transmit RF signals, which may be received at the second wireless device 104. Although a mixer 228 is illustrated, several mixers can be used to upconvert the filtered baseband signal to one or more intermediate frequencies and then upconvert the intermediate frequency signal to the frequency used for transmission.

[0040] RX path 222 may include a low-noise amplifier (LNA) 232, a mixer 234 (which may include one or more mixers), and a baseband filter (BBF) 236. RF signals received via antenna 220 (e.g., from the second wireless device 104) may be amplified by LNA 232, and mixer 234 mixes the amplified RF signal with a received local oscillator (LO) signal to convert the RF signal to a baseband frequency (e.g., down-conversion). The baseband signal output from mixer 234 may be filtered by BBF 236 and then converted to a digital I or Q signal by analog-to-digital converter (ADC) 238 for digital signal processing. Modem 210 may receive the digital I or Q signals and further process these digital signals, for example, demodulating them into information.

[0041] Some transceivers may employ a frequency synthesizer with a voltage-controlled oscillator (VCO) to generate a stable, tunable LO frequency with a specific tuning range. Therefore, the transmit LO frequency can be generated by frequency synthesizer 240, which may be buffered or amplified by an amplifier (not shown) before mixing with the baseband signal in mixer 228. Similarly, the receive LO frequency can be generated by frequency synthesizer 240, which may be buffered or amplified by an amplifier (not shown) before mixing with the RF signal in mixer 234. Separate frequency synthesizers may be used for TX path 218 and RX path 222.

[0042] When in receive mode, modem 210 can acquire the digitally converted signal via ADC 238 and RX path 222. As an example, in modem 210, the digital signal can be provided to DSP circuitry configured to acquire the received signal, for example, by detecting the presence of the signal and estimating the initial timing and frequency offset. The DSP circuitry is also configured to digitally condition the digital signal, for example, using channel (narrowband) filtering, analog impairment conditioning (such as correcting I / Q imbalance), and applying digital gain to ultimately obtain a narrowband signal. The output of the DSP circuitry can be fed to AGC, which is configured to use information extracted from the digital signal (e.g., in one or more received training fields) to determine the appropriate gain. The output of the DSP circuitry can also be coupled to a demodulator configured to extract modulated symbols from the signal and, for example, calculate the log-likelihood ratio (LLR) for each bit location of each subcarrier in each spatial stream. The demodulator can be coupled to a decoder configured to process the LLR to provide decoded bits. The decoded bits from all spatial streams can be fed to a demultiplexer for demultiplexing. The demultiplexed bits can be descrambled and provided to the media access control layer (e.g., processor 212) for processing, evaluation, or interpretation.

[0043] Modem 210 and / or processor 212 can control the transmission of signals via TX path 218 and / or the reception of signals via RX path 222. In some aspects, modem 210 and / or processor 212 can be configured to perform various operations, such as those associated with any of the methods described herein. Modem 210 and / or processor 212 may include a microcontroller, microprocessor, application processor, baseband processor, MAC processor, artificial intelligence (AI) processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA) or other programmable logic device (PLD), discrete gate or transistor logic unit, discrete hardware component, or any combination thereof. Memory 214 may store data and program code (e.g., processor-readable instructions) for performing wireless communications as described herein. In some cases, memory 214 may be external to modem 210 and / or processor 212 and / or incorporated into the modem and / or processor (as illustrated).

[0044] In some respects, one or more ML models 242 may be stored in memory 214 and accessible by processor 212. In some cases, different ML models 242 with different characteristics may be stored in memory 214, and a particular ML model 242 may be selected based on its characteristics and / or application and the characteristics and / or conditions of the first wireless device 102 (e.g., power state, mobility state, battery reserve, temperature, etc.). For example, ML models 242 may have different inference data and output pairings (e.g., different types of inference data produce different types of outputs), and predictions (e.g., Figure 6 The different accuracy levels (e.g., 80%, 90%, or 95% accuracy) associated with the output (614), the different latency associated with generating the prediction (e.g., less than 10 milliseconds (ms), 100 ms, or 1 second processing time), the different ML model sizes (e.g., file size), the different coefficients or weights, etc.

[0045] Processor 212 can use ML model 242 based on input data (e.g., Figure 6 The inferred data 612) is used to generate output data (e.g., Figure 6 The output is 614), for example, as in this article about Figure 6 The inference host 604 is described herein. The ML model 242 can be used to perform any AI augmentation task among various AI augmentation tasks, such as those described herein. As an example, the ML model 242 can determine one or more tuning parameters to configure an impedance-tuned circuit coupled to the output of PA 230, as described herein. Figure 3 Further description. It should be noted that other input and / or output data may be used as a supplement to or alternative to the examples described herein.

[0046] In some respects, model server 260 can perform any of the various ML model lifecycle management (LCM) tasks for the first wireless device 102 and / or the second wireless device 104. Model server 260 can operate as a model training host (e.g., as per [reference to...]). Figure 6 (As discussed), and uses the training data to update the ML model 242. In some cases, the model server 260 can operate as a data source (e.g., as regarding...). Figure 6 (As discussed) to collect and host training data, inference data, and / or performance feedback associated with ML model 242. In some respects, model server 260 can host various types and / or versions of ML model 242 for download by first wireless device 102 and / or second wireless device 104.

[0047] In some cases, model server 260 may monitor and evaluate the performance of ML model 242 to trigger one or more LCM tasks. For example, model server 260 may determine whether to activate or deactivate the use of a specific ML model at first wireless device 102 and / or second wireless device 104, and model server 260 may provide such instructions to the respective first wireless device 102 and / or second wireless device 104. In some cases, model server 260 may determine whether to switch to a different ML model 242 used at first wireless device 102 and / or second wireless device 104, and model server 260 may provide such instructions to the respective first wireless device 102 and / or second wireless device 104. In yet other examples, model server 260 may also act as a central server for decentralized machine learning tasks such as federated learning, as discussed further herein.

[0048] Figure 2 An example transceiver design is illustrated. It will be understood that other transceiver designs or architectures can be applied in conjunction with various aspects of this disclosure. For example, while the examples discussed herein utilize I and Q signals (e.g., quadrature modulation), those skilled in the art will understand that transceiver components can be configured to utilize any other suitable modulation, such as polarity modulation. Furthermore, circuit blocks can be... Figure 2 The configurations shown are arranged differently, and / or can be implemented in addition to or in place of the depicted blocks. Figure 2 Other circuit blocks not shown.

[0049] Some aspects described in this paper can be implemented, at least in part, using some form of artificial intelligence (AI), such as the process of using a machine learning (ML) model to infer or predict output data based on input data. Example ML models may include mathematical representations of one or more relationships between various objects to provide outputs representing one or more predictions or inferences. Once an ML model has been trained, it can be deployed to process data that is wholly or partially similar to or related to the training data, and to provide outputs representing one or more predictions or inferences based on the input data.

[0050] ML is typically characterized by a learning type that generates a specific type of learning model that performs a particular type of task. For example, different types of machine learning include supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning.

[0051] Supervised learning algorithms typically model the relationships and dependencies between input features (e.g., feature vectors) and one or more target outputs. Supervised learning uses labeled training data, which consists of data including one or more inputs and the desired output. Supervised learning can be used to train models to perform tasks such as classification (where the goal is to predict discrete values) or regression (where the goal is to predict continuous values). Some example supervised learning algorithms include nearest neighbor, Naive Bayes, decision trees, linear regression, support vector machines (SVM), and artificial neural networks (ANN).

[0052] Unsupervised learning algorithms process unlabeled input data and train models that take the input and transform it into output to solve real-world problems. Examples of unsupervised learning tasks are clustering (where the model's output might be cluster labels), dimensionality reduction (where the model's output is an output feature vector with fewer features than the input feature vector), and outlier detection (where the model's output is a value indicating how the input differs from typical examples in the dataset). An example unsupervised learning algorithm is k-means.

[0053] Semi-supervised learning algorithms process datasets containing both labeled and unlabeled examples, where the number of unlabeled examples is typically much greater than the number of labeled examples. However, the goal of semi-supervised learning is to achieve the goals of supervised learning. Typically, a semi-supervised model involves a model trained to generate pseudo-labels for unlabeled data, which are then combined with labeled data to train a second classifier that leverages a larger volume of overall training data to improve task performance.

[0054] Reinforcement learning algorithms use observations gathered by an agent from its interactions with the environment to take actions that maximize reward or minimize risk. Reinforcement learning is a continuous and iterative process in which the agent learns from its experience with the environment until it has explored, for example, the entire range of possible states. An example type of reinforcement learning algorithm is adversarial networks. Reinforcement learning can be particularly beneficial when used to improve or attempt to optimize the behavior of models deployed in dynamically changing environments, such as wireless communication networks.

[0055] ML models can be deployed in one or more devices (e.g., network entities such as base stations and / or user equipment) to support various wired and / or wireless communication aspects of a communication system. For example, ML models can be trained to identify patterns and relationships in data corresponding to networks, devices, air interfaces, etc. ML models can improve operations associated with one or more aspects, such as transceiver circuitry control, frequency synchronization, timing synchronization, channel state estimation, channel equalization, channel state feedback, modulation, demodulation, device location, transceiver tuning, beamforming, signal decoding / decoding, network routing, load balancing, and energy saving (to name just a few). AI-enhanced transceiver circuitry control may include, for example, filter tuning, transmit power control, gain control (including automatic gain control), phase control, power management, etc.

[0056] The aspects described herein can be used to describe technical solutions for performing certain tasks and various technical problems by applying specific types of ML models, such as ANNs. However, it should be understood that other types of AI models can be used as supplements or alternatives to ANNs. ML models can be examples of AI models, and any suitable AI model can be used as a supplement or alternative to any ML model described herein. Therefore, unless explicitly stated otherwise, the topic of ML models is not necessarily intended to be limited to ANN solutions or machine learning. Furthermore, it should be understood that, unless specifically stated otherwise, terms such as “AI model,” “ML model,” “AI / ML model,” or “trained ML model” are intended to be used interchangeably. Example of multi-stage impedance tuning in an RF transmitter

[0057] Various aspects of this disclosure provide techniques for multi-stage impedance tuning in RF transmitters that can improve the performance of the PA in the RF transmitter, such as in linearization, ACLR, EVM, power output efficiency, etc.

[0058] Figure 3 An example RF transmitter 300 with multi-stage impedance tuning is illustrated. The RF transmitter 300 can be an example of a TX path, such as... Figure 2 The TX path 218 or aspects thereof. The RF transmitter 300 may include one or more amplifiers 302 (hereinafter referred to as "amplifier 302"), one or more impedance tuning circuits 304 (hereinafter referred to as "impedance tuning circuit 304"), an antenna tuner 308, an antenna 310, and a controller 312. In some cases, the RF transmitter 300 may also include RF circuitry 306 coupled between the impedance tuning circuit 304 and the antenna tuner 308.

[0059] Amplifier 302 may be or include a power amplifier (PA) 314 that converts a lower-power RF signal to a higher-power RF signal. In some cases, amplifier 302 may be or include a driver amplifier (DA) 316 coupled to PA 314. For example, DA 316 and PA 314 may form a cascaded amplifier configuration, wherein the output 318 of DA 316 is coupled to the input 320 of PA 314. DA 316 and PA 314 may be used to provide multi-stage amplification of the RF signal. In some aspects, each of DA 316 and PA 314 may have a power rail supplying power to the respective amplifier. Power supply voltage and / or power supply current may be fed into the power rails of the amplifier. In this example, one or more power supplies 322, 324 may be fed into one or more power rails of the respective amplifiers 314, 316. As discussed further herein, the current and / or voltage levels of power supplies 322, 324 may be used as tuning parameters for amplifier 302. Power supplies 332 and 324 may correspond to power supply currents, power supply voltages, reference currents, quiescent currents, reference voltages, bias voltages, and / or bias currents fed to one or more transistors (not shown) of DA 316 and / or PA 314. In some respects, power supplies 322 and 324 may be used as bias currents and / or bias voltages applied to one or more transistors of DA 316 and / or PA 314. For example, bias currents and / or voltages may be fed to the collectors of one or more transistors of DA 316 and / or PA 314 to tune the collector impedance of the transistors of DA 316 and / or PA 314.

[0060] Impedance tuning circuitry 304 is disposed between output terminal 326 of amplifier 302 and antenna tuner 308. In this example, impedance tuning circuitry 304 is coupled to output terminal 326 of amplifier 302 (e.g., one or more output terminals). In some aspects, any element of tuning circuitry 304 may be disposed wholly or partially on or on the same chip, die, or module as amplifier 302. In some aspects, any element of tuning circuitry 304 may be partially or wholly integrated with amplifier 302. In some aspects, tuning circuitry 304 may be disposed in or within the same package or module as amplifier 302 (e.g., partially on or on the die of PA 314, or separately from the die of PA 314). In some aspects, tuning circuitry 304 may be completely separate from the chip, die, or module on or in which amplifier 302 is disposed.

[0061] In some aspects, the impedance tuning circuit 304 can tune the output impedance of the amplifier 302 based on the load impedance of subsequent RF circuitry in the transmit chain. For example, the impedance tuning circuit 304 can be configured to adjust its impedance to balance the impedances of the amplifier 302 and the RF circuitry 306. In some aspects, the impedance tuning circuit 304 can operate as a first impedance matching circuit 328, which is configured to increase power transfer to the antenna 310 and / or reduce power reflected back to the amplifier 302. In some cases, the first impedance matching circuit 328 can be configured to reduce the amplifier output impedance 330 ( Z PA_Out ) and the RF circuit impedance 332 of RF circuit 306 ( Z RF Effective matching. The impedance of the first impedance matching circuit 328 may depend on the output power of the amplifier 302, the tuning stage of the antenna tuner 308, any power loss across the RF circuit 306 and / or the antenna tuner 308, and / or the complex reflection coefficient at the antenna feed 332, each of which may vary due to the transmitted signal power, carrier frequency, and / or RF environment (e.g., surrounding objects that cause RF reflections).

[0062] Impedance tuning circuit 304 may be or include one or more resonant circuits formed via one or more reactive components, including, for example, one or more capacitors and / or one or more inductors. In this example, impedance tuning circuit 304 includes a first resonant circuit 336 and / or a bypass capacitor 338 (e.g., a decoupling capacitor). The first resonant circuit 336 is formed via a first capacitor 340 coupled in series with a first inductor 342. The first resonant circuit 336 may be or include a notch filter, for example, tuned or configured to suppress certain harmonic distortions, as discussed further below. The first capacitor 340 may include a variable capacitor, including, for example, a parallel array of capacitors in a capacitor bank and / or a varactor diode. The first resonant circuit 336 may be coupled between signal path 344 and reference potential node 346. Signal path 344 may be formed between input node 348 and output node 350 of impedance tuning circuit 304. Bypass capacitor 338 may be coupled between signal path 344 and reference potential node 346. Bypass capacitor 338 may include a variable capacitor, such as a parallel capacitor array in a capacitor bank and / or a varactor diode. As discussed further herein, the tunable capacitance of the variable capacitor in impedance tuning circuit 304 may be used as a tuning parameter of impedance tuning circuit 304.

[0063] In some respects, the impedance tuning circuit 304 may be formed using certain filters that suppress noise and / or distortion (such as certain harmonic distortions) of the RF transmitter 300. For example, the impedance tuning circuit 304 may be or include a harmonic notch filter tuned to suppress the second harmonic (e.g., 2fo) and / or the third harmonic (e.g., 3fo) of the RF carrier frequency. In this example, the first resonant circuit 336 may be configured to suppress the second harmonic of a specific RF carrier frequency. It should be noted that, as a complement or alternative to the first resonant circuit 336, the impedance tuning circuit 304 may include other filters and / or resonant circuits, such as low-pass filters, notch filters, harmonic notches, band-pass filters, etc.

[0064] RF circuit 306 may be or include an antenna switching module that selectively couples the output 326 of amplifier 302 to one or more antennas, such as antenna 310. In some cases, RF circuit 306 may be used to selectively couple the amplifier to one of multiple antennas (e.g., multiple low-frequency antennas, multiple mid-frequency antennas, and multiple high-frequency antennas). Therefore, RF circuit 306 may include a set of switches coupled to multiple antennas (not shown). In some aspects, RF circuit 306 may have impedances, including resistance and reactance (e.g., capacitance), which may affect the load impedance seen at the output of amplifier 302.

[0065] Antenna tuner 308 is coupled between amplifier 302 and antenna 310. In some cases, antenna tuner 308 is coupled between antenna feed 332 and one or more outputs 352 of impedance tuning circuit 304. In some cases, RF circuitry 306 may be arranged between amplifier 302 and antenna tuner 308. For example, antenna tuner 308 may be coupled to output 354 of RF circuitry 306.

[0066] Antenna tuner 308 includes a second impedance matching circuit 356 configured to adjust its impedance to balance the impedance of antenna 310 and the transmit chain circuit coupled to antenna tuner 308. As an example, the second impedance matching circuit 356 may be or include a variable capacitor (not shown) comprising a parallel array of capacitors in a capacitor bank and / or a varactor diode. In some aspects, the second impedance matching circuit 356 may be configured to increase power transfer to antenna 310 and / or reduce power reflected between antenna 310 and the transmit chain circuit. In some cases, the second impedance matching circuit 356 may be configured to reduce the transmit chain impedance 358 ( Z Tx ) and the antenna impedance 360 ​​of the antenna 310 coupled to the antenna tuner 308. Z AEffective matching is achieved. The transmit chain impedance 358 is the impedance of the transmit chain circuit coupled to the antenna tuner 308 (e.g., a cascade of transmit chain circuits including amplifier 302, impedance tuning circuit 304, and RF circuit 306). The impedance of the second impedance matching circuit 356 may depend on the transmit chain impedance 358 and the antenna impedance 360, each of which may vary due to the transmitted signal power and / or the RF carrier frequency.

[0067] Controller 312 controls multi-stage impedance tuning at impedance tuning circuit 304 and / or antenna tuner 308. In some aspects, controller 312 may include one or more processors 362 (hereinafter referred to as "processor 362") coupled to one or more memories 364 (hereinafter referred to as "memory 364"). Processor 362 may be an example of processor 212, and memory 364 may be... Figure 2 Example of memory 214.

[0068] The controller 312 may obtain feedback at the output 326 of the antenna feed 332 and / or amplifier 302, and the controller 312 may configure the impedance at the first impedance matching circuit 328 and / or the second impedance matching circuit 356 based on this feedback. This feedback may be or include an indication of the impedance load and / or load phase (e.g., antenna impedance 360) at the antenna feed 332. In some aspects, this feedback may be or include an indication of one or more performance metrics associated with the RF transmitter 300 (or any component thereof).

[0069] The controller 312 can be configured with a second impedance matching circuit 356 to match the impedance load and / or load phase observed at the antenna feed 332. The controller 312 can also be configured based on one or more performance metrics of the RF transmitter 300, the antenna impedance 360, etc. Z AThe impedance tuning circuit 304 is configured by a relationship or mapping between the tuning index of the RF transmitter 300 and / or the impedance tuning circuit 304. Performance metrics may include one or more of the following: operating temperature of the RF transmitter 300 (or any of its components, such as amplifier 302); output frequency of the RF transmitter 300 (e.g., characterized by a power spectral distribution associated with the output signal at output 302 and / or antenna feed 332); voltage standing wave ratio (VSWR) of the RF transmitter 300; error vector magnitude (EVM) of the RF transmitter 300; adjacent channel leakage ratio (ACLR) of the RF transmitter 300; peak-to-average power ratio (PAPR) of the RF transmitter 300; output power level of the RF transmitter 300; and power-added efficiency (PAE) of amplifier 302. The controller 312 can access a mapping of one or more performance metrics of the RF transmitter 300 (e.g., EVM, ACLR, power efficiency, etc.) to a combination of antenna loads (e.g., impedance loads and phase) at the antenna feed 332 and the tuning index of the impedance tuning circuit 304, for example, as described herein. Figure 4 and Figure 5 As described.

[0070] In some aspects, controller 312 may monitor performance metrics associated with RF transmitter 300, and controller 312 may adjust the impedance at first impedance matching circuit 328 and / or second impedance matching circuit 356 in response to changes in performance metrics. In some aspects, controller 312 may monitor performance metrics to generate a mapping between one or more performance metrics of RF transmitter 300 (e.g., EVM, ACLR, power efficiency, etc.) and a combination of antenna load (e.g., impedance load and phase) at antenna feed 332 and tuning index of impedance tuning circuit 304. As an example, controller 312 may monitor certain performance metrics via a feedback path between the transmit and receive chains (e.g., a feedback path for DPD calibration).

[0071] In some cases, controller 312 may search in the map for a tuning index corresponding to the current antenna impedance (e.g., impedance load and load phase) and the target performance metric (e.g., low ACLR, high PAE, low PAPR, and / or low EVM). The tuning index may correspond to one or more tuning parameters of at least one of amplifier 302 or impedance tuning circuit 304. For example, tuning parameters may include reference current, quiescent current, or bias current of power supplies 322, 324; reactance of at least one reactive component of impedance tuning circuit 304 (e.g., capacitance value of a variable capacitor); quiescent and / or bias current and / or voltage applied to the transistors of PA 314 and / or DA 316; and / or frequency variations associated with impedance tuning circuit 304.

[0072] In some respects, tuning parameters (e.g., bias current and / or capacitance) can be characterized for each complex impedance seen at the output 326 of amplifier 302. The impedance seen at the output 326 of amplifier can be monitored, and tuning parameters can be adjusted in response to the monitored impedance to deliver target performance metrics, which may include a combination of output power, ACLR, EVM, and / or power efficiency.

[0073] In some aspects, controller 312 may use one or more ML models 366 (hereinafter referred to as "ML model 366") to identify the tuning index of impedance tuning circuit 304, which may be stored in memory 364. ML model 366 may be an example of ML model 242. Controller 312 may provide input data to ML model 366, including antenna impedance, for example, as observed at antenna feed 332. In some aspects, the input data may include the tuning index and / or the current state of antenna impedance. In some aspects, the input data may include antenna impedance and target performance metrics of the RF transmitter, and the ML model may identify the tuning index based on antenna impedance and target performance metrics. Output data may include a tuning index predicted to achieve certain performance metrics in the current state of antenna impedance. Controller 312 may obtain output data including the tuning index of impedance tuning circuit from ML model 366. As an example, ML model 366 may be trained to identify the tuning index based on the mapping discussed above. Generally speaking, the ML model 366 can effectively determine the performance of the RF transmitter under various antenna loads and tuning indices, for example, in the form of the mapping discussed above. The controller 312 can use the ML model 366 to predict a specific tuning index that can achieve the target performance metric, given the current state of the antenna impedance as observed at the antenna feed 332 (e.g., provided to the controller 312 in the form of feedback).

[0074] Figure 4 Examples are given in this article regarding Figure 3 The described example mapping 400 is a combination of the EVM of the RF transmitter and the antenna impedance and tuning index of the impedance tuning circuit. Mapping 400 is shown as a heatmap of the EVM values ​​of the RF transmitter, where the y-axis represents different impedance loads and load phases, and the x-axis represents the impedance tuning circuit (e.g., ...). Figure 3Different tuning indices are used for the impedance tuning circuit 304. Each tuning index may correspond to a different combination of tuning parameters used to configure the impedance tuning circuit 304 and / or amplifier 302, such as the capacitance values ​​of power supplies 322, 324 and / or variable capacitors (e.g., first capacitor 340 and bypass capacitor 338). Assuming the current antenna impedance is observed at the load and load phase along horizontal line 402, controller 312 may search for (or identify using an ML model) tuning indices that achieve the lowest EVM for the corresponding antenna impedance, for example, at position 404. Furthermore, position 404 may allow for a relatively low EVM over a wide range of antenna impedances. Therefore, mapping 400 can be used to identify tuning indices that achieve the target performance metric over the range of antenna impedances. As the antenna impedance changes over time (e.g., along vertical line 406), controller 312 may repeatedly search for tuning indices that achieve the lowest EVM at the current antenna impedance. In some cases, Map 400 can be represented as a lookup table (or data structure) of a combination of EVM, antenna impedance, and tuning index. In some cases, an ML model (e.g., ML model 366) can be trained to identify the tuning index in Map 400 given the current antenna impedance and / or target EVM.

[0075] In some respects, an ML model can be trained to learn Map 400 or aspects thereof, and the ML model can identify the tuning index given the current antenna impedance and / or target EVM based on knowledge representing Map 400 (e.g., weights). In other words, Map 400 can be an example of the behavior of an RF transmitter, which can be trained to learn in the form of relationships between various combinations of antenna impedance, tuning index, and one or more performance metrics.

[0076] Figure 5 Example mapping 500 illustrates a combination of PA output power, antenna load, and tuning index. Mapping 500 can be used to identify, as described in this article... Figure 4 The tuning index of the described impedance tuning circuit. In some cases, controller 312 can search for (or identify using an ML model) the tuning index that maximizes the output power of the PA based on the current antenna impedance. In other cases, controller 312 can search for (or identify using an ML model) the tuning index that achieves a specific output power of the PA based on the current antenna impedance.

[0077] In some respects, controller 312 can search for tuning indices that achieve or are predicted to achieve multiple target performance metrics (e.g., EVM and output power) given the current state of the antenna impedance. For example, controller 312 can use mappings 400 and 500 to identify tuning indices that reduce EVM and increase output power given the current antenna impedance.

[0078] It should be noted that Figure 4 and Figure 5 The illustrated mappings are examples of mappings for certain performance metrics. Other mappings, such as those for ACLR, PAE, PAPR, linearization, etc., can be used as supplements or alternatives to those illustrated mappings. Example of artificial intelligence for wireless communication

[0079] Figure 6 This is a diagram illustrating an example AI architecture 600 that can be used for AI-enhanced wireless communication. As illustrated, architecture 600 includes multiple logical entities, such as a model training host 602, a model inference host 604, a data source 606, and an agent 608. The AI ​​architecture can be used in any of the various use cases of wireless communication, such as the multi-stage impedance tuning described herein.

[0080] The model inference host 604 in architecture 600 is configured to run an ML model based on inference data 612 provided by data source 606. Inference data 612 may be or include antenna impedance (e.g., impedance load and / or load phase) and / or target performance metrics. In some cases, antenna impedance may be or include multiple impedances measured, for example, within a time window. In other cases, antenna impedance may be or include peak values, minimum (lowest) values, average values, median values, etc., within a time window. The model inference host 604 may generate an output 614 (e.g., a prediction or inference, such as discrete or continuous values) based on the inference data 612, and then provide it as input to agent 608.

[0081] In some respects, agent 608 may be or include a processor, modem, or controller of an RF transmitter. For example, agent 608 may be... Figure 3 The controller 312. Additionally, the type of agent 608 may also depend on the type of task performed by the model inference host 604, the type of inference data 612 provided to the model inference host 604, and / or the type of output 614 generated by the model inference host 604.

[0082] After agent 608 receives output 614 from model inference host 604, agent 608 can determine whether to take action based on the output. For example, if agent 608 is a modem and the output from model inference host 604 identifies the tuning parameters (e.g., tuning index) of impedance tuning circuit 304 and / or amplifier 302, agent 608 can determine whether to configure the impedance tuning circuit based on output 614. If agent 608 determines to take action based on output 614, agent 608 can instruct at least one action subject 610 to take the action, for example, in the form of a control signal configured to tune the impedance tuning circuit.

[0083] Data source 606 can be configured to collect data that can be used as training data 616 for training an ML model or as inference data 612 for feeding the ML model's inference operations. Training data 616 may be or include antenna impedance and the corresponding expected performance metrics and / or tuning indices, as described herein. Figure 8 Further described. Specifically, data source 606 may collect data from any of a variety of entities, devices, or components that may include action agent 610 (e.g., an RF transmitter) and provide the collected data to model training host 602 for ML model training. For example, a modem may collect performance feedback associated with multi-stage impedance tuning (e.g., one or more performance metrics including VSWR, EVM, ACLR, PAPR, PAE, etc.) and provide such feedback to data source 606, wherein the performance feedback may be used by model training host 602 to monitor and / or evaluate ML model performance, such as whether the output 614 provided to agent 608 is accurate. In some examples, if the output 614 provided to agent 608 is inaccurate (or its accuracy is below an accuracy threshold), model training host 602 may determine, for example, to modify or retrain the ML model used by model inference host 604 via ML model deployment / update.

[0084] In some respects, model training host 602 may be deployed at the same or a different entity as the entity that deploys model inference host 604, or together with that entity. For example, to offload model training processing that might affect the performance of model inference host 604, model training host 602 may be deployed at a model server, as described in this document. Figure 2 As described. Furthermore, in some cases, training and / or inference can be distributed among devices in a decentralized or federated manner. Example Artificial Intelligence Model

[0085] Figure 7 This is an exemplary block diagram of an example artificial neural network (ANN) 700.

[0086] ANN 700 may receive input data 706, which may include one or more bits of data 702, preprocessed data (optionally) output from preprocessor 704, or some combination thereof. Here, data 702 may include training data, validation data, application-related data, etc., depending on the development and / or deployment phase of ANN 700. In some other embodiments, preprocessor 704 may be included within ANN 700. Preprocessor 704 may, for example, process all or part of data 702, which may result in some data in data 702 being changed, replaced, deleted, etc. In some embodiments, preprocessor 704 may add additional data to data 702.

[0087] ANN 700 includes at least one first layer 708 of artificial neurons 710 (e.g., perceptrons) to process input data 706 and provide the resulting first layer output data to at least a portion of at least one second layer 714 via edge 712. The second layer 714 processes the data received via edge 712 and provides the second layer output data to at least a portion of at least one third layer 718 via edge 716. The third layer 718 processes the data received via edge 716 and provides the third layer output data to at least a portion of a final layer 722 comprising one or more neurons via edge 720 to provide output data 724. All or part of the output data 724 may be further processed in some way by (optionally) a post-processor 726. Thus, in some examples, ANN 700 may provide output data 728 based on output data 724, post-processed data output from post-processor 726, or some combination thereof. In some other embodiments, post-processor 726 may be included within ANN 700. Postprocessor 726 may process all or part of output data 724, which may result in output data 728 being at least partially different from output data 724, for example, due to data being altered, replaced, deleted, etc. In some embodiments, postprocessor 726 may be configured to add additional data to output data 724. In this example, the second layer 714 and the third layer 718 represent intermediate or hidden layers that may be arranged in a layered or other similar structural manner. Although not explicitly shown, one or more additional intermediate layers may exist between the second layer 714 and the third layer 718.

[0088] The structure and training of the artificial neurons 710 in each layer can be tailored to the specific requirements of the application. Within a given layer of an ANN, some or all of the neurons can be configured to process the information provided to that layer and output corresponding transformed information from that layer. For example, the transformed information from a layer can represent a weighted sum of input information associated with a nonlinear activation function or another activation function used to “activate” the artificial neurons in the next layer, or otherwise based on that nonlinear activation function or the other activation function used to “activate” the artificial neurons in the next layer. Artificial neurons in such layers can be activated by or in response to weights and biases that can be adjusted during the training process. The weights of various artificial neurons can act as parameters controlling the connection strength between layers or between artificial neurons, while the biases can act as parameters controlling the connection direction between layers or between artificial neurons. Activation functions can select or determine whether an artificial neuron sends its output to the next layer in response to the data received by the artificial neuron. Different activation functions can be used to model different types of nonlinear relationships. By introducing nonlinearity into the ML model, activation functions allow the ML model to “learn” the input data (e.g., Figure 6 Complex patterns and relationships in (606). Some non-exhaustive example activation functions include linear functions, binary step functions, sigmoid, hyperbolic tangent (tanh), rectified linear unit (ReLU) and its variants, exponential linear unit (ELU), Swish, Softmax, etc.

[0089] Design tools (such as computer applications, programs, etc.) can be used to select the appropriate structure and number of layers for the ANN 700, as well as the number of artificial neurons in each layer, and to select activation functions, loss functions, training procedures, etc. Once the initial model is designed, it can be trained using training data. Training data may include one or more datasets within which the ANN 700 can detect, determine, identify, or discover patterns. Training data can represent various types of information, including written, visual, audio, environmental context, operational attributes, etc. During training, the parameters of the artificial neurons 710 can be changed, such as minimizing or otherwise reducing the loss function or cost function. The training process can be repeated multiple times to fine-tune the ANN 700 with each iteration.

[0090] Various ANN model architectures are available for consideration. For example, in a feedforward ANN architecture, each artificial neuron 710 in a layer receives information from the previous layer and similarly generates information for the next layer. In a convolutional ANN architecture, some layers can be organized as filters that extract features from data (e.g., training data and / or input data). In a recursive ANN architecture, some layers may have connections that allow data to be processed across time, such as for processing information with temporal structure (e.g., time series data prediction).

[0091] In the autoencoder ANN architecture, compact representations of data can be processed, and models can be trained to make predictions or potentially reconstruct the original data from a reduced set of features. The autoencoder ANN architecture can be used for tasks related to dimensionality reduction and data compression.

[0092] Generative adversarial network (GAN) architectures can include generator ANNs and discriminator ANNs trained to compete against each other. GANs are ANN architectures that can be used for tasks related to generating synthetic data or improving the performance of other models.

[0093] Transformer ANN structures utilize an attention mechanism that enables the model to process input sequences in a parallel and efficient manner. The attention mechanism allows the model to focus on different parts of the input sequence at different times. This attention mechanism can be implemented using a series of layers called attention layers to compute, compute, determine, or select a weighted sum of input features based on the similarity between different elements of the input sequence. A transformer ANN structure may include a series of feedforward ANN layers that "learn" the non-linear relationship between the input and output sequences. The output of the transformer ANN structure can be obtained by applying a linear transformation to the output of the final attention layer. Transformer ANN structures are particularly useful for tasks involving sequence modeling or other similar processing.

[0094] Another example type of ANN structure is a model with one or more invertible layers. This type of model can be inverted or "unfolded" to reveal the input data used to generate the output of the layers.

[0095] Other examples of ANN model architectures include fully connected neural networks (FCNN) and long short-term memory (LSTM) networks.

[0096] ANN 700 or other ML models can be implemented in various types of processing circuits, as well as their memory and applicable instructions, for example, as described in this paper. Figure 2 , Figure 3 and Figure 6 As described. For example, the model can be implemented using general-purpose hardware circuitry such as one or more central processing units (CPUs) and one or more graphics processing units (GPUs). One or more ML accelerators, such as tensor processing units (TPUs), embedded neural processing units (eNPUs) or other dedicated processors, and / or field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), etc., can also be used to develop ANN models. Various programming tools can be used to develop ANN models. Various aspects of artificial intelligence model training

[0097] There are deployable ML models (such as...) Figure 7 The various model training techniques and processes used at some point before or after the ANN 700.

[0098] As part of the model development process, information may be collected or otherwise created in a suitable training data format for training the ML model accordingly. For example, training data may be collected or otherwise created relating to data on received / transmitted signal strength, interference, and resource usage, as well as any other relevant data that can be used to train the model to solve one or more problems or challenges in a communication system. In some instances, all or part of the training data may originate from one or more user equipment (UEs), one or more network entities, or one or more other devices in a wireless communication system. In some cases, all or part of the training data may be aggregated from multiple sources, such as one or more UEs, one or more network entities, the Internet, etc. For example, wireless network architectures (such as self-organizing networks (SONs) or mobile-driven test (MDT) networks) may be adapted to support data collection for ML model applications. In another example, training data may be generated or collected online, offline, or both online and offline by UEs, network entities, or other devices, and all or part of such training data may be transmitted or shared (real-time or near real-time), such as through store-and-forward functions. Offline training may refer to, for example, creating and using a static training dataset in a batch manner, while online training may refer to collecting and using training data in real-time or near real-time. For example, online or offline training can be used to train and / or fine-tune the ML model at a network device (e.g., a UE). For offline training, data collection and training can occur offline at the network side (e.g., at a base station or other network entity) or at the UE side. For online training, training of the ML model at the UE side can be performed locally at the UE or by a server device (e.g., a server hosted by the UE vendor) based on data provided from the UE to the server device, in a real-time or near real-time manner.

[0099] In some instances, all or part of the training data may be shared within the wireless communication system, or even shared (or obtained from outside) the wireless communication system.

[0100] Once the ML model has been trained on the training data, its performance can be evaluated. In some scenarios, evaluation / validation tests can be conducted using a validation dataset, which may include data not present in the training data, to compare the model's performance against a baseline or other benchmark information. If the model's performance is deemed unsatisfactory, fine-tuning the model may be beneficial, for example, by changing its architecture, retraining it on the data, or using different optimization techniques. Once the model's performance is deemed satisfactory, it can be deployed accordingly. In some instances, the model may be updated in some way, such as by changing or replacing all or part of the model, or by undergoing further training, to name just a few examples.

[0101] As for use in ANNs (such as Figure 7 As part of the training process of an ANN (Advanced Neural Network 700), parameters affecting the operation of artificial neurons and layers can be tuned. For example, backpropagation can be used to train an ANN by iteratively adjusting the weights and / or biases of certain artificial neurons associated with the error between the model's predicted output and the expected output, which may be known or otherwise considered acceptable. Backpropagation may include forward propagation, loss function, backward propagation, and parameter updates that can be performed during training iterations. This process can be repeated a certain number of times for each training dataset until the weights of the artificial neurons / layers are properly tuned.

[0102] Backpropagation, associated with the loss function, measures how well a model can predict the desired output for a given input. Optimization algorithms can be used during training to adjust weights and / or biases to reduce or minimize the loss function, which improves model performance. Various optimization algorithms exist that can be used with backpropagation or other training techniques. Some initial examples include gradient descent-based and stochastic gradient descent-based optimization algorithms. Stochastic gradient descent (or ascent) can be used to adjust weights / biases to minimize or otherwise reduce the loss function. Mini-batch gradient descent, a variant of gradient descent, involves updating weights / biases using mini-batch training data instead of the entire dataset. Momentum techniques can accelerate the optimization process by adding momentum terms to update or otherwise influence certain weights / biases.

[0103] Adaptive learning rate techniques adjust the learning rate of an optimization algorithm that is associated with one or more characteristics of the training data. Batch normalization techniques can be used to normalize the input to a model in order to stabilize the training process and potentially improve the model's performance.

[0104] The "drop-out" technique can be used to randomly discard some of the artificial neurons from the model during the training process, for example, to reduce overfitting and potentially improve the model's generalization.

[0105] The “early stopping” technique can be used to stop an ongoing training process early, such as when the performance of a model using a validation dataset begins to degrade.

[0106] Another example technique includes data augmentation, which generates additional training data by applying transformations to all or part of the training information.

[0107] Transfer learning techniques can be used, which involve using a pre-trained model as a starting point for training a new model. This can be useful when training data is limited or when there are multiple tasks that are related to each other.

[0108] Multi-task learning techniques can be used, which involve training a model to perform multiple tasks simultaneously to potentially improve the model's performance on one or more of these tasks. In some instances, hyperparameters can be input and applied during the training process.

[0109] Another example technique that can be useful for ML models is some form of "pruning." Pruning techniques, which can be performed during the training process or after the model has been trained, involve removing unnecessary (e.g., because they have no effect on the output), less necessary (e.g., because their effect on the output is negligible), or potentially redundant features from the model. In some instances, pruning techniques can reduce the complexity of the model or improve its efficiency without compromising its expected performance.

[0110] Pruning techniques can be particularly useful in the context of wireless communication, where available resources, such as power and bandwidth, may be limited. Some example pruning techniques include weight pruning, neuron pruning, layer pruning, structural pruning, and dynamic pruning. Pruning techniques can, for example, reduce the amount of data corresponding to a model that may need to be transmitted or stored.

[0111] Weight pruning techniques may involve removing some weights from a model. Neuron pruning techniques may involve removing some neurons from a model. Layer pruning techniques may involve removing some layers from a model. Structural pruning techniques may involve removing some connections between neurons in a model. Dynamic pruning techniques may involve adapting a pruning strategy for a model to one or more characteristics associated with the data or environment. For example, in some wireless communication devices, dynamic pruning techniques may be more aggressive in pruning models used in low-power or low-bandwidth environments and less aggressive in pruning models used in high-power or high-bandwidth environments. In some aspects, pruning techniques may also be applied to training data, such as to remove outliers. In some specific implementations, preprocessing techniques on all or part of the training dataset can improve model performance or facilitate faster model convergence. For example, training data may be preprocessed to alter or remove unnecessary, irrelevant, incorrect, or otherwise identifiable data. Such preprocessing of training data can, for example, lead to a reduction in potential overfitting or otherwise improve the performance of the trained model.

[0112] One or more of the example training techniques presented above can be used as part of the training process. As mentioned above, some example training processes that can be used to train ML models include supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning techniques.

[0113] Decentralized, distributed, or shared learning, such as federated learning, enables training on data distributed across multiple devices or organizations without the need for centralized data or training. Federated learning can be particularly useful in data-sensitive or privacy-constrained situations, or when centralized data is impractical, inefficient, or expensive. For example, in the context of wireless communications, federated learning can be used to improve performance by allowing ML models to be trained on data collected from a wide range of devices and environments. For instance, ML models can be trained on data collected from a large number of wireless devices in a network, such as distributed wireless communication nodes, smartphones, or Internet of Things (IoT) devices, to improve network performance and efficiency. With federated learning, a user equipment (UE) or other device can receive a full or partial copy of the model and perform local training on such a copy of the model using locally available training data. Such devices can provide updated information about the locally trained model (e.g., trainable parameter gradients) to one or more other devices, such as network entities or servers, where updates from other similar devices (such as other UEs) can be aggregated and used to provide updates to shared models, etc. The federated learning process can be iteratively repeated until all or part of the model achieves a satisfactory level of performance. Federated learning enables devices to protect the privacy and security of local data while supporting full or partial collaboration on training and updating shared models. Aspects of training a machine learning model for multi-level tuning

[0114] Figure 8 An example architecture 800 is illustrated for training an ML model to determine a multi-level tuning configuration of an RF transmitter. Architecture 800 can be trained by a model training host (e.g., Figure 6 This is implemented using a model training host 602. In some cases, the model training host may be or include a modem 210, a processor 212, and / or a controller 312. In some aspects, the model training host may be or include a model server 260, which can collect training data from one or more wireless communication devices (e.g., the first wireless device 102).

[0115] The model training host obtains training data 802, which includes training input data 804 and / or corresponding labels 806 used for training input data 804. Training input data 804 may include samples of antenna impedance (e.g., representing the current antenna impedance fed back to controller 312). In some cases, the sample antenna impedance may be or include multiple impedances, for example, measured (or simulated) within a time window. The sample antenna impedance may include a time series of antenna impedances. In some cases, the sample antenna impedance may be or include peak (highest value), minimum (lowest value), average, median, etc., within a time window. In some cases, training input data 804 may include a target performance metric to be achieved by the tuning index. In some cases, training input data 804 may include one or more performance maps (e.g., maps 400, 500) for searching the tuning index given an antenna impedance. Training input data 804 may be simulated (e.g., computer-generated) and / or collected from performance characterizations (e.g., measurements) of RF transmitters and / or multiple RF transmitters, for example, under various operating conditions as further discussed herein.

[0116] The model training host can use labels 806 to evaluate the performance of ML model 808 and tune ML model 808 (e.g., the weights of ANN 700), as further described herein. Each label in label 806 may be associated with at least one antenna impedance in the antenna impedances of the training input data 804. In some cases, each label in label 806 may include the expected tuning index of the corresponding antenna impedance. In other cases, each label in label 806 may include a performance metric of the expected tuning index (e.g., EVM, output power, etc.).

[0117] The model training host provides training input data 804 to the ML model 808. In some respects, the ML model 808 may include a neural network. The ML model 808 may be what this paper describes. Figure 2 , Figure 3 , Figure 6 and Figure 7 An example of the described ML model. ML model 808 provides output data 810, which may include an indication of the tuning index of impedance tuning circuit 304.

[0118] The model training host can evaluate the performance of the ML model 808 and, for example, determine whether to update the ML model 808 based on the performance achieved using the predicted tuning exponent. The model training host can evaluate the quality and / or accuracy of the output data 810. In some cases, the model training host can determine whether the output data 810 matches the corresponding label 806 of the training input data 804. For example, the model training host can determine whether the predicted tuning exponent output by the ML model 808 matches the expected tuning exponent of the corresponding label 806. In some cases, performance evaluation can determine whether the predicted tuning exponent achieves the performance metric associated with the expected tuning exponent of the corresponding label 806.

[0119] In some aspects, the model training host may use a cost or loss function 812 (hereinafter referred to as "loss function 812") to evaluate the performance of the ML model 808. Loss function 812 may be or include a comparison between the expected tuning index corresponding to label 806 and the predicted tuning index output by the ML model 808. In some cases, loss function 812 may be or include a comparison between the expected performance corresponding to the tuning index of label 806 and the actual performance of the RF transmitter configured with the predicted tuning index of output data 810. In some aspects, loss function 812 may be or include the difference between the actual performance of the RF transmitter configured with the predicted tuning index and the expected performance corresponding to label 806, for example, as the mean squared error between the corresponding performance measures. Loss function 812 may provide a loss value or score 814 based on the comparison between output data 810 and label 806.

[0120] The model training host can provide a loss score 814 to an optimizer 816, which determines one or more updated weights 818 for the ML model 808. The optimizer 816 can adjust the ML model 808 (e.g., any weights in the layers of a neural network) to reduce the loss score 814 associated with the ML model 808. In some respects, the optimizer 816 can perform backpropagation to determine the updated weights 818. The model training host can continue to provide training input data 804 to the ML model 808 and adjust the ML model 808 using the weights 818 until the loss score 814 of the ML model 808 meets a threshold and / or reaches a minimum. The model training host can perform online training of the ML model 808 or train the ML model 808 using one or more batches of training data 802. In some respects, the optimizer 816 may be or include a root mean square propagation (RMSprop) optimizer, a gradient descent optimizer (e.g., stochastic gradient descent (SGD)), a momentum optimizer, an Adam optimizer, etc., to minimize the loss fraction 814 associated with ML-based multi-stage impedance tuning.

[0121] In some respects, the model training host can train the ML model 808 to meet certain criteria associated with multi-stage impedance tuning. As an example, the model training host can train the ML model 808 to identify the tuning index that achieves certain performance metrics, such as minimum EVM, minimum ACLR, maximum PAE, maximum output power, minimum PAPR, etc.

[0122] In some respects, the model training host can train multiple ML models to perform multi-level impedance tuning. The ML models can be trained or configured with different model performance characteristics, different operating environments (e.g., RF environments), and / or different input-output schemes (e.g., different input data and different output data). For example, ML models can be trained to predict the tuning index to meet different accuracy levels (e.g., 70%, 80%, or 99% accuracy) and / or different latency (e.g., processing time for predicting the tuning index) of the target performance metric. Therefore, the UE can select an ML model capable of predicting the tuning index based on certain performance characteristics, operating environments, and / or input-output schemes as described above.

[0123] It should be noted that training architecture 800 is an example of deep learning, and any suitable training architecture can be used to train ML model 808 as a supplement to or alternative to training architecture 800. Example operation for multi-stage impedance tuning

[0124] Figure 9 An example operation 900 for wireless communication is illustrated. Operation 900 may be performed, for example, by a wireless device (e.g., the first wireless device 102 in wireless communication system 100). In some aspects, operation 900 may be performed by an RF transmitter (e.g., RF transmitter 300). Operation 900 may be implemented in one or more processors (e.g., Figure 2 Modem 210 and / or processor 212 and / or Figure 3 Software components executed and running on the processor 362. Furthermore, the transmission and / or reception of signals by the wireless device in operation 900 may be, for example, by one or more antennas (e.g., Figure 2 This is achieved via antenna 220 and / or antenna 310. In some aspects, the transmission and / or reception of signals by the wireless device may be achieved via one or more processors (e.g., Figure 2 Modem 210 and / or processor 212 and / or Figure 3 The processor 362) obtains and / or outputs signals for receiving or transmitting through its bus interface.

[0125] Operation 900 may optionally begin at block 902, wherein the wireless device configures at least one parameter of at least one of one or more impedance tuning circuits (e.g., impedance tuning circuit 304) or one or more amplifiers (e.g., amplifier 302) based at least in part on at least one load characteristic at an antenna feed (e.g., antenna feed 332). In some aspects, the wireless device includes RF circuitry (e.g., RF transmitter 300) comprising: one or more amplifiers (e.g., amplifier 302); one or more impedance tuning circuits coupled to one or more outputs of the one or more amplifiers; and an antenna tuner (e.g., antenna tuner 308) coupled between the antenna feed and one or more outputs of the one or more impedance tuning circuits. In some aspects, at least one load characteristic includes one or more of the load impedance or load phase at the antenna feed.

[0126] At block 904, the wireless device transmits one or more signals via RF circuitry when at least one parameter is configured based at least in part on at least one load characteristic at the antenna feed. For example, the wireless device may transmit signals to another wireless communication device (e.g., Figure 1 The second wireless device (any of the second wireless devices 104 depicted herein) transmits a signal. This signal may indicate (or carry) any information of various kinds, such as data and / or control information. In some cases, the signal may indicate (or carry) one or more packets or data blocks.

[0127] In some aspects, a wireless device may be configured with at least one parameter to achieve (or be predicted to achieve) one or more performance metrics under at least one load characteristic. In some aspects, a wireless device may configure at least one parameter based at least in part on the relationship (e.g., mapping 400) between one or more performance metrics associated with RF circuitry and a combination of at least one parameter and at least one load characteristic.

[0128] In some aspects, at least one parameter includes the tuning index of one or more impedance tuning circuits. In some aspects, the tuning index corresponds to the reactance value of at least one reactive component of one or more impedance tuning circuits and / or a reference current (or bias voltage or bias current or quiescent current, which in some examples may be considered as a reference current) applied to one or more amplifiers. In some aspects, at least one parameter includes one or more of the following: the reactance of at least one reactive component of one or more impedance tuning circuits; or a reference current applied to one or more amplifiers. In some aspects, one or more amplifiers include a driver amplifier and a power amplifier, and the reference current includes a first reference current applied to the driver amplifier and a second reference current applied to the power amplifier.

[0129] In some aspects, one or more performance metrics include one or more of the following: operating temperature of the RF circuit; output frequency of the RF circuit; voltage standing wave ratio (VSWR) of the RF circuit; error vector magnitude (EVM) of the RF circuit; adjacent channel leakage ratio (ACLR) of the RF circuit; peak-to-average power ratio (PAPR) of the RF circuit; output power level of the RF circuit; or power-added efficiency (PAE) of one or more amplifiers. In some aspects, the wireless device may monitor one or more performance metrics and generate relationships based on the monitored performance metrics. In some aspects, the wireless device may monitor one or more performance metrics and configure at least one parameter in response to changes in the monitored performance metrics.

[0130] In some respects, the wireless device may set or configure the impedance of the antenna tuner based at least in part on at least one load characteristic. The wireless device may set the impedance of the antenna tuner to match the load impedance and load phase of at least one load characteristic.

[0131] To configure at least one parameter, a wireless device can use an ML model, such as the one described in this article. Figure 3 As described. A wireless device may provide input data to an ML model, the input data including at least one load characteristic and / or one or more performance metrics associated with an RF circuit. The wireless device may obtain output data from the ML model including at least one parameter. The ML model can be trained to predict at least one parameter that enables one or more performance metrics associated with the RF circuit. In some aspects, the input data also includes at least one parameter and at least one load characteristic in a first state (e.g., at a first moment), and the output data includes at least one parameter in a second state (e.g., in the future or at the next moment), the second state being predicted to enable one or more performance metrics associated with the RF circuit under at least one load characteristic. In some aspects, the wireless device may train the ML model at least partially based on a heatmap that maps multiple values ​​of one or more performance metrics to a combination of at least one load characteristic and at least one parameter.

[0132] In some aspects, to configure at least one parameter, a wireless device may search for at least one parameter in a heatmap that maps multiple values ​​of one or more performance metrics associated with RF circuitry to a combination of at least one load characteristic and at least one parameter. In some aspects, the wireless device may search for at least one parameter, as represented by the heatmap, that implements one or more performance metrics under at least one load characteristic. Example communication device

[0133] Figure 10 Various aspects of the example communication device 1000 are described. In some aspects, the communication device 1000 is a wireless communication device, as described above. Figure 1 and Figure 2 The first wireless device 102 is described. In some aspects, the communication device 1000 may include an RF transmitter configured to perform the multi-stage impedance tuning described herein, such as... Figure 3 RF transmitter 300.

[0134] Communication device 1000 includes devices coupled to transceiver 1008 (e.g., transmitter and / or receiver, and / or...). Figure 3 The processing system 1002 is an RF transmitter 300. The transceiver 1008 is configured to transmit and receive signals for the communication device 1000, such as the various signals described herein, via antenna 1010. The processing system 1002 may be configured to perform processing functions for the communication device 1000, including processing signals received by the communication device 1000 and / or to be transmitted by the communication device.

[0135] Processing system 1002 includes one or more processors 1020. In various aspects, the one or more processors 1020 may be represented as... Figure 2 Either the described modem 210 and / or processor 212. One or more processors 1020 are coupled to computer-readable medium / memory 1030 via bus 1006. In some aspects, the computer-readable medium / memory 1030 is configured to store instructions (e.g., computer-executable code) that, when executed by one or more processors 1020, cause one or more processors 1020 to perform actions related to... Figure 9 The described operation 900 or any aspect relating to the operation described herein. It should be noted that references to a processor performing the functions of communication device 1000 may include one or more processors performing the functions of communication device 1000. References to one or more processors performing multiple functions may include any one of those processors performing any one of the multiple functions.

[0136] In the depicted example, computer-readable medium / memory 1030 stores code 1031 (e.g., processor-executable instructions) for configuration, code 1032 for communication, code 1033 for providing, code 1034 for obtaining, code 1035 for training, code 1036 for searching, or any combination thereof. Processing of codes 1031-1036 causes the communication device 1000 to perform relative to... Figure 9 The operation described herein is 900 or any aspect relating to the operation described herein.

[0137] One or more processors 1020 include circuitry configured to implement (e.g., execute) code stored in computer-readable medium / memory 1030, including circuitry 1021 for configuration, circuitry 1022 for communication, circuitry 1023 for provisioning, circuitry 1024 for acquisition, circuitry 1025 for training, circuitry 1026 for searching, or any combination thereof. Processing using circuitry 1021 to 1026 enables communication device 1000 to perform operations related to… Figure 9 The operation described herein is 900 or any aspect relating to the operation described herein.

[0138] The various components of the communication device 1000 can provide for performing tasks related to... Figure 9 The components described are the operation 900 or any aspect of the operation described herein. For example, components for conveying, sending, transmitting, or outputting for transmission may include... Figure 2 The illustrated first wireless device 102's TX path 218 and / or antenna 220, and / or Figure 10 The communication device 1000 includes a transceiver 1008 and an antenna 1010. Components used for transmitting, receiving, or acquiring data may include... Figure 2 The illustrated first wireless device's RX path 222 and / or antenna 220, and / or Figure 10 The communication device 1000 includes a transceiver 1008 and an antenna 1010. Components for configuring, providing, acquiring, training, and / or searching may include one or more processors, such as... Figure 2 The modem 210 and / or processor 212 depicted therein, and / or Figure 10 The processor in it is 1020. Example

[0139] Specific implementation examples are described in the following numbered clauses:

[0140] Aspect 1: An apparatus configured for wireless communication, the apparatus comprising: a radio frequency (RF) circuit including: one or more amplifiers; one or more impedance tuning circuits coupled to one or more outputs of the one or more amplifiers; and an antenna tuner coupled between an antenna feed and one or more outputs of the one or more impedance tuning circuits; one or more memories; and one or more processors coupled to the one or more memories, the one or more processors being configured to cause the apparatus to: configure at least one parameter of at least one of the one or more impedance tuning circuits or the one or more amplifiers at least in part based on at least one load characteristic at the antenna feed; and transmit one or more signals via the RF circuit when the at least one parameter is configured at least in part based on the at least one load characteristic at the antenna feed.

[0141] Aspect 2: The apparatus according to aspect 1, wherein, in order to configure the at least one parameter, the one or more processors are configured to cause the apparatus to configure the at least one parameter to achieve one or more performance metrics under the at least one load characteristic.

[0142] Aspect 3: The apparatus according to aspect 1 or 2, wherein, in order to configure the at least one parameter, the one or more processors are configured such that the apparatus configures the at least one parameter at least in part based on the relationship between one or more performance metrics associated with the RF circuit and a combination of the at least one parameter and the at least one load characteristic.

[0143] Aspect 4: The apparatus according to aspect 2 or 3, wherein the one or more performance metrics include one or more of the following: the operating temperature of the RF circuit; the output frequency of the RF circuit; the voltage standing wave ratio (VSWR) of the RF circuit; the error vector magnitude (EVM) of the RF circuit; the adjacent channel leakage ratio (ACLR) of the RF circuit; the peak-to-average power ratio (PAPR) of the RF circuit; the output power level of the RF circuit; or the power-added efficiency (PAE) of the one or more amplifiers.

[0144] Aspect 5: The apparatus according to any one of Aspects 2 to 4, wherein: the at least one load characteristic includes one or more of the load impedance or load phase at the antenna feed; and the at least one parameter includes the tuning index of the one or more impedance tuning circuits.

[0145] Aspect 6: The apparatus according to aspect 5, wherein the tuning index corresponds to the reactance value of at least one reactance component of the one or more impedance tuning circuits.

[0146] Aspect 7: The apparatus according to any one of Aspects 1 to 6, wherein the at least one parameter includes one or more of the following: the reactance of at least one reactance component of the one or more impedance tuning circuits; or a reference current applied to the one or more amplifiers.

[0147] Aspect 8: The apparatus according to aspect 7, wherein: the one or more amplifiers include a driver amplifier and a power amplifier; and the reference current includes a first reference current applied to the driver amplifier and a second reference current applied to the power amplifier.

[0148] Aspect 9: The apparatus according to any one of Aspects 1 to 8, wherein the one or more processors are configured to set the impedance of the antenna tuner at least in part based on the at least one load characteristic.

[0149] Aspect 10: The apparatus according to aspect 9, wherein, in order to set the impedance of the antenna tuner, the one or more processors are configured to cause the apparatus to set the impedance of the antenna tuner to match the load impedance and load phase of the at least one load characteristic.

[0150] Aspect 11: An apparatus according to any one of aspects 1 to 10, wherein, in order to configure the at least one parameter, the one or more processors are configured to cause the apparatus to: provide input data including the at least one load characteristic to a machine learning (ML) model; and obtain output data including the at least one parameter from the ML model.

[0151] Aspect 12: The apparatus according to aspect 11, wherein the ML model is trained to predict the at least one parameter for implementing one or more performance metrics associated with the RF circuit.

[0152] Aspect 13: The apparatus according to aspect 11 or 12, wherein: the input data further includes the at least one parameter and the at least one load characteristic in a first state; and the output data includes the at least one parameter in a second state, the second state being predicted to achieve one or more performance metrics associated with the RF circuit under the at least one load characteristic.

[0153] Aspect 14: An apparatus according to any one of Aspects 11 to 13, wherein the one or more processors are configured to train the ML model at least in part based on a heatmap, the heatmap mapping multiple values ​​of one or more performance metrics associated with the RF circuit to a combination of the at least one load characteristic and the at least one parameter.

[0154] Aspect 15: An apparatus according to any one of Aspects 1 to 14, wherein, in order to configure the at least one parameter, the one or more processors are configured to cause the apparatus to search for the at least one parameter in a heatmap that maps multiple values ​​of one or more performance metrics associated with the RF circuitry to a combination of the at least one load characteristic and the at least one parameter.

[0155] Aspect 16: The apparatus according to aspect 15, wherein, in order to search for the at least one parameter, the one or more processors are configured to cause the apparatus to search for the at least one parameter that achieves the one or more performance metrics under the at least one load characteristic, as represented by the heatmap.

[0156] Aspect 17: A method of wireless communication by means of a device, the method comprising: configuring at least one parameter of at least one of one or more impedance tuning circuits or one or more amplifiers based at least in part on at least one load characteristic at an antenna feed, wherein the means includes radio frequency (RF) circuitry including: the one or more amplifiers; the one or more impedance tuning circuits coupled to one or more outputs of the one or more amplifiers; and an antenna tuner coupled between the antenna feed and one or more outputs of the one or more impedance tuning circuits; and transmitting one or more signals via the RF circuitry when the at least one parameter is configured at least in part on at least one load characteristic at the antenna feed.

[0157] Aspect 18: According to the method of aspect 17, configuring the at least one parameter includes: configuring the at least one parameter to achieve one or more performance metrics under the at least one load characteristic.

[0158] Aspect 19: The method according to aspect 17 or 18, wherein configuring the at least one parameter comprises: configuring the at least one parameter based at least in part on the relationship between one or more performance metrics associated with the RF circuit and a combination of the at least one parameter and the at least one load characteristic.

[0159] Aspect 20: The method according to aspect 18 or 19, wherein the one or more performance metrics include one or more of the following: the operating temperature of the RF circuit; the output frequency of the RF circuit; the voltage standing wave ratio (VSWR) of the RF circuit; the error vector magnitude (EVM) of the RF circuit; the adjacent channel leakage ratio (ACLR) of the RF circuit; the peak-to-average power ratio (PAPR) of the RF circuit; the output power level of the RF circuit; or the power-added efficiency (PAE) of the one or more amplifiers.

[0160] Aspect 21: The method according to any one of Aspects 18 to 20, wherein: the at least one load characteristic includes one or more of the load impedance or load phase at the antenna feed; and the at least one parameter includes the tuning index of the one or more impedance tuning circuits.

[0161] Aspect 22: According to the method of aspect 21, wherein the tuning index corresponds to the reactance value of at least one reactance component of the one or more impedance tuning circuits.

[0162] Aspect 23: The method according to any one of Aspects 17 to 22, wherein the at least one parameter includes one or more of the following: the reactance of at least one reactive component of the one or more impedance tuning circuits; or a reference current applied to the one or more amplifiers.

[0163] Aspect 24: The method according to aspect 23, wherein: the one or more amplifiers include a driver amplifier and a power amplifier; and the reference current includes a first reference current applied to the driver amplifier and a second reference current applied to the power amplifier.

[0164] Aspect 25: The method according to any one of aspects 17 to 24, the method further comprising setting the impedance of the antenna tuner at least in part based on the at least one load characteristic.

[0165] Aspect 26: According to the method of aspect 25, setting the impedance of the antenna tuner includes: setting the impedance of the antenna tuner to match the load impedance and load phase of the at least one load characteristic.

[0166] Aspect 27: The method according to any one of Aspects 17 to 26, wherein configuring the at least one parameter comprises: providing input data including the at least one load characteristic to a machine learning (ML) model; and obtaining output data including the at least one parameter from the ML model.

[0167] Aspect 28: The method according to aspect 27, wherein the ML model is trained to predict the at least one parameter for implementing one or more performance metrics associated with the RF circuit.

[0168] Aspect 29: The method according to aspect 27 or 28, wherein: the input data includes the at least one parameter and the at least one load characteristic in a first state; and the output data includes the at least one parameter in a second state, the second state being predicted to achieve one or more performance metrics associated with the RF circuit under the at least one load characteristic.

[0169] Aspect 30: The method according to any one of Aspects 27 to 29, the method further comprising training the ML model at least in part based on a heatmap that maps multiple values ​​of one or more performance metrics associated with the RF circuit to a combination of the at least one load characteristic and the at least one parameter.

[0170] Aspect 31: The method according to any one of Aspects 17 to 30, wherein configuring the at least one parameter comprises: searching for the at least one parameter in a heatmap that maps multiple values ​​of one or more performance metrics associated with the RF circuit to a combination of the at least one load characteristic and the at least one parameter.

[0171] Aspect 32: According to the method of aspect 31, searching for the at least one parameter includes: searching for the at least one parameter that achieves the one or more performance metrics under the at least one load characteristic as represented by the heatmap.

[0172] Aspect 33: A radio frequency transmitter comprising: one or more amplifiers; one or more impedance tuning circuits coupled to one or more outputs of the one or more amplifiers; an antenna tuner coupled between an antenna feed and one or more outputs of the one or more impedance tuning circuits; one or more memories; and one or more processors coupled to the one or more memories, the one or more processors being configured to: configure a tuning index of at least one of the one or more impedance tuning circuits or the one or more amplifiers at least in part based on at least one load characteristic at the antenna feed; and transmit one or more signals when the tuning index is configured at least in part based on the at least one load characteristic at the antenna feed.

[0173] Aspect 34: An apparatus comprising: a memory; and one or more processors configured to perform the method according to any one of aspects 17 to 32.

[0174] Aspect 35: An apparatus comprising components for performing the method according to any one of aspects 17 to 32.

[0175] Aspect 36: A non-transitory computer-readable medium comprising computer-executable instructions that, when executed by one or more processors of a processing system, cause the processing system to perform the method according to any one of aspects 17 to 32.

[0176] Aspect 37: A computer program product embodied on a computer-readable storage medium, the computer program product comprising code for performing the method according to any one of aspects 17 to 32. Additional Notes

[0177] The foregoing description is provided to enable any person skilled in the art to practice the various aspects described herein. The examples discussed herein do not limit the scope, applicability, or aspects set forth in the claims. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects. For example, the function and arrangement of the elements discussed may be changed without departing from the scope of this disclosure. Various processes or components may be omitted, substituted, or added as appropriate in the various examples. For example, the described methods may be performed in a different order than described, and various actions may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined in some other examples. For example, any number of aspects set forth herein may be used to implement an apparatus or practice. Moreover, the scope of this disclosure is intended to cover such apparatuses or methods practiced using other structures, functionalities, or structures and functionalities that complement or replace the various aspects of this disclosure set forth herein. It should be understood that any aspect of the disclosure herein may be embodied by one or more elements of the claims.

[0178] The various exemplary logic blocks, modules, and circuits described in this disclosure may be implemented or executed using a microcontroller, microprocessor, general-purpose processor, digital signal processor (DSP), artificial intelligence processor, application-specific integrated circuit (ASIC), field-programmable gate array (FPGA) or other programmable logic device (PLD), discrete gate or transistor logic unit, discrete hardware component, or any combination thereof, designed to perform the functions described herein. While a general-purpose processor may be a microprocessor, in alternative embodiments, the processor may be any commercially available processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors working in conjunction with a DSP core, a system-on-a-chip (SoC), or any other such configuration.

[0179] As used in this article, the phrase “at least one of” in a list of items refers to any combination of these entries, including a single member. As an example, “at least one of a, b, or c” is intended to cover: a, b, c, ab, ac, bc, and abc, as well as any combination with multiple identical elements (e.g., aa, aaa, aab, aac, abb, acc, bb, bbb, bbc, cc, and ccc, or any other ordering of a, b, and c).

[0180] As used herein, the term "determine" encompasses a wide variety of actions. For example, "determine" can include calculation, computation, processing, derivation, research, searching (e.g., looking in a table, database, or other data structure), ascertainment, and similar actions. Additionally, "determine" can include receiving (e.g., receiving information), accessing (e.g., accessing data in memory), etc. Furthermore, "determine" can include parsing, selecting, identifying, mapping, applying, picking, building, etc.

[0181] The methods disclosed herein include one or more actions for implementing the methods. These actions may be interchanged without departing from the scope of the claims. In other words, unless a specific order of actions is specified, the order and / or use of a particular action may be modified without departing from the scope of the claims. Furthermore, the various operations of the methods described above may be performed by any suitable component capable of performing the corresponding function. This component may include various hardware and / or software components and / or modules, including but not limited to circuits, application-specific integrated circuits (ASICs), or processors.

[0182] The following claims are not intended to be limited to the aspects shown herein, but should be given the full scope consistent with the language of the claims. Within the claims, unless specifically stated otherwise, reference to an element in the singular form is not intended to mean “one and only one,” but rather “one or more.” The use of a definite article (e.g., “the” or “the”) before an element is not intended to impose a singular meaning (e.g., “one and only one”) onto that element which would otherwise have a plural meaning (e.g., “one or more”), unless specifically stated otherwise. Unless specifically stated otherwise, the term “some” means one or more. No element of any claim is to be interpreted in accordance with 35 USC § 112(f) unless the element is explicitly stated using the phrase “part for…”. All structural and functional equivalents of elements throughout the various aspects described in this disclosure that are known to a person of ordinary skill in the art, or will later be known, are expressly incorporated herein by reference and are intended to be covered by the claims. Furthermore, nothing disclosed herein is intended to be offered to the public, whether or not such disclosure is explicitly recited in the claims.

Claims

1. An apparatus configured for wireless communication, the apparatus comprising: Radio frequency (RF) circuit, the radio frequency (RF) circuit comprising: One or more amplifiers, One or more impedance tuning circuits, said one or more impedance tuning circuits being coupled to one or more outputs of said one or more amplifiers, and An antenna tuner, said antenna tuner being coupled between an antenna feed and one or more outputs of said one or more impedance tuning circuits; One or more memories; and One or more processors, said one or more processors coupled to said one or more memories, said one or more processors being configured to cause the device to: At least one parameter of at least one of the one or more impedance tuning circuits or the one or more amplifiers is configured based at least in part on at least one load characteristic at the antenna feed; and One or more signals are transmitted via the RF circuit when the at least one parameter is configured at least in part based on the at least one load characteristic at the antenna feed.

2. The apparatus of claim 1, wherein, in order to configure the at least one parameter, the one or more processors are configured to cause the apparatus to configure the at least one parameter to achieve one or more performance metrics under the at least one load characteristic.

3. The apparatus of claim 1, wherein, in order to configure the at least one parameter, the one or more processors are configured such that the apparatus configures the at least one parameter at least in part based on a relationship between one or more performance metrics associated with the RF circuit and a combination of the at least one parameter and the at least one load characteristic.

4. The apparatus of claim 2, wherein the one or more performance metrics include one or more of the following: The voltage standing wave ratio (VSWR) of the RF circuit. Error vector magnitude (EVM) of the RF circuit. Adjacent Channel Leakage Ratio (ACLR) of the RF circuit. The peak-to-average power ratio (PAPR) of the RF circuit. The output power level of the RF circuit; or The power-added efficiency (PAE) of the one or more amplifiers.

5. The apparatus according to claim 2, wherein: The at least one load characteristic includes one or more of the load impedance or load phase at the antenna feed; and The at least one parameter includes the tuning index of the one or more impedance tuning circuits.

6. The apparatus of claim 5, wherein the tuning index corresponds to the reactance value of at least one reactance component of the one or more impedance tuning circuits.

7. The apparatus of claim 1, wherein the at least one parameter comprises one or more of the following: The reactance of at least one reactance component of the one or more impedance tuning circuits; or A reference current applied to the one or more amplifiers.

8. The apparatus according to claim 7, wherein: The one or more amplifiers include driver amplifiers and power amplifiers; and The reference current includes a first reference current applied to the drive amplifier and a second reference current applied to the power amplifier.

9. The apparatus of claim 1, wherein the one or more processors are configured to cause the apparatus to set the impedance of the antenna tuner at least in part based on the at least one load characteristic.

10. The apparatus of claim 9, wherein, in order to set the impedance of the antenna tuner, the one or more processors are configured to cause the apparatus to set the impedance of the antenna tuner to match the load impedance and load phase of the at least one load characteristic.

11. The apparatus of claim 1, wherein, in order to configure the at least one parameter, the one or more processors are configured to cause the apparatus to: Provide the machine learning (ML) model with input data including at least one of the load characteristics; and Output data, including at least one parameter, is obtained from the ML model.

12. The apparatus of claim 11, wherein the ML model is trained to predict the at least one parameter for implementing one or more performance metrics associated with the RF circuit.

13. The apparatus according to claim 11, wherein: The input data also includes at least one parameter and at least one load characteristic in the first state; and The output data includes at least one parameter in a second state, which is predicted to achieve one or more performance metrics associated with the RF circuit under the at least one load characteristic.

14. The apparatus of claim 11, wherein the one or more processors are configured to train the ML model at least in part based on a heatmap that maps multiple values ​​of one or more performance metrics associated with the RF circuitry to a combination of the at least one load characteristic and the at least one parameter.

15. The apparatus of claim 1, wherein, in order to configure the at least one parameter, the one or more processors are configured to cause the apparatus to search for the at least one parameter in a heatmap that maps multiple values ​​of one or more performance metrics associated with the RF circuitry to a combination of the at least one load characteristic and the at least one parameter.

16. The apparatus of claim 15, wherein, in order to search for the at least one parameter, the one or more processors are configured to cause the apparatus to search for the at least one parameter that achieves the one or more performance metrics under the at least one load characteristic, as represented by the heatmap.

17. A method for wireless communication by a device, the method comprising: At least one parameter of at least one of one or more impedance tuning circuits or one or more amplifiers is configured based at least in part on at least one load characteristic at the antenna feed, wherein the device includes radio frequency (RF) circuitry comprising: the one or more amplifiers; the one or more impedance tuning circuits coupled to one or more outputs of the one or more amplifiers; and an antenna tuner coupled between the antenna feed and one or more outputs of the one or more impedance tuning circuits; and One or more signals are transmitted via the RF circuit when at least one parameter is configured at least in part based on at least one load characteristic at the antenna feed.

18. The method of claim 17, wherein configuring the at least one parameter comprises: Configure the at least one parameter to achieve one or more performance metrics under the at least one load characteristic.

19. The method of claim 17, wherein: Configuring the at least one parameter includes configuring the at least one parameter at least in part based on the relationship between one or more performance metrics associated with the RF circuit and a combination of the at least one parameter and the at least one load characteristic; The at least one load characteristic includes one or more of the load impedance or load phase at the antenna feed; and The at least one parameter includes the tuning index of the one or more impedance tuning circuits.

20. A radio frequency transmitter, the radio frequency transmitter comprising: One or more amplifiers; One or more impedance tuning circuits, said one or more impedance tuning circuits being coupled to one or more outputs of said one or more amplifiers; An antenna tuner, said antenna tuner being coupled between an antenna feed and one or more outputs of said one or more impedance tuning circuits; One or more memory units; and One or more processors, said one or more processors coupled to said one or more memories, said one or more processors being configured to: The tuning index of at least one of the one or more impedance tuning circuits or the one or more amplifiers is configured at least in part based on at least one load characteristic at the antenna feed. as well as One or more signals are transmitted when the tuning index is configured at least in part based on the at least one load characteristic at the antenna feed.

Images