CN116488979A

CN116488979A - Frequency offset estimation method and related device

Info

Publication number: CN116488979A
Application number: CN202210047782.9A
Authority: CN
Inventors: 吴虹; 陈琢; 耿雪; 赵迎新; 张天虹; 黎超
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2022-01-17
Filing date: 2022-01-17
Publication date: 2023-07-25

Abstract

The application provides a frequency offset estimation method and a related device, wherein the method comprises the following steps: receiving first information from a second device; according to the symbol where the first information is located, carrying out N times of cyclic prediction to obtain a first frequency offset estimation value; data demodulation is carried out according to the first frequency offset estimation value; the ith cyclic prediction in the N cyclic predictions comprises the following steps: performing frequency offset compensation on the symbol where the first information is located in the i-1 th cyclic prediction according to the frequency offset estimation value obtained in the i-1 th cyclic prediction, so as to obtain the symbol where the first information is located in the i-1 th cyclic prediction; and predicting to obtain a frequency offset estimation value during the ith cyclic prediction according to the symbol of the first information during the ith cyclic prediction. By implementing the embodiment of the application, the demodulation performance is improved.

Description

Frequency offset estimation method and related device

Technical Field

The present disclosure relates to the field of communications technologies, and in particular, to a frequency offset estimation method and a related device.

Background

Along with the rapid development of social economy and the continuous penetration of informatization process, the demands of people on communication are increasingly improved, and a long-term evolution (long term evolution, LTE) system adopts an orthogonal frequency division multiplexing and multiple input multiple Output (OFDM) technology as a core technology, so that the system has the advantages of high-efficiency spectrum efficiency, robustness under a fading channel, simple receiver structure and the like, can meet the requirements of users on low time delay, large capacity, wide coverage, high speed and the like, and becomes a mobile communication standard widely used by people at present.

Synchronization is a problem faced by any communication system, the performance of which will have a serious impact on the overall communication system, and efficient synchronization techniques are the preconditions for reliable data transmission by the communication system, especially for LTE systems. The orthogonal frequency division multiplexing (orthogonal frequency division multiplexing, OFDM) technology adopted by LTE is a multi-carrier modulation technology with higher spectral efficiency, and modulates a high-speed serial data stream to a plurality of parallel sub-carriers with mutually orthogonal transmission rates and transmits the serial data stream simultaneously, if the multi-carriers cannot keep strict orthogonality, inter-sub-carrier interference (inter carrier interference, ICI) is introduced after demodulation of a receiving end, so that performance is reduced. In the system transmitter and receiver, the frequency of the local oscillator generating the carrier is not synchronous, and the frequency spectrum offset in the channel, such as Doppler frequency shift generated by the rapid movement of the communication terminal, can cause the destruction of the orthogonality of the sub-carriers, thereby causing ICI, so the OFDM system is very sensitive to frequency deviation. Accurate carrier synchronization techniques are therefore very important for LTE systems.

In general, frequency offset estimation may be performed using conventional synchronization algorithms or by means of deep learning. However, no matter the frequency offset estimation is performed by adopting a traditional synchronization algorithm or the existing frequency offset estimation is performed by a deep learning mode, a frame structure needs to be designed artificially. For example, a traditional synchronization algorithm needs to perform frequency offset estimation based on artificially designed reference signals; when the existing frequency offset estimation is performed based on a deep learning mode, an artificially designed reference signal is also required to be input into a trained model for performing the frequency offset estimation. It will be appreciated that because the reference signal needs to be designed artificially, the frame structure will be changed. Therefore, the subjective color of the current frequency offset estimation mode is too strong, which is likely to cause insufficient accuracy of frequency offset estimation and further cause low demodulation performance.

Disclosure of Invention

The frequency offset estimation method and the related device provided by the application avoid the over-strong subjective color of the frequency offset estimation mode, so that the frequency offset estimation is more accurate, and the demodulation performance is further improved.

In a first aspect, a method for estimating frequency offset is provided, where the method is applied to a first device, and the method includes: receiving first information from a second device; the first information includes at least one of: a primary synchronization signal, a secondary synchronization signal, and service data; according to the symbol where the first information is located, carrying out N times of cyclic prediction to obtain a first frequency offset estimation value; n is an integer greater than or equal to 1; data demodulation is carried out according to the first frequency offset estimation value; the ith cyclic prediction in the N cyclic predictions comprises the following steps that i is an integer greater than or equal to 1 and less than or equal to N: performing frequency offset compensation on the symbol where the first information is located in the i-1 th cyclic prediction according to the frequency offset estimation value obtained in the i-1 th cyclic prediction, so as to obtain the symbol where the first information is located in the i-1 th cyclic prediction; and predicting to obtain a frequency offset estimation value during the ith cyclic prediction according to the symbol of the first information during the ith cyclic prediction. That is, it can be seen that, by receiving at least one of the primary synchronization signal, the secondary synchronization signal and the service data from the second device, the first frequency offset estimation value can be obtained by performing N-time cyclic prediction according to the location of at least one of the primary synchronization signal, the secondary synchronization signal and the service data, so that the problem that the frequency offset estimation is not accurate enough due to the excessively strong subjective color when the frequency offset estimation is performed according to the artificially designed reference signal is avoided, and the frame structure is also avoided being changed. Namely, under the condition of not changing the frame structure, the accuracy of frequency offset estimation is improved, and the demodulation performance is further improved. In addition, by means of repeated cyclic prediction, the frequency offset is compensated by adopting a cyclic iteration method, the sensitivity of the algorithm to high frequency offset is weakened, and the estimation performance of the algorithm is further improved through repeated iteration.

Optionally, with reference to the first aspect, if i is equal to 1, the frequency offset estimation value in the ith cyclic prediction is an integer multiple of the frequency offset estimation value. Namely, it can be seen that the frequency offset estimation value in the 1 st cycle prediction is an integer multiple frequency offset estimation value, so that the frequency offset range which can be estimated is greatly widened.

Optionally, with reference to the first aspect, N is a maximum number of cycles; or the first frequency offset estimation value is smaller than a preset first threshold value; or, the difference between the frequency offset estimation value in the i-1 th cycle prediction and the frequency offset estimation value in the i-1 th cycle prediction is smaller than a preset second threshold value. By providing the condition for stopping cyclic prediction, the problem of overlarge energy consumption caused by continuous operation is avoided, namely, the energy consumption is saved.

Optionally, with reference to the first aspect, the method further includes: and sending capability information of the first device to the second device, wherein the capability information of the first device is used for indicating that the first device supports integral multiple frequency offset estimation. That is, it can be seen that by sending the capability information of the first device to the second device, the second device can learn that the first device supports frequency offset estimation of integer multiple, and further can issue at least one of a primary synchronization signal, a secondary synchronization signal and service data.

Optionally, with reference to the first aspect, the frequency offset estimation method is performed by a network model, where the network model is obtained by training the following steps: acquiring a symbol where the second information is located; the second information includes at least one of: a primary synchronization signal, a secondary synchronization signal, and service data; vectorizing the symbol where the second information is located to obtain a feature vector; inputting the feature vector into a network model, and predicting to obtain a frequency offset estimation value corresponding to the feature vector; and adjusting model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector so as to train the network model. That is, it can be seen that the two-dimensional feature vector is obtained by vectorizing the symbol where the second information is located, so that the data analysis problem is converted into the image recognition problem, that is, the two-dimensional feature vector corresponds to the two-dimensional data in the image recognition. Meanwhile, the feature vector is input into the network model, and the frequency offset estimation value corresponding to the feature vector is obtained through prediction, which is equivalent to extracting the image features by using the neural network, outputting the frequency offset estimation value, and reducing the algorithm calculation complexity. In addition, since the second information includes at least one of: the main synchronization signal, the auxiliary synchronization signal and the service data avoid the problem that the frequency offset estimation is not accurate enough due to the over-strong subjective color when the training is carried out according to the artificially designed reference signal, and also avoid the change of the frame structure.

Optionally, with reference to the first aspect, adjusting model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector, so as to train the network model, including: obtaining a first loss according to the frequency offset estimation value corresponding to the feature vector and the pre-marked frequency deviation corresponding to the feature vector; and according to the first loss, adjusting model parameters of the network model to train the network model. Therefore, in the technical scheme, training of the network model is realized.

Optionally, in combination with the first aspect, the first loss satisfies the following formula:

wherein loss is a first loss, R is an integer greater than or equal to 1 and less than or equal to a preset training frequency, t _i For the corresponding pre-labeled frequency deviation of the feature vector, y _i And the frequency offset estimation value corresponding to the feature vector is obtained. It can be seen that in the above technical solution, the adjustment of the network model by the loss function is realized.

In a second aspect, a method for estimating frequency offset is provided, the method comprising: transmitting first information to a first device; the first information includes at least one of: primary synchronization signal, secondary synchronization signal, service data. That is, the first device may be caused to acquire the first information.

Optionally, with reference to the second aspect, the method further includes: capability information of a first device is received from the first device, the capability information of the first device being used to indicate that the first device supports integer multiple frequency offset estimation. That is, the second device may be enabled to learn that the first device supports frequency offset estimation of integer multiples, and may further issue at least one of a primary synchronization signal, a secondary synchronization signal, and service data.

In a third aspect, a communication apparatus is provided, the communication apparatus being a first device, the first device comprising a transceiver module and a processing module, the transceiver module being configured to receive first information from a second device; the first information includes at least one of: a primary synchronization signal, a secondary synchronization signal, and service data; a processing module for:

according to the symbol where the first information is located, carrying out N times of cyclic prediction to obtain a first frequency offset estimation value; n is an integer greater than or equal to 1; data demodulation is carried out according to the first frequency offset estimation value; the ith cyclic prediction in the N cyclic predictions comprises the following steps that i is an integer greater than or equal to 1 and less than or equal to N: performing frequency offset compensation on the symbol where the first information is located in the i-1 th cyclic prediction according to the frequency offset estimation value obtained in the i-1 th cyclic prediction, so as to obtain the symbol where the first information is located in the i-1 th cyclic prediction; and predicting to obtain a frequency offset estimation value during the ith cyclic prediction according to the symbol of the first information during the ith cyclic prediction.

Optionally, with reference to the third aspect, if i is equal to 1, the frequency offset estimation value in the ith cyclic prediction is an integer multiple of the frequency offset estimation value.

Optionally, with reference to the third aspect, N is a maximum number of cycles; or the first frequency offset estimation value is smaller than a preset first threshold value; or, the difference between the frequency offset estimation value in the i-1 th cycle prediction and the frequency offset estimation value in the i-1 th cycle prediction is smaller than a preset second threshold value.

Optionally, with reference to the third aspect, the transceiver module is further configured to send capability information of the first device to the second device, where the capability information of the first device is used to indicate that the first device supports frequency offset estimation with integer multiple.

Optionally, with reference to the third aspect, the processing module is further configured to:

acquiring a symbol where the second information is located; vectorizing the symbol where the second information is located to obtain a feature vector; inputting the feature vector into a network model, and predicting to obtain a frequency offset estimation value corresponding to the feature vector; and adjusting model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector so as to train the network model.

Optionally, with reference to the third aspect, when adjusting model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector to train the network model, the processing module is configured to:

Obtaining a first loss according to the frequency offset estimation value corresponding to the feature vector and the pre-marked frequency deviation corresponding to the feature vector; and according to the first loss, adjusting model parameters of the network model to train the network model.

Optionally, in combination with the third aspect, the first loss satisfies the following formula:

wherein loss is a first loss, R is an integer greater than or equal to 1 and less than or equal to a preset training frequency, t _i For the corresponding pre-labeled frequency deviation of the feature vector, y _i And the frequency offset estimation value corresponding to the feature vector is obtained.

In a fourth aspect, a communication apparatus is provided, where the communication apparatus is a second device, and the second device includes a transceiver module, where the transceiver module is configured to send first information to a first device; the first information includes at least one of: primary synchronization signal, secondary synchronization signal, service data.

Optionally, with reference to the fourth aspect, the transceiver module is further configured to receive capability information of the first device from the first device, where the capability information of the first device is used to indicate that the first device supports frequency offset estimation with integer multiple.

In a fifth aspect, there is provided a communication device comprising a processor, a memory, an input interface for receiving information from a communication device other than the communication device, and an output interface for outputting information to a communication device other than the communication device, the processor invoking a computer program stored in the memory to perform a method as in any of the first or second aspects.

In one possible design, the communication device may be a chip or a device comprising a chip implementing the method of any of the first or second aspects.

In a sixth aspect, there is provided a computer readable storage medium having a computer program stored therein, which when executed, implements a method as in any one of the first or second aspects.

In a seventh aspect, there is provided a chip comprising at least one processor and an interface, the processor being for reading and executing instructions stored in a memory, which when executed cause the chip to perform a method as in any of the first or second aspects.

In an eighth aspect, there is provided a computer program product comprising instructions which, when executed on a computer, cause the method of any one of the first or second aspects to be performed.

In a ninth aspect, a communication system is provided, the communication system comprising a first device for implementing the method of any one of the first or second aspects. In addition, the communication system may further include a second device.

Drawings

The drawings that are necessary for use in the description of the embodiments will be briefly described below.

Wherein:

fig. 1A is an infrastructure of a communication system according to an embodiment of the present application;

fig. 1B is an architecture of yet another communication system provided in an embodiment of the present application;

fig. 1C is an architecture of yet another communication system provided in an embodiment of the present application;

fig. 1D is an architecture of yet another communication system provided in an embodiment of the present application;

fig. 2 is a schematic diagram of a hardware structure of a communication device applicable to the embodiment of the present application;

fig. 3 is a schematic flow chart of a frequency offset estimation method provided in an embodiment of the present application;

fig. 4 is a schematic flow chart of training a network model according to an embodiment of the present application;

fig. 5 is a schematic flow chart of another frequency offset estimation provided in an embodiment of the present application;

FIG. 6 is a graph showing the comparison of the RMSE effect of the scheme and the RMSE effect of the conventional PSS correlation algorithm according to the embodiment of the present application;

FIG. 7 is a graph showing the BER of the scheme and the BER of the conventional PSS correlation algorithm according to the embodiment of the present application;

FIG. 8 is a graph comparing the RMSE effect of yet another embodiment provided by the embodiments of the present application with that of a conventional PSS correlation algorithm;

FIG. 9 is a graph comparing BER of another embodiment provided in the present application with BER of a conventional PSS correlation algorithm;

Fig. 10 is a schematic structural diagram of a communication device according to an embodiment of the present application;

fig. 11 is a schematic structural diagram of a simplified terminal device according to an embodiment of the present application;

fig. 12 is a schematic structural diagram of a simplified network device according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. Wherein the terms "system" and "network" in embodiments of the present application may be used interchangeably. Unless otherwise indicated, "/" indicates that the associated object is an "or" relationship, e.g., A/B may represent A or B; the term "and/or" in this application is merely an association relation describing an association object, and means that three kinds of relations may exist, for example, a and/or B may mean: there are three cases, a alone, a and B together, and B alone, wherein a, B may be singular or plural. Also, in the description of the present application, unless otherwise indicated, "a plurality" means two or more than two. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b, or c may represent: a, b, c, a-b, a-c, b-c, or a-b-c, wherein a, b, c may be one or more. In addition, in order to facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, the words "first", "second", etc. are used to distinguish the network element from the same item or similar items having substantially the same effect. It will be appreciated by those of skill in the art that the words "first," "second," and the like do not limit the amount and order of execution, and that the words "first," "second," and the like do not necessarily differ.

Reference in the specification to "one embodiment" or "some embodiments" or the like means that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of the application. Thus, appearances of the phrases "in one embodiment," "in some embodiments," "in other embodiments," and the like in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments" unless expressly specified otherwise. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless expressly specified otherwise.

The following detailed description is provided for further details of the objects, technical solutions and advantageous effects of the present application, and it should be understood that the following description is only a detailed description of the present application, and is not intended to limit the scope of the present application, and any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solutions of the present application should be included in the scope of protection of the present application.

In the various embodiments of the application, if there is no specific description or logical conflict, terms and/or descriptions between the various embodiments are consistent and may reference each other, and features of the various embodiments may be combined to form new embodiments according to their inherent logical relationships.

At present, the traditional synchronization algorithm is mainly divided into blind synchronization and data auxiliary synchronization, the blind synchronization algorithm does not need additional data auxiliary, and realizes synchronization by means of the characteristics of signals, so that frequency bands can be utilized more effectively, and the effective transmission efficiency of data is higher; the data-aided synchronization algorithm realizes the estimation of the frequency offset according to the position of the peak value of the correlation peak based on the maximum likelihood (maximum likelihood, ML) criterion or the minimum mean square error (minimum mean squared error, MMSE) criterion by adding a training sequence in the time domain or adding a pilot frequency in the frequency domain. Document [1] proposes an OFDM system time-frequency synchronization algorithm using cyclic prefix, which does not require additional training data and estimates frequency offset using cyclic prefix. Because of fewer sampling points, the frequency offset estimation performance is lower, and only the frequency offset with the normalization range of minus 0.5 and 0.5 can be estimated; document [2] proposes an algorithm for estimating the frequency offset of an OFDM system according to a derived cost function which varies with the carrier frequency offset, and provides a method for simplifying iterative calculation; document [3] proposes an OFDM system integer multiple frequency offset estimation algorithm based on a maximum likelihood criterion, the symbols used for estimation may be predefined pilot or PSK modulated data symbols; document [4] proposes an integer frequency offset estimation method for ML-based LTE systems, which uses strong correlation between the received primary synchronization signal (primary synchronization signal, PSS) and the local PSS signal for estimation, but greatly increases the computational complexity.

However, the conventional algorithm has a plurality of limitations, and most of the algorithms depend on an accurate analysis model of the signal and are difficult to acquire in practice; and a lot of related operation calculation complexity is high based on the ML/MMSE criterion; meanwhile, when the signal-to-noise ratio is low, the correlation peak is submerged by noise, so that the estimation performance is drastically reduced, and the estimation performance under a fading channel needs to be improved.

Meanwhile, with the rapid development of deep learning, people start to consider the problem of a communication physical layer from a new angle, document [5] proposes an algorithm for estimating frequency offset according to an IO constellation of an OFDM system received signal based on machine learning, an improved K-means algorithm is utilized to search the centroid of each cluster, and the estimated accuracy of the constellation rotation angle is 100% in a (-45, 45) range; document [6] proposes a single carrier phase shift keying (phase shift keying, PSK) modulation system timing and frequency offset estimation algorithm based on convolutional neural network, but the performance is inferior to the traditional algorithm at high signal-to-noise ratio; document [7] proposes a deep learning based OFDMA uplink multi-frequency synchronization estimation and compensation algorithm that divides the entire carrier frequency offset (carrier frequency offset, CFO) into a set of sub-ranges, each sub-range being compensated by a dedicated CFO compensator; document [8] proposes a coarse frequency offset estimation algorithm of a multi-input multi-output (multiple input multiple output, MIMO) OFDM system based on an LSTM network, but the algorithm is only used for estimating the fractional normalized frequency offset of [ -0.5,0.5 ].

That is, most existing frequency offset estimation algorithms based on machine learning or deep learning are decimal frequency offset estimation, and no integer frequency offset estimation algorithm based on deep learning is found at present; in addition, most of the existing algorithms are OFDM systems that can be designed based on the frame structure by themselves and artificially add training sequences or pilots, and lack integer frequency bias estimation algorithms for LTE, new Radio (NR) systems with fixed frame structures.

In summary, no matter the frequency offset estimation is performed by adopting the traditional synchronization algorithm or the existing frequency offset estimation is performed by a deep learning-based manner, the frame structure needs to be designed artificially. Therefore, the subjective color of the current frequency offset estimation mode is too strong, which is likely to cause insufficient accuracy of frequency offset estimation and further cause low demodulation performance.

Based on this, the embodiment of the present application proposes a frequency offset estimation method to solve the above problem, and the embodiment of the present application is described in detail below.

It should be understood that the technical solutions of the embodiments of the present application may be applied to LTE architecture, fifth generation mobile communication technology (5th generation mobile networks,5G), 4.5 th generation mobile communication technology (the 4.5generation mobile networks,4.5G), wireless local area network (wireless local area networks, WLAN) system, internet of vehicles, V2X communication system, and so on. The technical solution of the embodiment of the present application may also be applied to other future communication systems, such as a 6G communication system, etc., in which the functions may remain the same, but the names may change.

The following describes the infrastructure of the communication system provided in the embodiments of the present application. The technical solution of the embodiment of the present application may be applied to the communication system shown in fig. 1A, 1B, 1C, or 1D. Fig. 1A, fig. 1B, fig. 1C, and fig. 1D are only schematic diagrams, and do not limit the application scenario of the technical solution provided in the present application. In a communication system including communication devices, wireless communication can be performed between the communication devices using air interface resources. The communication device may include a first device and a second device, where the first device may be a terminal device, and the second device may be a network device or a terminal device, and the network device may also be referred to as a base station device. The air interface resources may include at least one of time domain resources, frequency domain resources, code resources, and space resources.

Illustratively, in fig. 1A, the network device and the terminal device may communicate via an uplink and/or a downlink, and the terminal device may communicate via a side-link.

Also exemplary, in fig. 1B, the network device and the vehicle may communicate via an uplink and/or a downlink, and the vehicle may communicate via a side-link.

Also exemplary, in fig. 1C, the processing device or display device and the augmented reality device may communicate via a side-link, or the processing device or display device and the virtual reality device may communicate via a side-link, or the processing device or display device and the mixed reality device may communicate via a side-link.

Also exemplary, in fig. 1D, the router and the terminal device may communicate via an uplink and/or a downlink, the network device and the terminal device may communicate via an uplink and/or a downlink, and the terminal device may communicate via a side downlink.

The network device is an entity on the network side for sending signals, or receiving signals, or sending signals and receiving signals. The network device may be a means deployed in a radio access network (radio access network, RAN) for providing wireless communication functionality for the terminal device, e.g. a transmission reception point (transmission reception point, TRP), a base station, various forms of control nodes. Such as a network controller, a radio controller in a cloud radio access network (cloud radio access network, CRAN) scenario, etc. Specifically, the network device may be a macro base station, a micro base station (also referred to as a small station), a relay station, an Access Point (AP), a radio network controller (radio network controller, RNC), a Node B (NB), a base station controller (base station controller, BSC), a base transceiver station (base transceiver station, BTS), a home base station (e.g., home evolved nodeB, or home node B, HNB), a baseBand unit (BBU), a transmission point (transmitting and receiving point, TRP), a transmitting point (transmitting point, TP), a mobile switching center, or the like, or may be an antenna panel of the base station. The control node may connect to a plurality of base stations and configure resources for a plurality of terminals covered by the plurality of base stations. In systems employing different radio access technologies, the names of base station capable devices may vary. For example, the network device may be a base transceiver station (base transceiver station, BTS) in a global system for mobile communications (global system for mobile communication, GSM) or code division multiple access (code division multiple access, CDMA) network, an NB (NodeB) in a wideband code division multiple access (wideband code division multiple access, WCDMA), an evolved base station (evolutional node B, eNB or eNodeB) in an LTE system, a radio controller in a cloud radio access network (cloud radio access network, CRAN) scenario, a gNB in 5G, or a relay station, an access point, a vehicle device, a wearable device, a network device in a network after 5G, or a network device in a PLMN network for future evolution, and the specific name of the network device is not limited in this application. Secondly, in the embodiment of the present application, the device for implementing the function of the network device may be the network device; or may be a device, such as a system-on-a-chip, capable of supporting the network device to perform this function, which may be installed in the network device.

The terminal device is an entity on the user side for receiving signals or transmitting signals or receiving signals and transmitting signals. The terminal device is configured to provide one or more of a voice service and a data connectivity service to a user. The terminal device may be a device that includes a wireless transceiver function and may cooperate with a network device to provide a communication service for a user. In particular, a terminal device may refer to a User Equipment (UE), an access terminal, a subscriber unit, a subscriber station, a mobile station, a remote terminal, a mobile device, a terminal, a wireless communication device, a user agent, or a user equipment. The terminal device may also be an unmanned aerial vehicle, an internet of things (internet of things, ioT) device, a station in a WLAN, a cellular telephone (ST), a smart phone (smart phone), a cordless telephone, a wireless data card, a tablet, a session initiation protocol (session initiation protocol, SIP) phone, a wireless local loop (wireless local loop, WLL) station, a personal digital assistant (personal digital assistant, PDA) device, a laptop (lap computer), a machine type communication (machine type communication, MTC) terminal, a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, a vehicle-mounted device, a wearable device (also may be referred to as a wearable smart device), a Virtual Reality (VR) terminal, an augmented reality (augmented reality, AR) terminal, a wireless terminal in industrial control (industrial control), a wireless terminal in an unmanned aerial vehicle (self-driving) such as a vehicle, a wireless terminal in a remote medical medium, a smart grid (smart) terminal, a smart wireless terminal in a smart grid (smart home) or the like. The terminal device may also be a device-to-device (D2D) device, such as an electricity meter, water meter, etc. The terminal equipment can also be deployed on the water surface (such as a ship and the like); but may also be deployed in the air (e.g., on aircraft, balloon, satellite, etc.). The terminal device may also be a terminal in a 5G system, or may be a terminal in a next generation communication system, which is not limited in the embodiment of the present application. Secondly, in the embodiment of the present application, the device for implementing the function of the terminal device may be the terminal device; or a device, such as a chip system, capable of supporting the terminal device to realize the function, which may be installed in the terminal device. In the embodiment of the application, the chip system may be formed by a chip, and may also include a chip and other discrete devices.

The technical scheme provided by the embodiment of the application can also be applied to wireless communication among communication equipment. The wireless communication between the communication devices may include: wireless communication between a network device and a terminal device, wireless communication between a network device and a network device, and wireless communication between a terminal device and a terminal device. In this embodiment of the present application, the term "wireless communication" may also be simply referred to as "communication", and the term "communication" may also be described as "data transmission", "information transmission" or "transmission".

Alternatively, each device (e.g., the first device, the second device, etc.) in fig. 1A-1D may be implemented by one device, or may be implemented by a plurality of devices together, or may be a functional module in one device, which is not specifically limited in this embodiment of the present application. It will be appreciated that the above described functionality may be either a network element in a hardware device, a software function running on dedicated hardware, or a virtualized function instantiated on a platform (e.g., a cloud platform).

For example, each of the devices of fig. 1A-1D may be implemented by the communications apparatus 200 of fig. 2. Fig. 2 is a schematic hardware structure of a communication device applicable to the embodiment of the present application. The communication device 200 comprises at least one processor 201, communication lines 202, a memory 203 and at least one communication interface 204.

The processor 201 may be a general purpose central processing unit (central processing unit, CPU), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of programs in accordance with aspects of the present application.

Communication line 202 may include a pathway to transfer information between the aforementioned components.

The communication interface 204 is any transceiver-like device (e.g., antenna, etc.) for communicating with other devices or communication networks, such as ethernet, RAN, wireless local area network (wireless local area networks, WLAN), etc.

The memory 203 may be, but is not limited to, a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (random access memory, RAM) or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (electrically erasable programmable read-only memory, EEPROM), a compact disc read-only memory (compact disc read-only memory) or other optical disc storage, optical disc storage (including compact disc, laser disc, optical disc, digital versatile disc, blu-ray disc, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. The memory may be stand alone and be coupled to the processor via communication line 202. The memory may also be integrated with the processor. The memory provided by embodiments of the present application may generally have non-volatility.

The memory 203 is used for storing computer-executable instructions for executing the embodiments of the present application, and is controlled by the processor 201 to execute the instructions. The processor 201 is configured to execute computer-executable instructions stored in the memory 203, thereby implementing the methods provided in the embodiments described below.

Alternatively, the computer-executable instructions in the embodiments of the present application may be referred to as application program codes, which are not specifically limited in the embodiments of the present application.

In one possible implementation, processor 201 may include one or more CPUs, such as CPU0 and CPU1 of FIG. 2.

In one possible implementation, the communication device 200 may include multiple processors, such as the processor 201 and the processor 207 in fig. 2. Each of these processors may be a single-core (single-CPU) processor or may be a multi-core (multi-CPU) processor. A processor herein may refer to one or more devices, circuits, and/or processing cores for processing data (e.g., computer program instructions).

In one possible implementation, the communications apparatus 200 can further include an output device 205 and an input device 206. The output device 205 communicates with the processor 201 and may display information in a variety of ways. For example, the output device 205 may be a liquid crystal display (liquid crystal display, LCD), a light emitting diode (light emitting diode, LED) display device, a Cathode Ray Tube (CRT) display device, or a projector (projector), or the like. The input device 206 is in communication with the processor 201 and may receive user input in a variety of ways. For example, the input device 206 may be a mouse, a keyboard, a touch screen device, a sensing device, or the like.

The communication apparatus 200 may be a general-purpose device or a special-purpose device. In a specific implementation, the communication apparatus 200 may be a desktop, a portable computer, a network server, a palm computer (personal digital assistant, PDA), a mobile phone, a tablet computer, a wireless terminal device, an embedded device, or a device having a similar structure as in fig. 2. The embodiments of the present application are not limited to the type of communication device 200.

The technical solution provided in the embodiments of the present application will be described below by taking a first device as a terminal device and a second device as a network device, with reference to the accompanying drawings.

Referring to fig. 3, fig. 3 is a schematic flow chart of a frequency offset estimation method according to an embodiment of the present application. As shown in fig. 3, the method includes, but is not limited to, the steps of:

301. the network device sends the first information to the terminal device.

Accordingly, the terminal device receives the first information from the network device.

Wherein the first information comprises at least one of: primary synchronization signal (primary synchronization signal, PSS), secondary synchronization signal (secondary synchronization signal, SSS), traffic data. The first information is located on at least one subframe.

For example, for LTE, PSS may be located on subframe 0 and subframe 5. Specifically, the PSS may be located at the last OFDM symbol of the first slot on subframe 0 and subframe 5.

In addition, the specific generation modes of the PSS and SSS are not limited in this application. As in the 3GPP protocol, for LTE, PSS is a frequency-domain-based Zadoff-Chu sequence, which satisfies the following equation (1):

where u is the root sequence number of the Zadoff-Chu sequence, which is related to the cell Identification (ID). D according to formula (1) _u (n) is a constant amplitude zero auto-correlation sequence with length of 62, and has good auto-correlation characteristics, so that the constant amplitude zero auto-correlation sequence is used as a PSS synchronization signal of an LTE system to realize accurate synchronization and improve system performance.

The type of the service data and the like are not limited in the application.

302. And the terminal equipment carries out N times of cyclic prediction according to the symbol where the first information is located to obtain a first frequency offset estimation value.

Wherein N is an integer greater than or equal to 1.

Optionally, step 302 may include: and the terminal equipment carries out N times of cyclic prediction according to the real part and the imaginary part of the symbol where the first information is located to obtain a first frequency offset estimation value. Namely, the first frequency offset estimation value is obtained by carrying out cyclic prediction for N times according to the real part and the imaginary part of the symbol where the first information is located, so that input data is enriched, and the frequency offset estimation value can be determined more accurately.

The terminal device performs N-time cyclic prediction according to the real part and the imaginary part of the symbol where the first information is located to obtain a first frequency offset estimation value, which can be understood as: and the terminal equipment carries out cyclic prediction for N times according to the real part and the imaginary part of the symbol where the first information is located on at least one subframe to obtain a first frequency offset estimation value.

For example, the terminal device may perform N-cycle prediction according to the real part and the imaginary part of the symbol where the PSS is located on the subframe 0 to obtain the first frequency offset estimation value.

Optionally, the ith loop prediction in the N loop predictions includes the following steps: the terminal equipment performs frequency offset compensation on the symbol where the first information is located in the ith-1 cycle prediction according to the frequency offset estimation value obtained in the ith-1 cycle prediction, so as to obtain the symbol where the first information is located in the ith cycle prediction; and the terminal equipment predicts and obtains a frequency offset estimation value during the ith cyclic prediction according to the symbol where the first information is located during the ith cyclic prediction. Wherein i is an integer greater than or equal to 1 and less than or equal to N. The frequency offset is compensated by adopting a cyclic iteration method through repeated cyclic prediction, the sensitivity of the algorithm to high frequency offset is weakened, and the estimation performance of the algorithm is further improved through repeated iteration.

The terminal device performs frequency offset compensation on the symbol where the first information is located during the ith-1 th cyclic prediction according to the frequency offset estimation value obtained during the ith-1 th cyclic prediction, so as to obtain the symbol where the first information is located during the ith cyclic prediction, which can be understood as follows: and the terminal equipment performs frequency offset compensation on the real part and the imaginary part of the symbol where the first information is located in the ith-1 cycle prediction according to the frequency offset estimation value obtained in the ith-1 cycle prediction, so as to obtain the real part and the imaginary part of the symbol where the first information is located in the ith cycle prediction. Similarly, the terminal device predicts and obtains the frequency offset estimation value during the ith cyclic prediction according to the symbol where the first information is located during the ith cyclic prediction, which can be understood as: and the terminal equipment predicts and obtains a frequency offset estimation value during the ith cyclic prediction according to the real part and the imaginary part of the symbol where the first information is located during the ith cyclic prediction.

Optionally, if i is equal to 1, the frequency offset estimation value in the ith cyclic prediction is an integer multiple frequency offset estimation value. The integer multiple frequency offset estimation value is greater than or equal to a first frequency offset value and less than or equal to a second frequency offset value, wherein the first frequency offset value is-4, and the second frequency offset value is 4. In other words, the integer multiple of the frequency offset estimate may fall within a predetermined frequency offset range of [ -4,4]. Namely, it can be seen that the frequency offset estimation value in the 1 st cycle prediction is an integer multiple frequency offset estimation value, so that the frequency offset range which can be estimated is greatly widened.

Optionally, in the present application, the condition for stopping loop prediction includes that N is the maximum number of loops; or the first frequency offset estimation value is smaller than a preset first threshold value; or, the difference between the frequency offset estimation value in the i-1 th cycle prediction and the frequency offset estimation value in the i-1 th cycle prediction is smaller than a preset second threshold value. In other words, N is the maximum number of cycles, i.e. i is equal to N, and the terminal device determines the frequency offset estimation value during the ith cycle prediction as the first frequency offset estimation value, and stops the cycle prediction; or, the first frequency offset estimation value is smaller than a preset first threshold value, and the cyclic prediction is stopped; or, the difference between the frequency offset estimation value in the i-1 th cyclic prediction and the frequency offset estimation value in the i-1 th cyclic prediction is smaller than a preset second threshold value, and the cyclic prediction is stopped. The preset first threshold value and the preset second threshold value may be predefined thresholds of a protocol, or thresholds configured by the network device to the terminal device, which are not limited herein. It should be understood that, in the present application, the preset first threshold may be 1, 2 or 3, which is not limited herein. When the preset first threshold value is 2, the demodulation performance is best when data demodulation is performed according to the first frequency offset estimation value.

303. And the terminal equipment demodulates the data according to the first frequency offset estimation value. According to the technical scheme, at least one of the main synchronization signal, the auxiliary synchronization signal and the service data is received from the network equipment, so that the first frequency offset estimation value can be obtained by carrying out N-time cyclic prediction according to the symbol where at least one of the main synchronization signal, the auxiliary synchronization signal and the service data is located, the problem that the frequency offset estimation is not accurate enough due to the fact that the subjective color is too strong when the frequency offset estimation is carried out according to the artificially designed reference signal is avoided, and the frame structure is also prevented from being changed. Namely, under the condition of not changing the frame structure, the accuracy of frequency offset estimation is improved, and the demodulation performance is further improved. In addition, by means of repeated cyclic prediction, the frequency offset is compensated by adopting a cyclic iteration method, the sensitivity of the algorithm to high frequency offset is weakened, and the estimation performance of the algorithm is further improved through repeated iteration.

Optionally, the scheme may further include: the terminal device sends the capability information of the terminal device to the network device, and correspondingly, the network device receives the capability information of the terminal device from the terminal device. The capability information of the terminal equipment is used for indicating that the terminal equipment supports integral multiple frequency offset estimation. That is, the network device may be enabled to learn that the terminal device supports frequency offset estimation of integer multiples, and may further issue at least one of a primary synchronization signal, a secondary synchronization signal, and service data.

It should be noted that, the sending, by the terminal device, the capability information of the terminal device to the network device may be performed before step 301.

In one embodiment of the present application, performing N-cycle prediction according to the real part and the imaginary part of the symbol where the first information is located to obtain the first frequency offset estimation value is performed by using a network model. The process of the network model for N-cycle prediction is described in detail below in conjunction with the network structure of the network model.

Wherein the network model comprises: input layer, convolution layer, pooling layer, full connection layer and output layer. Specifically, the network model may be implemented in any of the following manners, which are not limited herein.

In the mode 1, the number of the convolution layers is 8, the convolution layers extract and learn the characteristics of the data, the convolution kernel sizes of the convolution layers are (4, 2), and the number of the filters is 16, 32, 64, 128 and 256 respectively; the activation function adopts a linear correction unit Relu function; the pooling layer adopts average pooling, the pooling core size is (2, 2), and the pooling layer compresses the input characteristic diagram. On one hand, the feature map is reduced, the network calculation complexity is simplified, and on the other hand, feature compression is carried out to extract main features; in addition, a batch specification (BatchNorm) layer is added to accelerate convergence, improve generalization ability, and prevent overfitting.

The number of the convolution layers is 6, the convolution layers extract and learn the characteristics of the data, the convolution kernel sizes of the convolution layers are (4, 2), and the number of the filters is 16, 32, 64 and 64 respectively; the activation function adopts a linear correction unit Relu function; the pooling layer adopts average pooling, the pooling core size is (2, 2), the pooling layer compresses the input feature map, on one hand, the feature map is reduced, the network calculation complexity is simplified, on the other hand, feature compression is carried out, and main features are extracted; in addition, the BatchNorm layer is added to accelerate convergence, improve generalization capability and prevent overfitting.

For example, for mode 1, see table 1, where table 1 is a structure of a network model provided in an embodiment of the present application. In conjunction with Table 1, it can be seen that the data dimension of the Input (Input) layer output is (1096,2). The output data of the input layer may sequentially pass through the convolution layer, the BatchNorm layer and the ReLU layer, and output data with dimension (548,1), that is, the output data of the input layer passes through the network layer including Conv2D, batchNorm, reLu, and may output data with dimension (548,1). Similarly, for the rest of the network layers in table 1 including Conv2D, batchNorm, reLu, data in the corresponding dimension may also be output, which is not described herein. Second, data of dimension (548,1) can be input to a pooling layer, such as AaveragePooling2D, outputting data of dimension (274,1). In addition, the data with the dimension of (3, 1) can sequentially pass through the full connection layer and the output layer to obtain the data with the dimension of 1.

Table 1: the embodiment of the application provides a network model structure

Network layer	Output dimension
		Input	(1096,2)
Conv2D+BatchNorm+ReLu	(548,1)
		AaveragePooling2D	(274,1)
Conv2D+BatchNorm+ReLu	(137,1)
		Conv2D+BatchNorm+ReLu	(69,1)
Conv2D+BatchNorm+ReLu	(35,1)
		Conv2D+BatchNorm+ReLu	(18,1)
Conv2D+BatchNorm+ReLu	(9,1)
		Conv2D+BatchNorm+ReLu	(5,1)
Conv2D+BatchNorm+ReLu	(3,1)
		Dense+Regression Output	1

For example, for mode 2, see table 2, where table 2 is a structure of yet another network model provided in an embodiment of the present application. In conjunction with Table 2, it can be seen that the data dimension of the Input (Input) layer output is (137,2). The output data of the input layer may sequentially pass through the convolution layer, the BatchNorm layer and the ReLU layer, and output data with dimension (67,1), that is, the output data of the input layer passes through the network layer including Conv2D, batchNorm, reLu, and may output data with dimension (67,1). Similarly, for the rest of the network layers in table 1 including Conv2D, batchNorm, reLu, data in the corresponding dimension may also be output, which is not described herein. Second, data of dimension (67,1) can be input to a pooling layer, such as AaveragePooling2D, outputting data of dimension (34,1). In addition, the data with the dimension of (2, 1) can sequentially pass through the full connection layer and the output layer to obtain the data with the dimension of 1.

Table 2: structure of another network model provided in the embodiment of the present application

Network layer	Output dimension
		Input	(137,2)
Conv2D+BatchNorm+ReLu	(67,1)
		AaveragePooling2D	(34,1)
Conv2D+BatchNorm+ReLu	(17,1)
		Conv2D+BatchNorm+ReLu	(9,1)
Conv2D+BatchNorm+ReLu	(5,1)
		Conv2D+BatchNorm+ReLu	(3,1)
Conv2D+BatchNorm+ReLu	(2,1)
		Dense+Regression Output	1

For example, when the network model in mode 1 is adopted, the terminal device predicts, according to the symbol where the first information is located during the ith cyclic prediction, a frequency offset estimation value during the ith cyclic prediction, and may include: the terminal equipment carries out vectorization processing on the symbol where the first information is located in the ith cyclic prediction to obtain a feature vector in the ith cyclic prediction; the terminal equipment inputs the feature vector in the ith cyclic prediction into a trained network model, and predicts the frequency offset estimation value in the ith cyclic prediction. The terminal device performs vectorization processing on the symbol where the first information is located during the ith cyclic prediction to obtain a feature vector during the ith cyclic prediction, and the method may include: and the terminal equipment carries out vectorization processing on the real part and the imaginary part of the symbol where the first information is located in the ith cyclic prediction to obtain a feature vector in the ith cyclic prediction.

In one embodiment of the present application, the network model is obtained by training a training sample that is constructed in advance. The process of constructing a training sample and training the network model using the training sample is described below with reference to the accompanying drawings.

Referring to fig. 4, fig. 4 is a schematic flow chart of a training network model according to an embodiment of the present application. The method includes, but is not limited to, the steps of:

401. the terminal equipment acquires the symbol where the second information is located.

Wherein the second information includes at least one of: the primary synchronization signal, the secondary synchronization signal, and the service data are not limited herein.

Optionally, step 401 may include: the terminal equipment acquires the real part and the imaginary part of the symbol where the second information is located. It will be appreciated that the second information is one sample of data in the training set. In the application, in the offline stage, i.e. during model training, for the network model in the above mode 1, a random bit stream range may be set to 1-500, a cell ID range to 0-2, and a normalized carrier frequency deviation range to-4, so as to generate 50000 sets of sample data altogether, and divide the sample data into a training set and a verification set according to a ratio of 8:2. Each sample data is 1096 sampling points of one OFDM symbol where at least one of the primary synchronization signal, the secondary synchronization signal and the service data is located, and is unfolded into a 1096 x 2 matrix form according to a real part and an imaginary part, and the corresponding normalized carrier frequency difference is used as a label. For the network model in the above mode 2, a random bit stream range may be set to 1-500, a cell ID range to 0-2, and a normalized carrier frequency deviation range to-4, so as to generate 50000 groups of sample data altogether, and divide the sample data into a training set and a verification set according to a ratio of 8:2. Each sample data is 137 sampling points of one OFDM symbol where at least one of a main synchronous signal, an auxiliary synchronous signal and service data is located, and is unfolded into a matrix form of 137 x 2 according to a real part and an imaginary part, and the corresponding normalized carrier frequency difference is used as a label.

In the present application, the normalized carrier frequency difference is a normalized carrier frequency deviation.

402. And the terminal equipment carries out vectorization processing on the symbol where the second information is located to obtain a feature vector.

Optionally, step 402 may include: and the terminal equipment carries out vectorization processing on the real part and the imaginary part of the symbol where the second information is located to obtain a feature vector.

403. And the terminal equipment inputs the feature vector into a network model, and predicts to obtain a frequency offset estimation value corresponding to the feature vector.

404. And the terminal equipment adjusts model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector so as to train the network model.

Optionally, step 404 may include: the terminal equipment obtains a first loss according to the frequency deviation estimation value corresponding to the feature vector and the pre-marked frequency deviation corresponding to the feature vector; and the terminal equipment adjusts the model parameters of the network model according to the first loss so as to train the network model.

The pre-labeled frequency deviation corresponding to the feature vector is the pre-labeled normalized carrier frequency deviation corresponding to the feature vector.

Wherein the first loss satisfies the following formula:

wherein loss is a first loss, R is an integer greater than or equal to 1 and less than or equal to a preset training frequency, t _i For the corresponding pre-labeled frequency deviation of the feature vector, y _i And the frequency offset estimation value corresponding to the feature vector is obtained. The preset number of training times may be, for example, 100.

It should be noted that, for the network models in the above modes 1 and 2, the neural network optimizer may be selected as the Adam optimizer, the initial learning rate is 0.001, the attenuation is 0.2 for each training 20 rounds, the small batch size (minibatch size) is 256, and the maximum training frequency is 100. Training and optimizing the network by adopting a training set, and verifying the effect on a testing set. The network effect is measured in terms of root mean square error (root mean square error, RMSE), where,considering that the prediction result should be an integer, RMSE is calculated after rounding the prediction result.

In addition, in the application, because each sample data in the training set is 1096 sampling points of one OFDM symbol where at least one of the main synchronization signal, the auxiliary synchronization signal and the service data is located, or each sample data in the training set is 137 sampling points of one OFDM symbol where at least one of the main synchronization signal, the auxiliary synchronization signal and the service data is located, the problem that frequency offset estimation is not accurate enough due to over-strong subjective color when training is carried out according to a reference signal designed by people is avoided, and the frame structure is also avoided. Meanwhile, by constructing a large amount of data with variability, the frequency offset is estimated in a data driving mode, and the generalization capability and the adaptability of the model are improved without depending on a specific system model and a signal structure.

The process of frequency offset estimation will be described below by taking the first information as PSS and the network model as the network model in mode 1 as an example.

Referring to fig. 5, fig. 5 is a schematic flow chart of another frequency offset estimation provided in an embodiment of the present application. As shown in fig. 5, the method includes, but is not limited to, the steps of:

501. the terminal device receives the PSS from the network device.

Correspondingly, the network device sends the PSS to the terminal device.

502. The terminal device sets the initial iteration number i to 1.

503. And the terminal equipment extracts the received data of the symbol where the PSS is located, and obtains the real part and the imaginary part of the symbol where the PSS is located in the ith cyclic prediction.

504. The terminal equipment inputs the real part and the imaginary part of the symbol of the PSS in the ith cyclic prediction into a trained network model, and predicts the frequency offset estimation value in the ith cyclic prediction.

505. The terminal equipment determines whether the frequency offset estimation value in the ith cyclic prediction is smaller than a preset first threshold value.

The presetting of the first threshold may be described with reference to step 302 of fig. 3, which is not described herein.

506. If yes, the terminal equipment demodulates the data according to the frequency offset estimation value in the ith cyclic prediction.

507. If not, the terminal equipment determines whether the difference between the frequency offset estimation value in the ith-1 loop prediction and the frequency offset estimation value in the ith loop prediction is smaller than a preset second threshold value.

The second threshold value may be preset, and will be described with reference to step 302 of fig. 3, which is not described herein.

508. If yes, the terminal equipment demodulates the data according to the frequency offset estimation value in the ith cyclic prediction.

509. If not, the terminal device determines whether i is less than or equal to the maximum number of cycles.

510. If yes, the terminal equipment demodulates the data according to the frequency offset estimation value in the ith cyclic prediction.

511. If not, the terminal equipment adopts the frequency offset estimation value in the ith cyclic prediction to perform frequency offset compensation on the real part and the imaginary part of the symbol where the PSS is located in the ith cyclic prediction, so as to obtain the real part and the imaginary part of the symbol where the PSS is located in the ith cyclic prediction after the frequency offset compensation, and the terminal equipment assigns i+1 to i, and returns to execute step 504.

It can be understood that, after the terminal device assigns i+1 to i, the real part and the imaginary part of the symbol where the PSS is located in the ith cyclic prediction in step 504 are the real part and the imaginary part of the symbol where the PSS is located in the ith+1 cyclic prediction after the frequency offset compensation.

512. And the terminal equipment ends the predicted frequency offset estimation value.

According to the technical scheme, the frequency offset is compensated by adopting a cyclic iteration method through repeated cyclic prediction, the sensitivity of the algorithm to high frequency offset is weakened, and the estimation performance of the algorithm is further improved through repeated iteration.

The beneficial effects produced by the present scheme are described below in connection with the simulation diagrams.

For the network model in embodiment 1, the simulation environment used in the simulation includes: the simulated LTE system is a single-input single-output (single input single output, SISO) system, the downlink transmission bandwidth is 10MHz, the FFT point number is 1024, the cyclic prefix type is a common cyclic prefix, the cyclic prefix length is 80 and 72, the number of resource blocks on a resource grid is 50, the subcarrier number of each resource block is 12, the subcarrier interval is 15KHz, the channel sampling rate is 15.36MHz, and the modulation mode is 16-quadrature amplitude modulation (quadrature amplitude modulation, QAM). The channel models are additive white gaussian noise (additive white gaussian noise, AWGN) channels and high mobility frequency selective rayleigh (rayleigh) channels. In the Rayleigh channel parameter setting, the number of multipath transmission paths is 5, the path delays are respectively 0us, 0.65us, 1.3us, 1.95us and 6.51us, the path gains are respectively 0dB, -3dB, -6dB, -8dB and-172 dB, and the Doppler expansion is 70Hz. The RMSE effect of the scheme is compared with that of the conventional PSS correlation algorithm under AWGN channel and rayleigh channel, and the simulation result is shown in fig. 6. Referring to fig. 6, it can be seen that signal-to-noise ratio (SNR) is achieved in the AWGN channel <At-4 dB, the performance of the scheme is far better than that of the traditional PSS correlation algorithm; when the SNR is more than or equal to 2dB, the normalized root mean square error of the two algorithms can be reduced to 10 ^-2 Below the magnitude; under the rayleigh channel, the performance of the scheme is obviously superior to that of the traditional PSS correlation algorithm. Secondly, the scheme passes through N times of circulationAnd predicting to obtain a first frequency offset estimation value, and comparing a Bit Error Rate (BER) of data demodulation based on the first frequency offset estimation value with the BER of a traditional PSS correlation algorithm, wherein a simulation result is shown in fig. 7. With reference to fig. 7, it can be seen that, under AWGN channel, the performance of the scheme is comparable to that of the conventional PSS correlation algorithm at low signal-to-noise ratio, at SNR>When 8dB is reached, the BER of the scheme and the BER of the traditional PSS correlation algorithm can be reduced to 0; under the rayleigh channel, the BER of the scheme is continuously reduced along with the increase of the SNR, and the effect is slightly stronger than that of the traditional PSS correlation algorithm under the high signal-to-noise ratio.

For the network model in mode 2, the simulation environment used in the simulation includes: the simulated LTE system is a SISO system, the downlink transmission bandwidth is 1.4MHz, the FFT point number is 128, the cyclic prefix type is selected to be a common cyclic prefix, the cyclic prefix length is 10 and 9, the number of resource blocks on a resource grid is 6, the number of subcarriers of each resource block is 12, the subcarrier interval is 15KHz, the channel sampling rate is 1.92MHz, and the modulation mode is 16QAM. The channel model is an additive white gaussian noise channel and a high mobility frequency selective rayleigh channel. In the Rayleigh channel parameter setting, the number of multipath transmission paths is 5, the path delays are respectively 0us, 0.65us, 1.3us, 1.95us and 6.51us, the path gains are respectively 0dB, -3dB, -6dB, -8dB and-172 dB, and the Doppler expansion is 70Hz. The RMSE effect of the scheme is compared with that of the conventional PSS correlation algorithm under AWGN channel and rayleigh channel, and the simulation result is shown in fig. 8. Referring to fig. 8, it can be seen that SNR under AWGN channel <At-7 dB, the performance of the scheme is far better than that of the traditional PSS correlation algorithm; when the SNR is more than or equal to 2dB, the normalized root mean square error of the two algorithms is reduced to 0; under the rayleigh channel, the performance of the scheme is obviously superior to that of the traditional PSS correlation algorithm. Next, the first frequency offset estimation value is obtained through N-time cyclic prediction, and Bit Error Rate (BER) of data demodulation based on the first frequency offset estimation value is compared with BER of a conventional PSS correlation algorithm, and a simulation result is shown in fig. 9. With reference to fig. 9, it can be seen that the performance of the scheme is slightly stronger than that of the conventional PSS correlation algorithm under the AWGN channel, and the SNR is>At 8dB, BER of the scheme and BER of the traditional PSS correlation algorithmCan all be reduced to 10 ^-4 The following are set forth; in rayleigh channel, BER of the scheme and BER of the unified PSS correlation algorithm continuously decrease with increasing SNR, but can only be kept at 3×10 finally ^-1 Left and right.

The foregoing description of the solution provided in this application has been presented primarily from the perspective of interaction between the devices. It will be appreciated that the above-described implementation of the various devices to implement the above-described functions includes corresponding hardware structures and/or software modules that perform the various functions. Those of skill in the art will readily appreciate that the various illustrative modules and algorithm steps described in connection with the embodiments disclosed herein may be implemented as hardware or combinations of hardware and computer software. Whether a function is implemented as hardware or computer software driven hardware depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

The embodiment of the present application may divide the functional modules of the first device or the second device according to the above method example, for example, each functional module may be divided corresponding to each function, or two or more functions may be integrated into one processing module, where the integrated modules may be implemented in a form of hardware or in a form of a software functional module. It should be noted that, in the embodiment of the present application, the division of the modules is schematic, which is merely a logic function division, and other division manners may be implemented in actual implementation.

In the case of using an integrated module, referring to fig. 10, fig. 10 is a schematic structural diagram of a communication device according to an embodiment of the present application. The communication device 1000 may be applied to the method shown in fig. 3, 4 or 5, as shown in fig. 10, where the communication device 1000 includes a transceiver module 1001 and a processing module 1002, or where the communication device 1000 includes the transceiver module 1001. The transceiver module 1001 may be a transceiver or a communication interface, and the processing module 1002 may be one or more processors. The communication means may be used to implement the first device or the second device involved in any of the method embodiments described above, or to implement the functionality of the device involved in any of the method embodiments described above. For example, the communication means may be the first device or the second device. The first device or the second device may be either a network element in a hardware device, a software function running on dedicated hardware, or a virtualized function instantiated on a platform (e.g., a cloud platform). Optionally, the communication device 1000 may further comprise a storage module 1003 for storing program code and data of the communication device 1000.

Alternatively, when the communication apparatus is used as the first device or is a chip applied in the first device, the communication apparatus 1000 includes a transceiver module 1001 and a processing module 1002, and performs the steps performed by the first device in the above-described method embodiment. The transceiver module 1001 is configured to support communication with a second device, etc., and specifically perform the sending and/or receiving actions performed by the first device in fig. 3-5, which are not described herein. Such as other processes that support the first device to perform the techniques described herein. The processing module 1002 may be configured to support the communication device 1000 to perform the processing actions in the above-described method embodiments, which are not described herein. For example, the first device is enabled to perform step 303, and/or other processes for the techniques described herein.

Illustratively, a transceiver module 1001 for receiving first information from a second device; the first information includes at least one of: a primary synchronization signal, a secondary synchronization signal, and service data; a processing module 1002, configured to:

Optionally, when the communication apparatus is used as the second device or is a chip applied in the second device, the communication apparatus 1000 includes a transceiver module 1001 and a processing module 1002, and performs the steps performed by the second device in the above-described method embodiment. The transceiver module 1001 is configured to support communication with the first device, and the like, and specifically perform the sending and/or receiving actions performed by the second device in fig. 3-5, which are not described herein. Such as supporting the second device to perform step 301, and/or other processes of the techniques described herein. The processing module 1002 may be configured to support the communication device 1000 to perform the processing actions in the above-described method embodiments, which are not described herein. For example, other processes that support the second device to perform the techniques described herein.

The transceiver module 1001 is configured to send first information to a first device, where the first information includes at least one of: primary synchronization signal, secondary synchronization signal, service data.

In one possible implementation, when the communication device is a chip, the transceiver module 1001 may be an interface, pin, circuit, or the like. The interface may be used to input data to be processed to the processor, and may output a processing result of the processor to the outside. In a specific implementation, the interface may be a general purpose input output (general purpose input output, GPIO) interface, which may be connected to a plurality of peripheral devices (e.g., a display (LCD), a camera (cam), a Radio Frequency (RF) module, an antenna, etc.). The interface is connected with the processor through a bus.

The processing module 1002 may be a processor that may execute computer-executable instructions stored by the memory module to cause the chip to perform the methods of the embodiments of fig. 3 or fig. 4 or fig. 5.

Further, the processor may include a controller, an operator, and a register. Illustratively, the controller is primarily responsible for instruction decoding and issues control signals for the operations to which the instructions correspond. The arithmetic unit is mainly responsible for performing fixed-point or floating-point arithmetic operations, shift operations, logic operations, and the like, and may also perform address operations and conversions. The register is mainly responsible for storing register operands, intermediate operation results and the like temporarily stored in the instruction execution process. In particular implementations, the hardware architecture of the processor may be an application specific integrated circuit (application specific integrated circuits, ASIC) architecture, a microprocessor (microprocessor without interlocked piped stages architecture, MIPS) architecture of an interlocking-free pipeline stage architecture, an advanced reduced instruction set machine (advanced RISC machines, ARM) architecture, or a network processor (network processor, NP) architecture, among others. The processor may be single-core or multi-core.

The memory module 1003 may be a memory module within the chip, such as a register, a cache, etc. The Memory module may also be a Memory module located outside the chip, such as a Read Only Memory (ROM) or other type of static storage device that can store static information and instructions, a random access Memory (Random Access Memory, RAM), etc.

It should be noted that, the functions corresponding to the processor and the interface may be implemented by hardware design, or may be implemented by software design, or may be implemented by a combination of software and hardware, which is not limited herein.

Fig. 11 is a schematic structural diagram of a simplified terminal device according to an embodiment of the present application. For easy understanding and convenient illustration, in fig. 11, a mobile phone is taken as an example of the terminal device. As shown in fig. 11, the terminal device includes at least one processor, and may further include a radio frequency circuit, an antenna, and an input-output device. The processor may be used for processing communication protocols and communication data, controlling the terminal device, executing a software program, processing data of the software program, and the like. The terminal device may also comprise a memory for storing mainly software programs and data, which programs may be loaded into the memory at the time of shipment of the communication device or reloaded into the memory at a later time when needed. The radio frequency circuit is mainly used for converting a baseband signal and a radio frequency signal and processing the radio frequency signal. The antenna is mainly used for receiving and transmitting radio frequency signals in the form of electromagnetic waves. Input and output devices, such as touch screens, display screens, keyboards, etc., are mainly used for receiving data input by a user and outputting data to the user. It should be noted that some kinds of terminal apparatuses may not have an input/output device.

When data need to be sent, the processor carries out baseband processing on the data to be sent and then outputs a baseband signal to the radio frequency circuit, and the radio frequency circuit carries out radio frequency processing on the baseband signal and then sends the radio frequency signal outwards in the form of electromagnetic waves through the antenna. When data is sent to the terminal equipment, the radio frequency circuit receives a radio frequency signal through the antenna, converts the radio frequency signal into a baseband signal, and outputs the baseband signal to the processor, and the processor converts the baseband signal into data and processes the data. For ease of illustration, only one memory and processor is shown in fig. 11. In an actual end device product, there may be one or more processors and one or more memories. The memory may also be referred to as a storage medium or storage device, etc. The memory may be provided separately from the processor or may be integrated with the processor, which is not limited by the embodiments of the present application.

In the embodiment of the present application, the antenna and the radio frequency circuit having the transceiver function may be regarded as a receiving unit and a transmitting unit (may also be collectively referred to as a transceiver unit) of the terminal device, and the processor having the processing function may be regarded as a processing unit of the terminal device. As shown in fig. 11, the terminal device includes a receiving module 31, a processing module 32, and a transmitting module 33. The receiving module 31 may also be referred to as a receiver, a receiving circuit, etc., and the transmitting module 33 may also be referred to as a transmitter, a transmitting circuit, etc. The processing module 32 may also be referred to as a processor, processing board, processing device, etc.

For example, the processing module 32 is configured to perform the functions of the terminal device in the embodiment shown in fig. 3 or fig. 4 or fig. 5.

Fig. 12 is a schematic structural diagram of a simplified network device according to an embodiment of the present application. The network device includes a radio frequency signal transceiving and converting part and a part 42, which in turn includes a receiving module 41 part and a transmitting module 43 part (which may also be collectively referred to as transceiving modules). The radio frequency signal receiving and transmitting and converting part is mainly used for receiving and transmitting radio frequency signals and converting radio frequency signals and baseband signals; the 42 part is mainly used for baseband processing, control of network equipment and the like. The receiving module 41 may also be referred to as a receiver, a receiving circuit, etc., and the transmitting module 43 may also be referred to as a transmitter, a transmitting circuit, etc. Portion 42 is typically a control center of the network device, and may be generally referred to as a processing module, for controlling the network device to perform the steps performed with respect to the network device in fig. 3-5 described above. See for details the description of the relevant parts above.

The portion 42 may include one or more boards, each of which may include one or more processors and one or more memories, the processors being configured to read and execute programs in the memories to implement baseband processing functions and control of the network device. If there are multiple boards, the boards can be interconnected to increase processing power. As an alternative implementation manner, the multiple boards may share one or more processors, or the multiple boards may share one or more memories, or the multiple boards may share one or more processors at the same time.

For example, the sending module 43 is configured to perform the functions of the network device in any of the embodiments shown in fig. 3-5.

The embodiment of the application also provides a communication device, which comprises a processor, a memory, an input interface and an output interface, wherein the input interface is used for receiving information from other communication devices except the communication device, the output interface is used for outputting information to other communication devices except the communication device, and the processor calls a computer program stored in the memory to realize the method as shown in any one of fig. 3, fig. 4 or fig. 5.

In one possible design, the communication device may be a chip or a device comprising a chip implementing the method of any of fig. 3 or fig. 4 or fig. 5.

Embodiments of the present application also provide a computer readable storage medium, in which a computer program is stored, which when executed, implements a method as in any one of fig. 3 or fig. 4 or fig. 5.

The embodiment of the application also provides a chip, the chip comprises at least one processor and an interface, the processor is used for reading and executing the instructions stored in the memory, and when the instructions are executed, the chip is caused to execute the method shown in any one of fig. 3, fig. 4 or fig. 5.

Embodiments of the present application also provide a computer program product comprising instructions which, when executed on a computer, cause the method of any of fig. 3 or fig. 4 or fig. 5 to be performed.

The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the embodiments of the present application. In addition, each network element unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be realized in the form of hardware or in the form of software network element units.

The integrated units described above, if implemented in the form of software network element units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be a contributing part in essence, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a cloud server, or a network device, etc.) to perform all or part of the steps of the above-described method of the various embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes. While the invention has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions of equivalents may be made and equivalents will be apparent to those skilled in the art without departing from the scope of the invention. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method of frequency offset estimation, the method being applied to a first device, the method comprising:

receiving first information from a second device; the first information includes at least one of: a primary synchronization signal, a secondary synchronization signal, and service data;

according to the symbol where the first information is located, carrying out N times of cyclic prediction to obtain a first frequency offset estimation value; n is an integer greater than or equal to 1;

data demodulation is carried out according to the first frequency offset estimation value;

the ith cyclic prediction in the N cyclic predictions comprises the following steps that i is an integer greater than or equal to 1 and less than or equal to N:

performing frequency offset compensation on a symbol where first information is located in the i-1 th cyclic prediction according to a frequency offset estimated value obtained in the i-1 th cyclic prediction, so as to obtain the symbol where the first information is located in the i-1 th cyclic prediction;

and predicting to obtain a frequency offset estimation value during the ith cyclic prediction according to the symbol where the first information is located during the ith cyclic prediction.

2. The method of claim 1 wherein the frequency offset estimate for the ith cyclic prediction is an integer multiple of the frequency offset estimate if i is equal to 1.

3. A method according to claim 1 or 2, characterized in that,

N is the maximum number of cycles; or alternatively, the first and second heat exchangers may be,

the first frequency offset estimation value is smaller than a preset first threshold value; or alternatively, the first and second heat exchangers may be,

the difference between the frequency offset estimation value during the i-1 th cycle prediction and the frequency offset estimation value during the i-1 th cycle prediction is smaller than a preset second threshold value.

4. A method according to any one of claims 1-3, characterized in that the method further comprises:

and sending the capability information of the first device to the second device, wherein the capability information of the first device is used for indicating that the first device supports integral multiple frequency offset estimation.

5. The method of any of claims 1-4, wherein the frequency offset estimation method is performed by a network model that is trained by:

acquiring a symbol where the second information is located; the second information includes at least one of: a primary synchronization signal, a secondary synchronization signal, and service data;

vectorizing the symbol where the second information is located to obtain a feature vector;

inputting the feature vector into the network model, and predicting to obtain a frequency offset estimation value corresponding to the feature vector;

and adjusting model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector so as to train the network model.

6. The method of claim 5, wherein adjusting model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector to train the network model comprises:

obtaining a first loss according to the frequency deviation estimation value corresponding to the feature vector and the pre-marked frequency deviation corresponding to the feature vector;

and according to the first loss, adjusting model parameters of the network model to train the network model.

7. The method of claim 6, wherein the first penalty satisfies the following equation:

wherein loss is the first loss, R is an integer greater than or equal to 1 and less than or equal to the preset training times, t _i For the pre-noted frequency deviation, y, corresponding to the feature vector _i And the frequency offset estimation value corresponding to the characteristic vector is obtained.

8. A communication device is characterized in that the communication device is a first device, the first device comprises a transceiver module and a processing module,

the transceiver module is used for receiving the first information from the second equipment; the first information includes at least one of: a primary synchronization signal, a secondary synchronization signal, and service data;

The processing module is used for:

9. The apparatus of claim 8 wherein the frequency offset estimate for the ith cyclic prediction is an integer multiple of the frequency offset estimate if i is equal to 1.

10. The device according to claim 8 or 9, wherein,

n is the maximum number of cycles;

or, the first frequency offset estimation value is smaller than a preset first threshold value; or alternatively, the first and second heat exchangers may be,

11. The device according to any one of claims 8-10, wherein,

the transceiver module is further configured to send capability information of the first device to the second device, where the capability information of the first device is used to indicate that the first device supports frequency offset estimation with integer multiple.

12. The apparatus of any of claims 8-11, wherein the processing module is further configured to:

inputting the feature vector into a network model, and predicting to obtain a frequency offset estimation value corresponding to the feature vector;

13. The apparatus of claim 12, wherein the processing module, when adjusting the model parameters of the network model according to the frequency offset estimation value corresponding to the feature vector to train the network model, is configured to:

14. The apparatus of claim 13, wherein the first penalty satisfies the following equation:

15. A communication device comprising a processor which invokes a computer program stored in a memory to implement a method according to any one of claims 1 to 7.

16. A computer readable storage medium, characterized in that a computer program is stored in the computer readable storage medium, which, when being executed, implements the method according to any of claims 1-7.

17. A chip comprising at least one processor and an interface, the processor being configured to read and execute instructions stored in a memory, which when executed, cause the chip to perform the method of any of claims 1-7.