WO2019165969A1 - 用于信道均衡的方法、设备和计算机可读介质 - Google Patents

用于信道均衡的方法、设备和计算机可读介质 Download PDF

Info

Publication number
WO2019165969A1
WO2019165969A1 PCT/CN2019/076291 CN2019076291W WO2019165969A1 WO 2019165969 A1 WO2019165969 A1 WO 2019165969A1 CN 2019076291 W CN2019076291 W CN 2019076291W WO 2019165969 A1 WO2019165969 A1 WO 2019165969A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
direct association
channel equalization
subcarriers
payload
Prior art date
Application number
PCT/CN2019/076291
Other languages
English (en)
French (fr)
Inventor
叶晨晖
胡小锋
张东旭
张凯宾
Original Assignee
上海诺基亚贝尔股份有限公司
诺基亚通信公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 上海诺基亚贝尔股份有限公司, 诺基亚通信公司 filed Critical 上海诺基亚贝尔股份有限公司
Priority to US16/975,912 priority Critical patent/US11516053B2/en
Publication of WO2019165969A1 publication Critical patent/WO2019165969A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L25/03165—Arrangements for removing intersymbol interference using neural networks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L25/03012—Arrangements for removing intersymbol interference operating in the time domain
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L25/03159—Arrangements for removing intersymbol interference operating in the frequency domain
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L25/03821—Inter-carrier interference cancellation [ICI]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L2025/0335—Arrangements for removing intersymbol interference characterised by the type of transmission
    • H04L2025/03375—Passband transmission
    • H04L2025/03414—Multicarrier

Definitions

  • Embodiments of the present disclosure are primarily directed to communication technologies and, more particularly, to methods, apparatuses, and computer readable media for channel equalization.
  • the transmitting end performs corresponding processing on the signal, such as modulation, encoding, etc., before transmitting the signal, and transmits the processed signal to the receiving end via the communication channel.
  • the signal may experience various impairments with noise and/or distortion.
  • the receiving end typically uses a corresponding equalization algorithm to reduce or reduce these impairments.
  • a scheme for channel equalization is provided.
  • a method for channel equalization includes receiving, at a first device, a first signal from a second device over a plurality of subcarriers over a communication channel. The method also includes sampling the first signal to obtain sampled symbols. The method further includes generating a second signal based on the obtained sampled symbols using a predefined direct association between the sampled symbols and the payload, the second signal indicating that the first signal is carried on a valid one of the plurality of subcarriers Payload.
  • an apparatus for channel equalization includes a processor; and a memory coupled to the processor, the memory having instructions stored therein.
  • the instructions when executed by the processor, cause the device to: receive the first signal from the other device via the plurality of subcarriers on the communication channel; sample the first signal to obtain the sampled symbols; and utilize the pre-sampling between the sampled symbol and the payload A defined direct association generates a second signal based on the obtained sampled symbols, the second signal indicating a payload carried by the first signal on a valid one of the plurality of subcarriers.
  • a computer readable medium stores computer executable instructions that, when executed by one or more processors, cause one or more processors to perform the methods of the foregoing first aspect.
  • FIG. 1 is a schematic diagram of a communication system in which embodiments of the present disclosure may be implemented
  • FIGS. 2A and 2B show schematic diagrams of a conventional system for channel equalization
  • FIG. 3 shows a schematic diagram of a system for channel equalization, in accordance with some embodiments of the present disclosure
  • FIG. 4 shows a schematic diagram of a channel equalization model for implementing the integrated equalizer of FIG. 3, in accordance with some embodiments of the present disclosure
  • FIG. 5 illustrates a schematic diagram of a process for channel equalization, in accordance with some embodiments of the present disclosure
  • FIG. 6 shows a schematic diagram of a process for determining a direct association, in accordance with some embodiments of the present disclosure
  • FIG. 7 shows a schematic diagram of a system for training the channel equalization model of FIG. 4, in accordance with some embodiments of the present disclosure
  • FIG. 8 illustrates a flow diagram of a process for channel equalization, in accordance with some embodiments of the present disclosure
  • Figure 9 shows a simplified block diagram of a device suitable for implementing embodiments of the present disclosure.
  • the term “comprises” and the like are to be understood as open-ended, ie, “including but not limited to”.
  • the term “based on” should be understood to mean “based at least in part.”
  • the term “one embodiment” or “an embodiment” should be taken to mean “at least one embodiment.”
  • the terms “first,” “second,” and the like may refer to different or identical objects. Other explicit and implicit definitions may also be included below.
  • a “base station” may represent a Node B (NodeB or NB), an evolved Node B (eNodeB or eNB), a Remote Radio Unit (RRU), a Radio Head (RH), a Remote Radio Head (RRH), a repeater, Or low power nodes such as pico base stations, femto base stations, and the like.
  • terminal device refers to any terminal device capable of wired or wireless communication with or between network devices.
  • the terminal device may include an aircraft, a mobile terminal (MT), a subscriber station (SS), a portable subscriber station (PSS), a mobile station (MS), or an access terminal (AT), and the above-described equipment on board.
  • MT mobile terminal
  • SS subscriber station
  • PSS portable subscriber station
  • MS mobile station
  • AT access terminal
  • the terminal device may be any type of mobile terminal, fixed terminal or portable terminal, including mobile phone, site, unit, device, multimedia computer, multimedia tablet, internet node, communicator, desktop computer, laptop computer, notebook computer, netbook Computer, tablet computer, personal communication system (PCS) device, personal navigation device, personal digital assistant (PDA), audio/video player, digital camera/camcorder, positioning device, television receiver, radio receiver, e-book device , gaming devices, Internet of Things (IoT) devices or other devices that can be used for communication, or any combination of the above.
  • PCS personal communication system
  • PDA personal digital assistant
  • audio/video player digital camera/camcorder
  • positioning device television receiver
  • radio receiver radio receiver
  • e-book device gaming devices
  • IoT Internet of Things
  • the communication system 100 includes a device 110 and a device 120 that can communicate with each other to transmit various service data, control information, and the like.
  • device 110 transmits a signal to device 120, and signals transmitted by device 110 are propagated to device 120 via communication channel 102.
  • device 110 performs processing on the payload data to be transmitted, which is illustrated in FIG. 1 as being performed by signal processing module 112.
  • the signal processing module 112 can perform various processing, including one or more of preprocessing, modulation, time-frequency transform, addition of cyclic prefix, amplification, etc., on the data to be transmitted.
  • the device 110 also includes a transmitter 114 for transmitting the processed signals to the device 120 via the communication channel 102.
  • device 120 On the receiving side, device 120 includes a receiver 122 for receiving signals transmitted over communication channel 102. Since the payload data transmitted by the device 110 undergoes signal processing of the device 110, transmission of the transmitter, transmission of the link, and reception of the receiver upon reaching the device 120, it is possible to withstand various damages in these processes, thereby Noise, distortion, etc. Accordingly, the device 120 further includes an equalization processing module 124 for performing equalization processing on the signals received by the receiver 122 to determine the payload transmitted by the device 110.
  • Communication system 100 can be a wired communication system or a wireless communication system, and thus communication channel 102 can be a wired communication link or a wireless communication link.
  • wired communication systems may include backbone transport networks, such as fiber optic communication networks (active optical networks or passive optical networks), analog forward communication networks, and the like.
  • wireless communication systems can include any radio communication system.
  • Communication in communication system 100 can be implemented in accordance with any suitable communication protocol, including, but not limited to, first generation (1G), second generation (2G), third generation (3G), fourth generation (4G), and Cellular communication protocols such as fifth generation (5G), wireless local area network communication protocols such as Institute of Electrical and Electronics Engineers (IEEE) 802.11, and/or any other protocols currently known or developed in the future.
  • 5G fifth generation
  • IEEE 802.11 Institute of Electrical and Electronics Engineers
  • Devices 110 and 120 can be any device capable of communicating in communication system 100.
  • device 110 may be one of a network device and a terminal device (eg, a UE), and device 120 may be another device of the network device and the terminal device.
  • device 100 may be one of an Optical Line Terminal (OLT) and an Optical Network Unit (ONU), and device 120 may be the other of an OLT and an ONU.
  • OLT Optical Line Terminal
  • ONU Optical Network Unit
  • device 120 is also capable of transmitting signals to device 110.
  • device 120 may have some of the functionality of device 110 described above with respect to FIG. 1, while device 110 may have some of the functionality of device 120 described above with respect to FIG.
  • a cascading module is used to describe the entire communication link through which the signal passes, and then a cascading module is employed to compensate for different aspects.
  • 2A and 2B show an equalization processing system for channel modeling and compensation in a wireless communication system and a wired communication system, respectively.
  • the equalization processing system 200 includes a wireless channel modeling module 210 and a compensation module 220.
  • the wireless channel modeling module 210 includes a plurality of sub-modules for modeling the processing and channels experienced by the received signals.
  • the pre-processing sub-module 211 is used to model pre-processing at the transmitting end, such as pre-processing of the signal in the frequency domain;
  • the time-frequency transform sub-module 212 is used to model the time-frequency transform at the transmitting end, such as inverse fast Fourier transform (iFFT); transmit antenna sub-module 213 for modeling multiple input multiple output (MIMO) and/or power leakage experienced at the transmit antenna; air interface sub-module 214 for modeling during wireless channel propagation Frequency domain and time domain impairments experienced; receive antenna sub-module 215 for modeling power leakage or phase offset introduced in signal reception at the receiving end; sampling and equalization sub-module 216 for modeling symbols in communication process Interference (ISI).
  • the wireless channel modeling sub-module 210 may also include more other sub-modules for modeling other processing experienced by the signal, as desired.
  • the compensation module 220 includes a plurality of cascaded sub-modules for compensating signals from different aspects to achieve channel equalization.
  • the clock recovery sub-module 221 is configured to suppress ISI
  • the de-MIMO sub-module 222 is configured to suppress inter-carrier interference (ICI)
  • the feed-forward equalization sub-module 223 is configured to implement time-domain processing
  • the time-frequency transform sub-module 224 is configured to perform A time-frequency transform (eg, FFT); and a zero-forcing equalization sub-module 225 is used to implement the linear step size.
  • the compensation module 220 may also include more other sub-modules for compensating for signals from other aspects as needed.
  • the equalization processing system 202 illustrated in Figure 2B is suitable for networks based on fiber optic communications.
  • the equalization processing system 202 includes a fiber optic link modeling module 230 that includes a plurality of sub-modules for modeling the processing and channels experienced by the received signals.
  • the electro-optic modulation sub-module 231 is used to model the signal modulation from electrical to optical at the transmitting end;
  • the optical amplification sub-module 232 is used to model the optical amplification at the transmitting end to determine the nonlinear distortion that may be introduced therein.
  • a fiber link sub-module 233 for modeling the fiber link to determine distortions that may be experienced in the fiber link; a light detection sub-module 234 for modeling phase-to-power conversion at the receiving end; and sampling And equalization sub-module 235 is used to model inter-symbol interference (ISI) in the communication process.
  • the fiber link modeling sub-module 230 may also include more other sub-modules for modeling other processes experienced by the signal, as desired.
  • the compensation module 240 includes a plurality of cascaded sub-modules for compensating signals from different aspects to achieve channel equalization.
  • the clock recovery sub-module 241 is used to implement realignment;
  • the N-tap equalizer 242 is used to cancel the ISI;
  • the reverse light detection sub-module 243 is used to determine the reverse response;
  • the nonlinear management sub-module 244 is used to determine the reverse optics.
  • Amplifying the response; and the zero-forcing equalization sub-module 245 is configured to perform frequency domain compensation.
  • the compensation module 240 may also include more other sub-modules for compensating for signals from other aspects as needed.
  • SSM-based solutions are at the expense of power, which is difficult to accept in multi-point access scenarios, such as passive optical network systems or distributed antenna system preamble networks, because these scenarios require higher power levels of the transmitted signal to Make sure that multiple receivers can receive the signal.
  • the problem of solving channel equalization based on artificial intelligence (AI) technology has been proposed.
  • AI artificial intelligence
  • the advantage of the AI-based scheme is the ability to optimize posterior probability estimates without prior knowledge of the channel model, thereby achieving higher performance.
  • the resulting model has reduced computational complexity requirements and can therefore be implemented by lower cost processing devices such as DSPs.
  • the proposed scheme has mainly focused on channel equalization of single carrier formatted signals because channel equalization can be performed on single carrier formatted signals in the pure time domain.
  • channel equalization for multi-carrier formatted signals has not been proposed.
  • a scheme for channel equalization is proposed.
  • the scheme is adapted to perform channel equalization on signals based on multi-carrier transmission.
  • a device at the receiving end receives a signal from another device via a plurality of subcarriers on a communication channel, and samples the received signal to obtain a sampled symbol.
  • another signal is generated based on the obtained sampled symbols, the signal indicating the received payload carried on the valid subcarriers of the plurality of subcarriers.
  • channel equalization refers to suppressing or eliminating impairments such as noise, distortion, etc. introduced in a received signal at the receiving end in order to obtain a payload that the transmitting end desires to transmit.
  • FIG. 3 shows a schematic diagram of a system for channel equalization, in accordance with some embodiments of the present disclosure.
  • the system 300 can be implemented at the device 120 at the receiving end in FIG. 1, for example, in the equalization processing module 124.
  • system 300 will be described with reference to FIG.
  • System 300 includes a sampler 310 for acquiring a signal 302 from a receiver 122 of device 120 (also referred to as a first signal for ease of discussion).
  • the first signal 302 is a frequency domain signal.
  • Sampler 310 samples signal 302 to obtain a series of sample symbols 304.
  • receiver 122 of device 120 receives signal 302 from device 110 over a plurality of subcarriers over communication channel 102.
  • Signal 302 may be an Orthogonal Frequency Division Multiplexing (OFDM) modulated signal, a discrete multi-carrier (DMT) modulated signal, or the like, depending on the processing for multiple subcarriers.
  • OFDM Orthogonal Frequency Division Multiplexing
  • DMT discrete multi-carrier
  • Signal 302 may introduce some noise, distortion, etc. damage at device 110, communication channel 102, and device 120 at the receiving end before being sampled.
  • device 110 may desire to send a payload to device 120.
  • device 110 may perform some processing on the payload, including frequency domain pre-processing, frequency time conversion (such as iFFT), cyclic prefix (CP) addition, amplification, antenna transmission, and the like.
  • frequency domain pre-processing such as iFFT
  • CP cyclic prefix
  • the binary data representing the payload is modulated to the corresponding subcarrier and converted into a time domain signal for transmission in the communication channel 102.
  • Processing in device 110, propagation in communication channel 102, and signal reception at device 120 may all cause damage in signal 302.
  • System 300 also includes an integrated equalizer 320 for obtaining sample symbols 304 from sampler 310 and generating a signal 306 based on the obtained sample symbols 304 using a predefined direct association between the sample symbols and the payload ( Easy to discuss, called the second signal).
  • Signal 306 indicates the payload carried by the first signal on valid subcarriers of the plurality of subcarriers.
  • the payload is data that device 110 desires to transmit to device 120.
  • Signal 306 can be considered a frequency domain signal.
  • channel equalization can be performed on the sample symbols 304 that experience various impairments (such as noise, distortion), thereby extracting the payload transmitted by the device 110.
  • the sampled symbols with the payload by directly correlating the sampled symbols with the payload, it can be used to map the sampled symbols directly to the payload without concern for various intermediates from the time-domain sampled symbols to the frequency-domain payloads.
  • Cascade processing Compared to using multiple cascading modules to model the communication channel and perform corresponding compensation, by directly correlating the sampled symbols with the payload, it is possible to avoid the non-existence in the case where different cascading modules achieve channel equalization in different aspects. Linear damage problem.
  • the channel equalization scheme of the present disclosure can also be advantageously applied to cost sensitive wired communication networks, such as fiber optic communication networks, analog forward communication networks, and the like.
  • the direct association between the sampled symbols and the payload can be learned from previously known sample symbols and payloads.
  • a learning model (referred to as a channel equalization model) can be constructed and the set of parameters of the learning model determined using known sample symbols and payloads to indicate a direct correlation between the sampled symbols and the payload.
  • FIG. 4 illustrates an example architecture of an integrated equalizer 320, in accordance with some embodiments of the present disclosure.
  • the integrated equalizer 320 includes a channel equalization model 420 for generating a payload on the output effective subcarrier based on the input sample symbols.
  • the channel equalization model 420 can be constructed as a multi-layer learning network.
  • the term "learning network” refers to a model that is capable of learning the correlation between the corresponding inputs and outputs from the training data so that the pair of parameters obtained based on the training after the training is completed The given input is processed to generate a corresponding output.
  • the "learning network” can sometimes also be referred to as "neural network,” “learning model,” “network,” or “model.” These terms are used interchangeably herein.
  • the channel equalization model 420 includes an input layer 422, an output layer 424, and one or more hidden layers between the input layer 422 and the output layer 444.
  • Each of the layers in channel equalization model 420 includes a plurality of nodes 426 (also referred to as neurons), each node processing a respective input with a respective excitation function to determine an output.
  • the multiple layers of the channel equalization model 420 are arranged in a hierarchical structure, and the output of the previous layer is adjusted as a parameter to the input of the next layer.
  • Nodes of different layers can be fully connected (ie, each node of the latter layer is connected to all nodes of the previous layer), not fully connected (ie, each node of the latter layer is connected to a part of the node of the previous layer) or a pair A connection (ie, each node of the latter layer is connected to a node of the previous layer).
  • Signals are transmitted at nodes connected in adjacent layers.
  • the parameter set of the channel equalization model 420 includes parameters that map the input employed symbols to the input layer 422, parameters of mapping between respective adjacent layers in the channel equalization model 420, and maps the output of the output layer 424 to the payload. parameter.
  • the number of nodes (denoted as I) of the input layer 422 of the channel equalization model 420 is determined by the parameters of the time-frequency transform in the device 110 and the length of the cyclic prefix.
  • the number of nodes of the input layer 422 can be equal to the sum of the size of the time-varying transform (eg, the length of the iFFT) and the length of the cyclic prefix. For example, if the size of the time-frequency transform in the device 110 at the transmitting end is 1024, and the length of the cyclic prefix is 176, the number of nodes of the input layer may be 1200.
  • the number of nodes (denoted as K) of the output layer 424 of the channel equalization model 420 is determined by the number of valid subcarriers. In general, if device 110 and device 120 are assigned multiple subcarriers for communication, device 110 may select some or all of the subcarriers as valid subcarriers and carry the payload on these active subcarriers. Since the channel equalization model 420 is intended to output the payload on these effective subcarriers, the number of nodes of the output layer can be the same as the number of effective subcarriers. For example, if the size of the frequency transform is 1024, this means that the number of all subcarriers is 1024.
  • device 110 may select a portion of the subcarriers from all of the 1024 subcarriers for carrying payloads (eg, 600, 800, 1000, or other values).
  • payloads eg, 600, 800, 1000, or other values.
  • the determined number of nodes of the input layer 422 of the channel equalization model 420 is greater than the number of nodes of the output layer 424, so the channel equalization model 420 is a model with a wide time domain input and a narrow frequency domain output.
  • sampler 310 may continue to perform sampling on time domain signal 302 to obtain a series of sample symbols 304 at different points in time.
  • the integrated equalizer 320 can also include a plurality of delay units 410 for performing delays on previously sampled symbols. When sampler 310 samples a predetermined number of sample symbols 304, the delayed sample symbols are simultaneously mapped to input layer 422. It should be understood that although in the example of FIG.
  • the number of sample symbols 304 is the same as the number of nodes of the input layer 422, in other examples, the number of sample symbols 304 may be greater or smaller than the number of nodes of the input layer 422, but may be via The parameters are mapped to respective nodes of the input layer 422.
  • the set of parameters of the channel equalization model 420 can be determined using known sample symbols as well as the payload, which relates to the model training process and will be described in detail below.
  • the device 120 may utilize the channel equalization model 420 to perform channel equalization to determine the payload from the received signal 302.
  • the output signal 306 of the channel equalization model 420 can be composed of the output of each output node, and the output of each output node indicates the payload carried on one active subcarrier.
  • the payload carried on each valid subcarrier can be represented by a complex value.
  • FIG. 1 illustrates an example structure of the channel equalization model 420
  • this example is for illustrative purposes only.
  • channel equalization model 420 can be constructed with more layers, more nodes, and/or other structures.
  • the channel equalization model 420 can utilize other model structures in addition to the multi-layer learning network. Embodiments of the present disclosure are not limited in this regard.
  • FIG. 5 illustrates a flow diagram of a process 500 for channel equalization, in accordance with some embodiments of the present disclosure.
  • Process 500 relates to device 110 and device 120 of FIG.
  • Process 500 involves utilizing a direct correlation between the determined sample symbols and the payload to achieve channel equalization at the receiving end.
  • device 110 transmits signal 302 to device 120 over a plurality of subcarriers over a communication channel.
  • Signal 302 is a time domain signal.
  • Device 120 receives signal 302 using receiver 122 and samples signal 302 through sampler 310 to obtain sample symbols 304.
  • the integrated equalizer 320 of the device 120 then generates a signal 306 based on the sampled symbols 304 of the signal 302 using a direct association between the predefined sample symbols and the payload to indicate the payload carried on the valid subcarriers.
  • the integrated equalizer 320 can utilize the channel equalization model 420 as shown in FIG. 4 to generate the signal 306.
  • the set of parameters in the channel equalization model 420 indicates a direct correlation between the input sample symbols and the output payload.
  • channel equalization is implemented by device 120 in the event that a direct correlation between sampled symbols and payload (e.g., a parameter set of channel equalization model 420) has been determined. How to determine such a direct association will be described in detail below.
  • payload e.g., a parameter set of channel equalization model 420
  • determining a direct association between a sampled symbol and a payload involves learning a mapping relationship between the known input and output.
  • the goal of learning is to enable direct associations to be used to implement input-based sampling symbols to generate payloads.
  • determining the directly associated training data can include known sampling symbols and the payload corresponding to the sampling symbols.
  • the direct association between the sampled symbols and the payload can be determined online such that the determined direct association is more suitable for channel equalization of signals transmitted from device 110 to device 120.
  • FIG. 6 illustrates a flow diagram of a process 600 for determining a direct association, in accordance with some embodiments of the present disclosure.
  • Process 600 relates to device 110 and device 120 of FIG.
  • device 110 transmits a training signal to device 120 over communication channel 102 via a plurality of subcarriers.
  • the training signal is used to determine the purpose of the direct association between the sampled symbols and the payload.
  • device 110 processes the payload of the training signal in the same manner as subsequent normal communications, in the same manner.
  • the communication channel 102 and the plurality of subcarriers for the transmission of the training signal are also the channels and subcarriers to be employed in subsequent normal communication of the device 110 with the device 120. In this way, a direct association for channel equalization can be better determined for communication of device 110 with device 120.
  • device 110 can also transmit a reference signal to device 120 indicating the payload of the training signal carried on the active subcarrier.
  • the reference signal can also be used by device 120 to determine a direct association.
  • device 110 does not perform additional processing on the reference signal, such as frequency domain pre-processing, time-frequency transform, cyclic prefix addition, amplification, etc., but instead sends the payload directly to device 120.
  • device 110 can send a reference signal directly to device 120.
  • device 110 may also transmit a reference signal to device 120 via other devices.
  • the reference signal is transmitted by device 110 to device 120 in the example of FIG. However, in other examples, the reference signal may also be pre-configured in device 120. In these examples, step 610 in process 600 may be omitted.
  • Device 120 receives the training signal and the reference signal. At 615, device 120 samples the received training signal to obtain a training sample signal. At 620, device 120 determines a direct association based on the training sample symbols and the reference signals. Device 120 may utilize a variety of training algorithms to determine direct associations. The determination of the direct association may depend on the model constructed to represent the direct association. In the training process of the channel equalization model 420 of FIG. 4, the determination of direct association may be implemented by a variety of machine learning algorithms, such as a stochastic gradient descent algorithm, a backward propagation algorithm, and the like.
  • the device 110 since the number of nodes of the input layer of the channel equalization model 420 is related to the parameters of the time-frequency transform and the length of the cyclic prefix in the device 110, and the number of nodes of the output layer is selected by the device 110.
  • the device 110 also receives information from the device 120 indicating the parameters of the time-frequency transform and the length of the cyclic prefix (referred to as first information) and indicating valid actor Information on the number of carriers (referred to as second information).
  • first information the parameters of the time-frequency transform and the length of the cyclic prefix
  • second information indicating valid actor Information on the number of carriers
  • the parameters of the time-frequency transform, the length of the cyclic prefix, and the number of valid sub-carriers that device 110 may employ may be pre-configured into device 120 without being received from device 110.
  • the device 120 may also send a determination indication to the device 110 at 625 to indicate that the direct association has been determined. After device 110 receives such an acknowledgment indication, normal transmission with device 120 can be initiated.
  • FIG. 7 shows a schematic diagram of a system 700 for training the channel equalization model 420 of FIG.
  • the architecture of channel equalization model 420 has been constructed to include the number of layers of the model, the number of nodes per layer, and the node connections between layers.
  • the parameter set for this model can then be determined by a training process, such as process 600 described in FIG.
  • the parameter set of the channel equalization model 420 is initialized.
  • the training signal 702 received by device 110 is sampled (e.g., sampled by sampler 310) to obtain training sample symbols. Since the training signal 702 is a time domain signal, the training sample symbols at an earlier point in time are delayed by the respective delay unit 410.
  • a corresponding number of training sample symbols are input to the input layer 422 of the channel equalization model 420.
  • the training sample symbols are mapped to the output layer 424 via the current parameter set of the channel equalization model 420 to obtain the current payload output.
  • System 700 includes a training library 710 in which reference signals 706 corresponding to training signals 702 (i.e., payloads on respective valid subcarriers) are stored.
  • System 700 also includes a plurality of transfer elements 720.
  • the training library 710 provides the payload indicated by the reference signal 706 to the corresponding delivery element 720.
  • a plurality of error calculation components 730 in system 700 are used to calculate the payload of each node output of output layer 424 and the reference payload obtained by transfer component 720 to determine an error between the two (denoted as e1 to eK, Where K represents the number of nodes of the output layer 424). These errors are summed by summing unit 740 to calculate the sum of the errors. This sum of errors can be used to adjust the parameter set of the channel equalization model 420, such as mapping parameters between different layers, possible bias parameters, and the like.
  • System 700 can include a parameter adjustment unit (not shown) for applying adjustments to respective parameters of channel equalization model 420 (such as modulation 1, adjustment 2, ... adjustment n, n representing adjacent two layers The number of maps between)).
  • a plurality of training signals can be continuously input to the channel equalization model 420 so that the parameter set of the channel equalization model 420 is continuously optimized through the above process.
  • a loss function eg, a cross entropy function
  • the value of the loss function can be continuously reduced.
  • the channel equalization model 420 can be considered to have converged, and the current set of parameters can be used to indicate a direct association between the sampled symbols and the payload.
  • FIG. 7 shows only one specific example system for training the channel equalization model 420, some of which are used only during the training phase and may be hidden during the use phase. In other embodiments, other training or learning methods may also be employed to determine the parameter set of the channel equalization model 420.
  • FIG. 8 shows a flow diagram of a process 800 for channel equalization, in accordance with some embodiments of the present disclosure. It will be appreciated that process 800 can be implemented, for example, at device 120 as shown in FIG. 1 and can be implemented by system 300 of FIG. For convenience of description, the process 800 will be described below in conjunction with FIGS. 1 and 3.
  • the receiver 122 of the device 120 receives a signal 302 (also referred to as a first signal) from the device 110 (also referred to as a second device) over a plurality of subcarriers over a communication channel.
  • sampler 310 of device 120 samples first signal 302 to obtain sample symbol 304.
  • the integrated equalizer 320 of the device 120 generates a signal 306 (also referred to as a second signal) based on the obtained sample symbols 304 using a predefined direct association between the sampled symbols and the payload.
  • the second signal 306 indicates the payload carried by the first signal 302 on valid subcarriers of the plurality of subcarriers.
  • the process 800 can also include receiving, on the communication channel 102, the training signal from the second device 110 via the plurality of subcarriers; sampling the training signal to obtain the training sample symbol; acquiring the reference signal, the reference signal indicating the training signal A payload carried on a valid subcarrier; and determining a direct association based on the trained sample symbols and the reference signal.
  • determining the direct association can include determining a parameter set of the channel equalization model based on the training sample symbols and the reference signal, the parameter set indicating a direct association.
  • determining the direct association may further comprise: receiving first information from the second device 110, the first information indicating a parameter of the time-frequency transform and a length of the cyclic prefix in the second device 110; and a frequency-time transform based
  • the parameters and the length of the cyclic prefix determine the number of nodes in the input layer in the channel equalization model.
  • determining the direct association may further comprise: receiving second information from the second device 110, the second information indicating a number of valid subcarriers; and determining an output layer in the channel equalization model based on the number of effective subcarriers The number of nodes.
  • obtaining the reference signal can include receiving the reference signal from the second device 110.
  • the process 800 can also include transmitting an acknowledgement indication to the second device 110 to indicate that the direct association has been determined.
  • communication channel 102 can include a wired communication channel or a wireless communication channel.
  • FIG. 9 shows a simplified block diagram of an apparatus 900 suitable for implementing an implementation of the present disclosure.
  • Device 900 can be used to implement a communication device, such as device 120 shown in FIG.
  • device 900 includes one or more processors 910 coupled to one or more memories 920 of processor(s) 910, coupled to one or more transmitters and/or receivers of processor 910 (TX/RX) 940.
  • processors 910 coupled to one or more memories 920 of processor(s) 910, coupled to one or more transmitters and/or receivers of processor 910 (TX/RX) 940.
  • TX/RX transmitters and/or receivers of processor 910
  • Processor 910 can have any type suitable for the local technical environment, and can include, by way of non-limiting example, one or more of the following: general purpose computers, special purpose computers, microprocessors, digital signal processors (DSPs), and multi-core based processors
  • Device 900 can have multiple processors, such as an application specific integrated circuit chip that follows the clock synchronized with the main processor in time.
  • the memory 920 can be of any type suitable for the local technical environment and can be implemented using any suitable data storage technology, such as non-transitory computer readable storage media, semiconductor based storage devices, magnetic memory devices, and Systems, optical memory devices and systems, fixed memory, and removable memory.
  • suitable data storage technology such as non-transitory computer readable storage media, semiconductor based storage devices, magnetic memory devices, and Systems, optical memory devices and systems, fixed memory, and removable memory.
  • Memory 920 stores at least a portion of program 930.
  • the TX/RX 940 is used for two-way communication.
  • the TX/RX 940 has at least one antenna for facilitating communication.
  • the communication interface can represent any interface necessary to communicate with other devices.
  • program 930 includes program instructions that, when executed by associated processor 910, cause device 900 to perform the implementation of the present disclosure as discussed above with respect to Figures 3-8. That is, implementations of the present disclosure may be implemented by computer software executable by processor 910 of device 900, or by a combination of software and hardware.
  • the various example implementations of the disclosure can be implemented in hardware or special purpose circuits, software, logic, or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which can be executed by a controller, microprocessor or other computing device.
  • various examples (eg, methods, apparatus, or devices) of the present disclosure may be implemented in part or in whole on a computer readable medium.
  • the various aspects of the implementation of the present disclosure are illustrated or described as a block diagram, a flowchart, or some other graphical representation, it will be understood that the blocks, devices, systems, techniques, or methods described herein may be considered as non-limiting examples. Implemented in hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controllers or other computing devices, or some combination thereof.
  • implementations of the present disclosure may be described in the context of computer-executable instructions, such as in a program module that is executed in a device on a physical or virtual processor of a target.
  • program modules include routines, programs, libraries, objects, classes, components, data structures, and the like that perform particular tasks or implement particular abstract data structures.
  • the functionality of the program modules can be combined or divided between the described program modules.
  • Computer-executable instructions for program modules can be executed within a local or distributed device. In a distributed device, program modules can be located in both local and remote storage media.
  • Computer program code for implementing the methods of the present disclosure can be written in one or more programming languages.
  • the computer program code can be provided to a general purpose computer, a special purpose computer or a processor of other programmable data processing apparatus such that the program code, when executed by a computer or other programmable data processing apparatus, causes a flowchart and/or block diagram.
  • the functions/operations specified in are implemented.
  • the program code can execute entirely on the computer, partly on the computer, as a stand-alone software package, partly on the computer and partly on the remote computer or entirely on the remote computer or server.
  • a computer readable medium can be any tangible medium that contains or stores a program for use with or in connection with an instruction execution system, apparatus, or device.
  • the computer readable medium can be a machine readable signal medium or a machine readable storage medium.
  • the computer readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination thereof. More detailed examples of machine readable storage media include electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only Memory (EPROM or flash memory), optical storage device, magnetic storage device, or any suitable combination thereof.

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)

Abstract

根据本公开的示例实施例，提供了用于信道均衡的方法、设备和计算机可读介质。该方法包括在第一设备处，在通信信道上经由多个子载波从第二设备接收第一信号，并且对第一信号进行采样，以获得采样符号。该方法还包括利用采样符号与有效载荷之间的预定义的直接关联，基于所获得的采样符号来生成第二信号。第二信号指示第一信号在多个子载波中的有效子载波上承载的有效载荷。通过采样符号与有效载荷之间的直接关联的利用，可以以更简单、可靠且低成本的方式实现信道均衡，以提取接收信号中的有效载荷。

Description

用于信道均衡的方法、设备和计算机可读介质 技术领域
本公开的实施例主要涉及通信技术，并且更具体地，涉及用于信道均衡的方法、设备和计算机可读介质。
背景技术
在电信通信过程中，发送端在发送信号之前对信号执行相应的处理，诸如调制、编码等，并且经由通信信道将处理后的信号发送到接收端。在发送端的处理以及信号传输过程中，信号可能经受各种损害，从而带有噪声和/或失真。为了能够从接收到的信号中恢复出有效载荷数据，接收端通常会采用相应均衡算法来减小或降低这些损害。
在一些通信系统中，特别是在无线通信系统(具有极具价值的频谱资源)和骨干传送系统(诸如光纤通信系统，其特点在于在长距离传输中仍保持可靠性和容量方面的优势)中，采用信道建模和基于建模结果进行补偿的方案。虽然已经开发了非常准确信道模型用于描述信号经过的整个通信链路，但是这种方案在可靠性、计算复杂度以及成本方面仍然存在局限性。因此，期望能够提供更可靠且计算复杂度和成本更低的方案用于实现信道均衡。
发明内容
根据本公开的示例实施例，提供了一种用于信道均衡的方案。
在本公开的第一方面中，提供了一种用于信道均衡的方法。该方法包括在第一设备处，在通信信道上经由多个子载波从第二设备接收第一信号。该方法还包括对第一信号进行采样，以获得采样符号。该方法进一步包括利用采样符号与有效载荷之间的预定义的直 接关联，基于所获得的采样符号来生成第二信号，第二信号指示第一信号在多个子载波中的有效子载波上承载的有效载荷。
在本公开的第一方面中，提供了一种用于信道均衡的设备。该设备包括处理器；以及与处理器耦合的存储器，存储器具有存储于其中的指令。指令在被处理器执行时使设备：在通信信道上经由多个子载波从另一设备接收第一信号；对第一信号进行采样，以获得采样符号；以及利用采样符号与有效载荷之间的预定义的直接关联，基于所获得的采样符号来生成第二信号，第二信号指示第一信号在多个子载波中的有效子载波上承载的有效载荷。
在本公开的第三方面中，提供了一种计算机可读介质。计算机可读介质上存储有计算机可执行指令，计算机可执行指令在由一个或多个处理器执行时使一个或多个处理器执行前述第一方面的方法。
应当理解，发明内容部分中所描述的内容并非旨在限定本公开的实施例的关键或重要特征，亦非用于限制本公开的范围。本公开的其它特征将通过以下的描述变得容易理解。
附图说明
结合附图并参考以下详细说明，本公开各实施例的上述和其他特征、优点及方面将变得更加明显。在附图中，相同或相似的附图标记表示相同或相似的元素，其中：
图1是本公开描述的实施例可以在其中被实现的通信系统的示意图；
图2A和图2B示出了用于信道均衡的传统系统的示意图；
图3示出了根据本公开的一些实施例的用于信道均衡的系统的示意图；
图4示出了根据本公开的一些实施例的用于实现图3的一体化均衡器的信道均衡模型的示意图；
图5示出了根据本公开的一些实施例的用于信道均衡的过程的 示意图；
图6示出了根据本公开的一些实施例的用于确定直接关联的过程的示意图；
图7示出了根据本公开的一些实施例的训练图4的信道均衡模型的系统的示意图；
图8示出了根据本公开的一些实施例的用于信道均衡的过程的流程图；以及
图9示出了适合实现本公开的实施例的设备的简化框图。
在所有附图中，相同或相似参考数字表示相同或相似元素。
具体实施方式
下面将参照附图更详细地描述本公开的实施例。虽然附图中显示了本公开的某些实施例，然而应当理解的是，本公开可以通过各种形式来实现，而且不应该被解释为限于这里阐述的实施例，相反提供这些实施例是为了更加透彻和完整地理解本公开。应当理解的是，本公开的附图及实施例仅用于示例性作用，并非用于限制本公开的保护范围。
在本公开的实施例的描述中，术语“包括”及其类似用语应当理解为开放性包含，即“包括但不限于”。术语“基于”应当理解为“至少部分地基于”。术语“一个实施例”或“该实施例”应当理解为“至少一个实施例”。术语“第一”、“第二”等等可以指代不同的或相同的对象。下文还可能包括其他明确的和隐含的定义。
在此使用的术语“网络设备”是指在通信网络中具有特定功能的其他实体或节点。“基站”(BS)可以表示节点B(NodeB或者NB)、演进节点B(eNodeB或者eNB)、远程无线电单元(RRU)、射频头(RH)、远程无线电头端(RRH)、中继器、或者诸如微微基站、毫微微基站等的低功率节点等等。
在此使用的术语“终端设备”是指能够与网络设备之间或者彼此之间进行有线或无线通信的任何终端设备。作为示例，终端设备可 以包括飞行器、移动终端(MT)、订户台(SS)、便携式订户台(PSS)、移动台(MS)或者接入终端(AT)，以及车载的上述设备。终端设备可以是任意类型的移动终端、固定终端或便携式终端，包括移动手机、站点、单元、设备、多媒体计算机、多媒体平板、互联网节点、通信器、台式计算机、膝上型计算机、笔记本计算机、上网本计算机、平板计算机、个人通信系统(PCS)设备、个人导航设备、个人数字助理(PDA)、音频/视频播放器、数码相机/摄像机、定位设备、电视接收器、无线电广播接收器、电子书设备、游戏设备、物联网(IoT)设备或可用于通信的其他设备、或者上述的任意组合。
图1描述了本公开的实施例可以在其中被实现的通信系统100该通信系统100包括设备110和设备120，这两个设备可以互相通信，以传输各种业务数据、控制信息等等。在图1的示例中，设备110向设备120发送信号，由设备110发送的信号经由通信信道102传播到设备120。
在发送侧，设备110对待发送的有效载荷数据执行处理，这在图1中被图示为由信号处理模块112执行。信号处理模块112可以对待发送的数据执行各种处理，包括预处理、调制、频时变换、循环前缀的添加、放大等中的一项或多项处理。设备110还包括发射器114，用于将经处理的信号经由通信信道102发送至设备120。
在接收侧，设备120包括接收器122，用于接收通信信道102上传输的信号。由于设备110所发送的有效载荷数据在到达设备120时经历了设备110的信号处理、发射器的发射、链路传输以及接收器的接收，在这些过程中有可能经受各种损害，从而带有噪声、失真等。因此，设备120还包括均衡处理模块124，用于对接收器122接收到的信号进行均衡处理，以确定设备110发送的有效载荷。
通信系统100可以是有线通信系统或无线通信系统，并且因此通信信道102可以是有线通信链路或无线通信链路。有线通信系统的示例可以包括骨干传送网络，诸如光纤通信网络(有源光网络或无源光网络)、模拟前传通信网络等。无线通信系统的示例可以包 括任何无线电通信系统。在通信系统100中的通信可以根据任何适当的通信协议来实施，包括但不限于，第一代(1G)、第二代(2G)、第三代(3G)、第四代(4G)和第五代(5G)等蜂窝通信协议、诸如电气与电子工程师协会(IEEE)802.11等的无线局域网通信协议、和/或目前已知或者将来开发的任何其他协议。
设备110和120可以是在通信系统100中能够通信的任何设备。例如，在无线通信系统中，设备110可以是网络设备和终端设备(例如，UE)中的一个设备，而设备120可以是网络设备和终端设备中的另一个设备。在基于光纤的通信中，设备100可以是光线路终端(OLT)和光网络单元(ONU)中的一个，而设备120可以是OLT和ONU中的另一个。
应当理解，虽然图1中示出了由设备110向设备120发送信号，但在其他示例中，设备120也能够向设备110发送信号。在这些示例中，设备120可以具有以上参照图1描述的设备110的一些功能，而设备110可以具有以上参照图1描述的设备120的一些功能。
为了在接收侧处实现均衡处理，很多系统采用的是基于信道建模和基于建模结果进行补偿的方案。在这些方案中，采用级联模块来描述信号经过的整个通信链路，然后再采用级联模块来针对不同方面进行补偿。图2A和图2B分别示出了在无线通信系统和有线通信系统中的信道建模和补偿的均衡处理系统。
如图2A所示的均衡处理系统200中，均衡处理系统200包括无线信道建模模块210和补偿模块220。无线信道建模模块210包括多个子模块，用于对接收信号经历过的处理和信道进行建模。具体地，预处理子模块211用于建模发送端处的预处理，诸如信号在频域中的预处理；频时变换子模块212用于建模发送端处的频时变换，诸如逆快速傅里叶变换(iFFT)；发射天线子模块213用于建模在发射天线处经历的多输入多输出(MIMO)和/或功率泄露；空口子模块214用于建模在无线信道传播过程中经历的频域和时域损害；接收天线子模块215用于建模在接收端处的信号接收中引入的功率泄 露或相位偏移；采样和均衡子模块216用于建模通信过程中的符号间干扰(ISI)。根据需要，无线信道建模子模块210还可以包括更多其他子模块，用于对信号经历的其他处理进行建模。
基于无线信道建模模块210的建模结果，补偿模块220包括多个级联子模块，用于从不同方面对信号进行补偿，以实现信道均衡的目的。具体地，时钟恢复子模块221用于抑制ISI；去MIMO子模块222用于抑制载波间干扰(ICI)；前馈均衡子模块223用于实现时域处理；时频变换子模块224用于执行时频变换(例如，FFT)；并且迫零均衡子模块225用于实现线性步长。根据需要，补偿模块220还可以包括更多其他子模块，用于从其他方面来对信号进行补偿。
图2B示出的均衡处理系统202适于基于光纤通信的网络。均衡处理系统202包括光纤链路建模模块230包括多个子模块，用于对接收信号经历过的处理和信道进行建模。具体地，电-光调制子模块231用于建模发送端处从电学到光学的信号调制；光学放大子模块232用于建模发送端处的光学放大，以确定其中可能引入的非线性失真；光纤链路子模块233，用于建模光纤链路，以确定在光纤链路中可能经历的失真；光检测子模块234，用于建模接收端处的相位到功率的转换；并且采样和均衡子模块235用于建模通信过程中的符号间干扰(ISI)。根据需要，光纤链路建模子模块230还可以包括更多其他子模块，用于对信号经历的其他处理进行建模。
基于无线信道建模模块230的建模结果，补偿模块240包括多个级联子模块，用于从不同方面对信号进行补偿，以实现信道均衡的目的。具体地，时钟恢复子模块241用于实现重新对齐；N抽头均衡器242用于消除ISI；反向光检测子模块243用于确定反向响应；非线性管理子模块244用于确定反向光学放大响应；并且迫零均衡子模块245用于执行频域补偿。根据需要，补偿模块240还可以包括更多其他子模块，用于从其他方面来对信号进行补偿。
包括图2A和图2B所示的示例在内的常规方案在可靠性、复杂 度和成本方面均存在局限性。首先，在可靠性方面，由于接收信号中所有类型的损害可能是彼此耦合的，在一些特殊使用情况(诸如良好线性条件)下能够确保数据传输质量，从而确保均衡处理结果的可靠度或有效性。在此方面，进一步提出了将通信系统限制在小信号模型(SSM)下进行操作，以使得通信链路不会引入非线性问题。然而，基于SSM的方案以牺牲功率为代价，这在多点接入场景、诸如无源光网络系统或分布式天线系统前传网络中难以被接受，因为这些场景要求更高功率水平的发射信号以确保多个接收器均能接收到信号。
此外，常规方案对计算复杂度要求很高。为了确保建模和补偿的性能，可以采用更复杂的基于递归或迭代的算法(诸如对比特序列的最大似然估计等)。这将导致这些方案受限于处理设备的计算能力并且难以由低成本的处理设备(诸如数字信号处理单元)来实现。在这种情况下，难以被应用到对成本敏感的有线通信网络，光纤通信网络、模拟前传通信网络等。
在另外一些常规方案中，已经提出了基于人工智能(AI)技术来解决信道均衡的问题。基于AI的方案的优势在于能够在没有信道模型的先验知识的情况下优化后验概率估计，从而实现较高的性能。此外，在经过足够的学习之后，所获得的模型对于计算复杂度的要求降低，因此可以由成本较低的处理设备(诸如DSP)来实现。然而，已经提出的方案主要集中于对单载波格式化信号的信道均衡，因为可以在纯时域中对基于单载波格式化信号执行信道均衡。然而，还未提出针对基于多载波格式化信号的信道均衡。
根据本公开的实施例，提出了一种用于信道均衡的方案。该方案适于对基于多载波传输的信号执行信道均衡。在接收端处的设备在通信信道上经由多个子载波从另一设备接收信号，并且对接收到的信号进行采样，以获得采样符号。利用采样符号与有效载荷之间的预定义的直接关联，基于所获得的采样符号来生成另一信号，该信号指示接收到的在多个子载波中的有效子载波上承载的有效载 荷。通过预先定义(学习)采样符号与有效载荷之间的直接关联，可以避免由不同级联模块在不同方面实现信道均衡的情况中存在的非线性损害问题。此外，该直接关联的利用可以以更简单、可靠且低成本的方式实现信道均衡，以提取接收信号中的有效载荷。
如本文中所使用的，“信道均衡”指的是在接收端处对接收信号中被引入的噪声、失真等损害进行抑制或消除，以便获得发送端期望传输的有效载荷。
以下将参照附图来详细描述本公开的示例实施例。
图3示出了根据本公开的一些实施例的用于信道均衡的系统的示意图。该系统300可以被实现在图1中接收端处的设备120处，例如，被实现在均衡处理模块124中。为了讨论的目的，将参考图1来描述系统300。
系统300包括采样器310，用于从设备120的接收器122获取信号302(为便于讨论，也称为第一信号)。通常，第一信号302是频域信号。采样器310对信号302进行采样，以获得一系列的采样符号304。在本公开的实施例中，设备120的接收器122在通信信道102上经由多个子载波从设备110接收信号302。取决于针对多个子载波的处理，信号302可以是正交频分复用(OFDM)调制的信号、离散多载波(DMT)调制的信号等等。
信号302在被采样之前，在发送端处的设备110、通信信道102以及接收端处的设备120处均可能被引入一些噪声、失真等损害。例如，设备110可能期望向设备120发送有效载荷。为了有效传输，设备110可以对有效载荷执行一些处理，包括频域预处理、频时转换(诸如iFFT)、循环前缀(CP)添加、放大、天线发射等。经过这些处理，表示有效载荷的二进制数据被调制到相应的子载波，并且被变换为时域信号以在通信信道102中发送。设备110中的处理、通信信道102中的传播以及设备120处的信号接收都可能带来信号302中的损害。
系统300还包括一体化均衡器320，用于从采样器310获得采样 符号304，并且利用采样符号与有效载荷之间的预定义的直接关联，基于所获得的采样符号304来生成信号306(为便于讨论，称为第二信号)。信号306指示第一信号在多个子载波中的有效子载波上承载的有效载荷。有效载荷是设备110期望向设备120传输的数据。信号306可以被认为是频域信号。通过一体化均衡器320，可以实现对经历各种损害(诸如噪声、失真)的采样符号304执行信道均衡，从而提取出设备110发送的有效载荷。
根据本公开的实施例，通过采样符号与有效载荷之间的直接关联，可以用于将采样符号直接映射到有效载荷，而无需关心从时域的采样符号到频域的有效载荷的各种中间级联处理。相比于利用多个级联模块来建模通信信道和执行相应补偿，通过采样符号与有效载荷之间的直接关联，可以避免由不同级联模块在不同方面实现信道均衡的情况中存在的非线性损害问题。在均衡处理之前信号所经历的其他处理以及物理环境引发的损害可以被认为是“黑盒子”。通过一体化均衡可以直接实现从采样符号到期望的有效载荷的映射，而无需关于系统的先验知识和复杂的算法用于信道建模和补偿，甚至都无需执行信号的时频变换。通过这种方式，可以以更简单、可靠且低成本的方式实现信道均衡。因此，本公开的信道均衡方案还可以有利地被应用到对成本敏感的有线通信网络，诸如光纤通信网络、模拟前传通信网络等。
在一些实施例中，采样符号与有效载荷之间的直接关联可以从预先已知的采样符号以及有效载荷中学习得到。因此，在一些实现中，可以构建学习模型(被称为信道均衡模型)，并且利用已知的采样符号以及有效载荷来确定学习模型的参数集，以指示采样符号与有效载荷之间的直接关联。
图4示出了根据本公开的一些实施例的一体化均衡器320的示例架构。在图4的示例中，一体化均衡器320中包括信道均衡模型420，用于基于输入的采样符号来生成输出的有效子载波上的有效载荷。信道均衡模型420可以被构建为多层学习网络。如本文所使用 的，术语“学习网络”指的是这样的一个模型，该模型能够从训练数据中学习到相应的输入与输出之间的关联，从而在训练完成后基于训练得到的参数集对给定的输入进行处理以生成对应的输出。“学习网络”有时也可以被称为“神经网络”、“学习模型”、“网络”或“模型”。这些术语在本文中可互换地使用。
信道均衡模型420包括输入层422、输出层424以及输入层422与输出层444之间的一个或多个隐含层。信道均衡模型420中的每个层均包括多个节点426(也被称为神经元)，每个节点利用相应的激励函数对各自的输入进行处理，以确定输出。信道均衡模型420的多个层以层级结构布置，前一层的输出经由参数调整后作为下一层的输入。不同层的节点可以完全连接(即后一层的每个节点连接到前一层的全部节点)、不完全连接(即后一层的每个节点连接到前一层的部分节点)或一对一连接(即后一层的每个节点连接到前一层的一个节点)。相邻层中相连的节点存在信号传递。信道均衡模型420的参数集包括将输入的采用符号映射到输入层422的参数、信道均衡模型420中的各个相邻层之间的映射的参数、以及将输出层424的输出映射到有效载荷的参数。
在一些实施例中，信道均衡模型420的输入层422的节点数目(被表示为I)由设备110中的频时变换的参数和循环前缀的长度决定。例如，输入层422的节点数目可以等于频时变换的尺寸(例如，iFFT的长度)和循环前缀的长度之和。例如，如果发送端处的设备110中的频时变换的尺寸是1024，并且循环前缀的长度是176，则输入层的节点数目可以是1200个。
在一些实施例中，信道均衡模型420的输出层424的节点数目(被表示为K)由有效子载波的数目决定。通常，如果设备110和设备120被分配多个子载波进行通信，设备110可以选择其中的一些或全部子载波作为有效子载波，并且将有效载荷承载在这些有效子载波上。由于信道均衡模型420目标在于输出这些有效子载波上的有效载荷，因此输出层的节点数目可以与有效子载波的数目相同。 例如，如果频时变换的尺寸是1024，这意味着全部子载波的数目是1024。然而，设备110可能从全部1024个子载波中选择一部分子载波用于承载有效载荷(例如，600、800、1000或者其他数值)。通常，所确定的信道均衡模型420的输入层422的节点数目大于输出层424的节点数目，因此信道均衡模型420是具有宽的时域输入和窄的频域输出的模型。
由于设备110接收时域信号302，其可以被表示为电压或电流幅度波形，如图4所示。采样器310可以对时域信号302持续执行采样，以获得不同时间点的一系列采样符号304。一体化均衡器320还可以包括多个延迟单元410，用于对先前采样到的符号执行延迟。当采样器310采样到预定数目的采样符号304后，这些被延迟的采样符号被同时映射到输入层422。应当理解，虽然在图4的示例中，采样符号304的数目与输入层422的节点数目相同，但在其他示例中，采样符号304的数目可以大于或小于输入层422的节点数目，但是可以经由参数被映射到输入层422的相应节点。
如以上提及的，信道均衡模型420的参数集可以利用已知的采样符号以及有效载荷来确定，这涉及模型训练过程，并且将在以下详细描述。在信道均衡模型420的参数集已经确定的情况下，可以设备120可以利用信道均衡模型420来执行信道均衡，以从接收到的信号302中确定有效载荷。信道均衡模型420的输出信号306可以由每个输出节点的输出组成，并且每个输出节点的输出指示一个有效子载波上承载的有效载荷。如图4的示例所示，每个有效子载波上承载的有效载荷可以由复数值表示。
应当理解，虽然图1示出了信道均衡模型420的示例结构，但该示例仅为了解释说明的目的。在其他示例中，信道均衡模型420可以被构建为具有更多层、更多节点和/或其他结构。还应当理解，除了多层学习网络之外，信道均衡模型420还可以利用其它模型结构。本公开的实施例在此方面不受限制。
图5示出了根据本公开的一些实施例的用于信道均衡的过程500 的流程图。过程500涉及图1的设备110和设备120。过程500涉及利用已经确定的采样符号与有效载荷之间的直接关联来实现接收端处的信道均衡。
在505，设备110在通信信道上经由多个子载波向设备120发送信号302。信号302是时域信号。设备120利用接收器122接收信号302，并且通过采样器310对信号302进行采样，以获得采样符号304。设备120的一体化均衡器320然后利用预先定义的采样符号与有效载荷之间的直接关联，基于信号302的采样符号304来生成信号306，以指示有效子载波上承载的有效载荷。一体化均衡器320可以利用如图4所示的信道均衡模型420来生成信号306。信道均衡模型420中的参数集指示输入的采样符号与输出的有效载荷之间的直接关联。
以上描述了在采样符号与有效载荷之间的直接关联(例如，信道均衡模型420的参数集)已经确定的情况下由设备120实现信道均衡。下文将详细描述如何确定这样的直接关联。
在本公开的实施例中，确定采样符号与有效载荷之间的直接关联涉及从已知输入与输出之间学习两者的映射关系。学习的目标在于使得直接关联能够用于实现基于输入的采样符号而生成有效载荷。因此，确定直接关联的训练数据可以包括已知的采样符号以及这些采样符号对应的有效载荷。在通信系统中，由于不同设备之间的信道条件不同、设备的配置或设置参数不同、所采样的传输资源也不同，因此线下利用已知数据来确定直接关联可能难以适合所有场景下的信道均衡。在一些实施例中，可以在线确定采样符号与有效载荷之间的直接关联，从而使得所确定的直接关联更适合于从设备110到设备120传输的信号的信道均衡。
接下来将参照图6和图7来描述对采样符号与有效载荷之间的直接关联的在线确定的示例。图6示出了根据本公开的一些实施例的用于确定直接关联的过程600的流程图。过程600涉及图1的设备110和设备120。
在605，设备110在通信信道102上经由多个子载波向设备120发送训练信号。训练信号用于确定采样符号与有效载荷之间的直接关联的目的。在一些实施例中，设备110以与后续正常通信相同的方式对训练信号的有效载荷进行处理，以相同方式发送。此外，用于训练信号的传输的通信信道102和多个子载波也是设备110与设备120的后续正常通信中将被采用的信道和子载波。通过这种方式，可以更好地针对设备110与设备120的通信来确定用于信道均衡的直接关联。
在610，设备110还可以向设备120发送参考信号，参考信号指示训练信号在有效子载波上承载的有效载荷。参考信号也可由设备120用于确定直接关联。相较于训练信号，设备110不对参考信号执行额外处理，诸如频域预处理、频时变换、循环前缀添加、放大等等，而是直接将有效载荷发送至设备120。在一些实施例中，设备110可以直接将参考信号发送给设备120。在另外一些实施例中，设备110还可以经由其他设备来将参考信号发送给设备120。在图6的示例中描述了参考信号由设备110发送至设备120。然而，在其他示例中，参考信号也可以被预先配置在设备120中。在这些示例中，过程600中的步骤610可以省略。
设备120接收训练信号和参考信号。在615，设备120对接收到的训练信号进行采样，以获得训练采样信号。在620，设备120基于训练采样符号和参考信号来确定直接关联。设备120可以利用多种训练算法来确定直接关联。直接关联的确定可以取决于所构建的用于表示直接关联的模型。在图4的信道均衡模型420的训练过程，直接关联的确定可以由多种机器学习算法来实现，诸如随机梯度下降算法、后向传播算法等。
在图4的信道均衡模型420示例中，由于信道均衡模型420的输入层的节点数目与设备110中的频时变换的参数和循环前缀的长度有关，并且输出层的节点数目与设备110选择用于有效载荷传输的有效子载波的数目有关，在一些实施例中，设备110还从设备120 接收指示频时变换的参数和循环前缀的长度的信息(被称为第一信息)和指示有效子载波的数目的信息(被称为第二信息)。在一些实施例中，设备110可能采用的频时变换的参数、循环前缀的长度以及有效子载波的数目可以由被预先配置到设备120中，而无需从设备110接收。
在一些实施例中，如果已经确定采样符号与有效载荷之间的直接关联，设备120还可以在625向设备110发送确定指示，以指示直接关联已经被确定。设备110接收到这样的确认指示之后，可以发起与设备120的正常传输。
图7示出了用于训练图4的信道均衡模型420的系统700的示意图。在系统700中，信道均衡模型420的架构已经被构建，包括模型的层数、每个层的节点数以及层与层之间的节点连接。然后，可以通过训练过程、诸如图6所述的过程600来确定这个模型的参数集。
具体地，在初始阶段，信道均衡模型420的参数集被初始化。设备110接收到的训练信号702被采样(例如，由采样器310采样)，以获得训练采样符号。由于训练信号702是时域信号，较早时间点处的训练采样符号被各个延迟单元410延迟。当采样得到相应数目的训练采样符号被输入到信道均衡模型420的输入层422。训练采样符号经由信道均衡模型420的当前参数集映射到输出层424，以获得当前的有效载荷输出。系统700包括训练库710，其中存储与训练信号702对应的参考信号706(即，相应有效子载波上的有效载荷)。系统700还包括多个传递元件720。训练库710将参考信号706指示的有效载荷提供给相应的传递元件720。
系统700中的多个误差计算元件730用于计算输出层424的每个节点输出的有效载荷与传递元件720获得的参考有效载荷，以确定两者之间的误差(被表示为e1至eK，其中K表示输出层424的节点数目)。这些误差由加和单元740加和，以计算误差总和。该误差总和可以用于调整信道均衡模型420的参数集，例如不同层之间 的映射参数、可能的偏置参数等。系统700可以包括参数调整单元(未示出)，用于向信道均衡模型420的相应参数施加调整(诸如，调制1、调整2、......调整n，n表示相邻两层之间的映射的数目)。
在训练过程中，可以将多个训练信号不断输入到信道均衡模型420，以便使得信道均衡模型420的参数集通过上述过程不断优化。在一些实施例中，可以定义一个损失函数(例如，交叉熵函数)，该函数可以与加和单元740计算的误差总和相关。通过参数集的调整，可以不断减小损失函数的值。当损失函数达到最小化时，可以认为信道均衡模型420已经收敛，并且当前的参数集可以用于指示采样符号与有效载荷之间的直接关联。
应当理解，图7仅示出了用于训练信道均衡模型420的一个具体示例系统，该系统中的一些元件仅在训练阶段使用，并且在使用阶段可以隐去。在其他实施例中，还可以采用其他训练或学习方法来确定信道均衡模型420的参数集。
图8示出了根据本公开的一些实施例的用于信道均衡的过程800的流程图。可以理解，过程800可以例如在如图1所示的设备120处实施，并且可以由图3的系统300来实现。为描述方便，下面结合图1和图3对过程800进行说明。
在810，设备120(也称为第一设备)的接收器122在通信信道上经由多个子载波从设备110(也称为第二设备)接收信号302(也称为第一信号)。在820，设备120的采样器310对第一信号302进行采样，以获得采样符号304。在830，设备120的一体化均衡器320利用采样符号与有效载荷之间的预定义的直接关联，基于所获得的采样符号304来生成信号306(也称为第二信号)。第二信号306指示第一信号302在多个子载波中的有效子载波上承载的有效载荷。
在一些实施例中，过程800还可以包括在通信信道102上经由多个子载波从第二设备110接收训练信号；对训练信号进行采样，以获得训练采样符号；获取参考信号，参考信号指示训练信号在有效子载波上承载的有效载荷；以及基于训练采样符号和参考信号来 确定直接关联。
在一些实施例中，确定直接关联可以包括：基于训练采样符号和参考信号来确定信道均衡模型的参数集，参数集指示直接关联。
在一些实施例中，确定直接关联可以进一步包括：从第二设备110接收第一信息，第一信息指示第二设备110中的频时变换的参数和循环前缀的长度；以及基于频时变换的参数和循环前缀的长度来确定信道均衡模型中的输入层的节点数目。
在一些实施例中，确定直接关联可以进一步包括：从第二设备110接收第二信息，第二信息指示有效子载波的数目；以及基于有效子载波的数目来确定信道均衡模型中的输出层的节点数目。
在一些实施例中，获取参考信号可以包括：从第二设备110接收参考信号。
在一些实施例中，过程800还可以包括：向第二设备110发送确认指示，以指示直接关联已经被确定。
在一些实施例中，通信信道102可以包括有线通信信道或无线通信信道。
图9示出了适合实现本公开的实现的设备900的简化框图。设备900可以用来实现通信设备，例如图1所示的设备120。如所示出的，设备900包括一个或多个处理器910，耦合到(多个)处理器910的一个或多个存储器920，耦合到处理器910的一个或多个发射器和/或接收器(TX/RX)940。
处理器910可以具有适合于本地技术环境的任何类型，并且作为非限制性示例可以包括以下一个或多个：通用计算机、专用计算机、微处理器、数字信号处理器(DSP)和基于多核处理器架构的处理器。设备900可以具有多个处理器，诸如在时间上跟随与主处理器同步的时钟进行从动的专用集成电路芯片。
存储器920可以具有适合于本地技术环境的任何类型并且可以使用任何适合的数据存储技术来实施，作为非限制性示例，诸如非暂态计算机可读存储介质、基于半导体的存储设备、磁存储器设备 和系统、光存储器设备和系统、固定存储器和可移除存储器。
存储器920存储程序930的至少一部分。TX/RX 940用于双向通信。TX/RX 940具有至少一个天线，用于促进通信。通信接口可以表示与其他设备通信必要的任何接口。
假定程序930包括程序指令，这些程序指令在由相关联的处理器910执行时，使设备900执行如以上参照图3至图8所讨论的本公开的实现。也就是说，本公开的实现可以由设备900的处理器910可执行的计算机软件来实现，或者由软件与硬件的组合来实现。
一般而言，本公开的各种示例实现可以在硬件或专用电路、软件、逻辑，或其任何组合中实施。某些方面可以在硬件中实施，而其他方面可以在可以由控制器、微处理器或其他计算设备执行的固件或软件中实施。例如，在一些实现中，本公开的各种示例(例如方法、装置或设备)可以部分或者全部被实现在计算机可读介质上。当本公开的实现的各方面被图示或描述为框图、流程图或使用某些其他图形表示时，将理解此处描述的方框、装置、系统、技术或方法可以作为非限制性的示例在硬件、软件、固件、专用电路或逻辑、通用硬件或控制器或其他计算设备，或其某些组合中实施。
作为示例，本公开的实现可以在计算机可执行指令的上下文中被描述，计算机可执行指令诸如包括在目标的物理或者虚拟处理器上的器件中执行的程序模块中。一般而言，程序模块包括例程、程序、库、对象、类、组件、数据结构等，其执行特定的任务或者实现特定的抽象数据结构。在各实现中，程序模块的功能可以在所描述的程序模块之间合并或者分割。用于程序模块的计算机可执行指令可以在本地或者分布式设备内执行。在分布式设备中，程序模块可以位于本地和远程存储介质二者中。
用于实现本公开的方法的计算机程序代码可以用一种或多种编程语言编写。这些计算机程序代码可以提供给通用计算机、专用计算机或其他可编程的数据处理装置的处理器，使得程序代码在被计算机或其他可编程的数据处理装置执行的时候，引起在流程图和/或 框图中规定的功能/操作被实施。程序代码可以完全在计算机上、部分在计算机上、作为独立的软件包、部分在计算机上且部分在远程计算机上或完全在远程计算机或服务器上执行。
在本公开的上下文中，计算机可读介质可以是包含或存储用于或有关于指令执行系统、装置或设备的程序的任何有形介质。计算机可读介质可以是机器可读信号介质或机器可读存储介质。计算机可读介质可以包括但不限于电子的、磁的、光学的、电磁的、红外的或半导体系统、装置或设备，或其任意合适的组合。机器可读存储介质的更详细示例包括带有一根或多根导线的电气连接、便携式计算机磁盘、硬盘、随机存储存取器(RAM)、只读存储器(ROM)、可擦除可编程只读存储器(EPROM或闪存)、光存储设备、磁存储设备，或其任意合适的组合。
另外，尽管操作以特定顺序被描绘，但这并不应该理解为要求此类操作以示出的特定顺序或以相继顺序完成，或者执行所有图示的操作以获取期望结果。在某些情况下，多任务或并行处理会是有益的。同样地，尽管上述讨论包含了某些特定的实施细节，但这并不应解释为限制任何发明或权利要求的范围，而应解释为对可以针对特定发明的特定实现的描述。本说明书中在分开的实现的上下文中描述的某些特征也可以整合实施在单个实现中。反之，在单个实现的上下文中描述的各种特征也可以分离地在多个实现或在任意合适的子组合中实施。
尽管已经以特定于结构特征和/或方法动作的语言描述了主题，但是应当理解，所附权利要求中限定的主题并不限于上文描述的特定特征或动作。相反，上文描述的特定特征和动作是作为实现权利要求的示例形式而被公开的。

Claims (17)

  1. 一种用于信道均衡的方法，包括：
    在第一设备处，在通信信道上经由多个子载波从第二设备接收第一信号；
    对所述第一信号进行采样，以获得采样符号；以及
    利用采样符号与有效载荷之间的预定义的直接关联，基于所获得的采样符号来生成第二信号，所述第二信号指示所述第一信号在所述多个子载波中的有效子载波上承载的有效载荷。
  2. 根据权利要求1所述的方法，进一步包括：
    在所述通信信道上经由所述多个子载波从所述第二设备接收训练信号；
    对所述训练信号进行采样，以获得训练采样符号；
    获取参考信号，所述参考信号指示所述训练信号在所述有效子载波上承载的有效载荷；以及
    基于所述训练采样符号和所述参考信号来确定所述直接关联。
  3. 根据权利要求2所述的方法，其中确定所述直接关联包括：
    基于所述训练采样符号和所述参考信号来确定信道均衡模型的参数集，所述参数集指示所述直接关联。
  4. 根据权利要求3所述的方法，其中确定所述直接关联进一步包括：
    从所述第二设备接收第一信息，所述第一信息指示所述第二设备中的频时变换的参数和循环前缀的长度；以及
    基于所述频时变换的参数和所述循环前缀的长度来确定所述信道均衡模型中的输入层的节点数目。
  5. 根据权利要求3所述的方法，其中确定所述直接关联进一步包括：
    从所述第二设备接收第二信息，所述第二信息指示所述有效子载波的数目；以及
    基于所述有效子载波的数目来确定所述信道均衡模型中的输出层的节点数目。
  6. 根据权利要求2所述的方法，其中获取所述参考信号包括：
    从所述第二设备接收所述参考信号。
  7. 根据权利要求2所述的方法，进一步包括：
    向所述第二设备发送确认指示，以指示所述直接关联已经被确定。
  8. 根据权利要求1所述的方法，其中所述通信信道包括有线通信信道或无线通信信道。
  9. 一种用于信道均衡的设备，包括：
    处理器；以及
    与所述处理器耦合的存储器，所述存储器具有存储于其中的指令，所述指令在被处理器执行时使所述设备：
    在通信信道上经由多个子载波从另一设备接收第一信号；
    对所述第一信号进行采样，以获得采样符号；以及
    利用采样符号与有效载荷之间的预定义的直接关联，基于所获得的采样符号来生成第二信号，所述第二信号指示所述第一信号在所述多个子载波中的有效子载波上承载的有效载荷。
  10. 根据权利要求9所述的设备，进一步包括：
    在所述通信信道上经由所述多个子载波从所述另一设备接收训练信号；
    对所述训练信号进行采样，以获得训练采样符号；
    获取参考信号，所述参考信号指示所述训练信号在所述有效子载波上承载的有效载荷；以及
    基于所述训练采样符号和所述参考信号来确定所述直接关联。
  11. 根据权利要求10所述的设备，其中确定所述直接关联包括：
    基于所述训练采样符号和所述参考信号来确定信道均衡模型的参数集，所述参数集指示所述直接关联。
  12. 根据权利要求11所述的设备，其中确定所述直接关联进一 步包括：
    从所述另一设备接收第一信息，所述第一信息指示所述另一设备中的频时变换的参数和循环前缀的长度；以及
    基于所述频时变换的参数和所述循环前缀的长度来确定所述信道均衡模型中的输入层的节点数目。
  13. 根据权利要求11所述的设备，其中确定所述直接关联进一步包括：
    从所述另一设备接收第二信息，所述第二信息指示所述有效子载波的数目；以及
    基于所述有效子载波的数目来确定所述信道均衡模型中的输出层的节点数目。
  14. 根据权利要求10所述的设备，其中获取所述参考信号包括：
    从所述另一设备接收所述参考信号。
  15. 根据权利要求10所述的设备，进一步包括：
    向所述另一设备发送确认指示，以指示所述直接关联已经被确定。
  16. 根据权利要求9所述的设备，其中所述通信信道包括有线通信信道或无线通信信道。
  17. 一种计算机可读介质，其上存储有计算机可执行指令，所述计算机可执行指令在由一个或多个处理器执行时使所述一个或多个处理器执行根据权利要求1至8中任一项所述的方法的步骤。
PCT/CN2019/076291 2018-02-27 2019-02-27 用于信道均衡的方法、设备和计算机可读介质 WO2019165969A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/975,912 US11516053B2 (en) 2018-02-27 2019-02-27 Method and device for channel equalization, and computer-readable medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810162566.2A CN110198282B (zh) 2018-02-27 2018-02-27 用于信道均衡的方法、设备和计算机可读介质
CN201810162566.2 2018-02-27

Publications (1)

Publication Number Publication Date
WO2019165969A1 true WO2019165969A1 (zh) 2019-09-06

Family

ID=67750855

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/076291 WO2019165969A1 (zh) 2018-02-27 2019-02-27 用于信道均衡的方法、设备和计算机可读介质

Country Status (3)

Country Link
US (1) US11516053B2 (zh)
CN (1) CN110198282B (zh)
WO (1) WO2019165969A1 (zh)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11284277B2 (en) * 2018-11-07 2022-03-22 DeepSig Inc. Communications and measurement systems for characterizing radio propagation channels
EP3959848A4 (en) 2019-04-23 2022-06-22 Deepsig Inc. COMMUNICATION SIGNAL PROCESSING BY MEANS OF A MACHINE LEARNING NETWORK
CN113141216B (zh) * 2020-01-20 2022-11-15 上海诺基亚贝尔股份有限公司 信号处理方法、设备、装置和计算机可读存储介质
CN116458103A (zh) * 2020-12-31 2023-07-18 华为技术有限公司 一种神经网络的训练方法以及相关装置
EP4282078A1 (en) * 2021-01-25 2023-11-29 Marvell Asia Pte, Ltd. Ethernet physical layer transceiver with non-linear neural network equalizers
WO2024007109A1 (en) * 2022-07-04 2024-01-11 Nokia Shanghai Bell Co., Ltd. Apparatus, method and computer program

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2001067697A2 (en) * 2000-03-08 2001-09-13 Telefonaktiebolaget L M Ericsson (Publ) Technique for efficiently equalizing a transmission channel in a data transmission system
CN101312443A (zh) * 2007-05-24 2008-11-26 中国科学院微电子研究所 一种用于正交频分复用通信均衡与解调的系统及方法
US7995665B2 (en) * 2006-06-26 2011-08-09 Ralink Technology (Singapore) Corporation Pte. Ltd. Method and apparatus for reception in a multi-input-multi-output (MIMO) orthogonal frequency domain modulation (OFDM) wireless communication system
CN105553898A (zh) * 2015-12-18 2016-05-04 中国电子科技集团公司第三研究所 均衡器及反馈均衡方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7254171B2 (en) * 2000-01-20 2007-08-07 Nortel Networks Limited Equaliser for digital communications systems and method of equalisation
CN101026433B (zh) * 2006-02-24 2010-08-25 上海无线通信研究中心 一种用于自适应调制编码的信噪比估算方法
GB0721424D0 (en) * 2007-10-31 2007-12-12 Icera Inc A radio receiver in a wireless communications system
CN102487370B (zh) * 2010-12-02 2015-05-13 无锡物联网产业研究院 一种传输模式调整方法、传输模式控制器和传输设备
GB2514174B (en) * 2013-05-17 2015-12-02 Cambium Networks Ltd Improvements to adaptive modulation
CN105227512B (zh) * 2015-10-19 2018-03-27 宁波大学 一种ofdm水声通信系统中的脉冲噪声估计方法
US10171272B2 (en) * 2017-04-28 2019-01-01 Intel Corporation Computationally efficient algorithm for mitigating phase noise in OFDM receivers

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2001067697A2 (en) * 2000-03-08 2001-09-13 Telefonaktiebolaget L M Ericsson (Publ) Technique for efficiently equalizing a transmission channel in a data transmission system
US7995665B2 (en) * 2006-06-26 2011-08-09 Ralink Technology (Singapore) Corporation Pte. Ltd. Method and apparatus for reception in a multi-input-multi-output (MIMO) orthogonal frequency domain modulation (OFDM) wireless communication system
CN101312443A (zh) * 2007-05-24 2008-11-26 中国科学院微电子研究所 一种用于正交频分复用通信均衡与解调的系统及方法
CN105553898A (zh) * 2015-12-18 2016-05-04 中国电子科技集团公司第三研究所 均衡器及反馈均衡方法

Also Published As

Publication number Publication date
CN110198282B (zh) 2020-06-23
US20210091979A1 (en) 2021-03-25
CN110198282A (zh) 2019-09-03
US11516053B2 (en) 2022-11-29

Similar Documents

Publication Publication Date Title
WO2019165969A1 (zh) 用于信道均衡的方法、设备和计算机可读介质
US20210105548A1 (en) Method and device for communication in passive optical network, and computer-readable medium
KR101241824B1 (ko) Ｏｆｄｍ 통신 시스템의 수신 장치 및 그의 위상 잡음 완화 방법
US20130230092A1 (en) Sparse and reconfigurable floating tap feed forward equalization
US10164803B2 (en) Method and apparatus for controlling interference in QAM-FBMC system
CN110611628B (zh) 信号处理方法、网络设备及计算机可读存储介质
WO2016195328A1 (ko) 무선 통신 시스템에서의 필터 뱅크 다중 반송파 심벌들을 검출하는 장치 및 방법
CN115378769B (zh) 数据传输方法、装置、通信设备及存储介质
WO2015131501A1 (zh) 一种相干光通信信道估计方法和系统
WO2017005161A1 (zh) 功率分配方法和装置
EP3090519B1 (en) Methods and devices for doppler shift compensation in a mobile communication system
WO2022062904A1 (zh) 参考信号传输方法、装置、通信节点及存储介质
CN108880691B (zh) 用于解调信号的方法、设备及计算机可读介质
US9667449B1 (en) Channel estimator, demodulator and method for channel estimation
CN108173611A (zh) 一种基于ofdm体制卫星转发器的evm测试优化方法
US11050449B1 (en) System and method for extensionless adaptive transmitter and receiver windowing
Güven et al. CNN-aided channel and carrier frequency offset estimation for HAPS-LEO links
CN109792426B (zh) 使空循环前缀适合于频域空单载波通信系统的方法
CN113630667B (zh) 用于光通信的方法、设备、装置和计算机可读介质
WO2011035592A1 (zh) 一种正交频分复用接入系统的信道估计方法和装置
WO2024135883A1 (ko) 증폭기의 비선형성 보상을 위한 다중 파일럿 운용 방법 및 장치
CN116488969B (zh) 信道均衡方法、装置、设备和存储介质
CN112311714B (zh) 数据帧传输方法、装置、电子设备和计算机可读介质
KR101346282B1 (ko) 단일 반송파 ｍｉｍｏ 시스템 기반의 pn 부호열을 이용한 반송파 주파수 오차 추정 방법
WO2023272653A1 (en) Channel estimation compensation with constellation

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19761569

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19761569

Country of ref document: EP

Kind code of ref document: A1