Information processing method, related apparatus and storage medium

Abstract

The present application belongs to the technical field of wireless communications. Provided are an information processing method, a related apparatus and a readable storage medium. The method is applied to a first device, wherein the first device is deployed with a CSI encoder or decoder using an AI / ML model, and a second device is correspondingly deployed with a CSI decoder or encoder using the AI / ML model. The method comprises: correcting cached CSI of a first device, so as to maintain synchronization with cached-CSI information of a second device; and on the basis of the cached CSI that has been corrected and the next measured CSI, determining measured-CSI differential information, or on the basis of the cached CSI that has been corrected and the next recovered-CSI differential information, determining the next recovered CSI.

Description

Information processing method, related apparatus and storage medium TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, and particularly relates to an information processing method, related apparatus and storage medium. BACKGROUND

[0002] In the communication system framework of Massive-MIMO, the measurement and feedback of channel state information (CSI) is one of the core key technologies of the physical layer. Accurate measurement and feedback of CSI information can help gNB to implement the correct data modulation and coding mode, directly improving the spectral efficiency of the communication system.

[0003] CSI compression is a study item (SI) in 3GPP NR Rel-18. In a framework of bilateral artificial intelligence (AI) / machine-learning (ML) based on CSI compression, a first part of the architecture is implemented on a user equipment (UE), and a second part of the architecture is implemented on a base station node (e.g., a gNodeB (gNB) of a wireless carrier network). In the first part of the architecture, the UE can pre-process CSI input into a form suitable for compression, and then compress the pre-processed or unprocessed CSI into an abstract representation of semantic features of the CSI using an AI / ML based encoder. In the second part of the architecture, the base station node receives the abstract representation of the CSI as feedback from the UE. Then, the base station node uses an AI / ML based decoder to decompress the abstract representation to reconstruct the CSI.

[0004] Temporal CSI compression utilizes the time-domain correlation of CSI, and historical CSI information is buffered on both the UE and the gNB. Only the difference between the current CSI and the buffered historical CSI is compressed and decompressed in the current time slot, which further reduces the air interface load of CSI reporting compared with the traditional spatial-frequency CSIro compression method. However, there is an error before and after CSI compression and recovery, which will accumulate over time in the temporal CSI compression and cause greater error, i.e., error propagation, ultimately leading to the loss of step of the bilateral buffered CSI, causing the CSI information recovery failure on the gNB side; in addition, the loss of uplink control information (UCI) will also cause the CSI recovery failure on the gNB side. Therefore, how to improve the CSI recovery accuracy and reduce the communication overhead is the current research focus.

[0005] SUMMARY

[0006] Embodiments of the present application provide an information processing method, related apparatus and storage medium to improve CSI recovery accuracy and reduce uplink and downlink signaling overhead.

[0007] To achieve the above object, the present application adopts the following technical solutions:

[0008] The first aspect of the present application provides an information processing method applied to a first device, wherein the first device is deployed with a channel state information (CSI) encoder or decoder applying an artificial intelligence (AI) / machine learning (ML) model, and a second device is correspondingly deployed with a CSI decoder or encoder applying an AI / ML model, and the method comprises: correcting cached CSI of the first device to keep synchronization with cached CSI information of the second device; and determining measurement CSI difference information based on the corrected cached CSI and next measurement CSI, or determining next recovery CSI based on the corrected cached CSI and next recovery CSI difference information.

[0009] The second aspect of the embodiments of the present application further provides an information processing method applied to a second device, wherein the second device is deployed with a CSI decoder or encoder applying an AI / ML model, and a first device is correspondingly deployed with a CSI encoder or decoder applying an AI / ML model, and the method comprises: generating CSI synchronization information; and sending the CSI synchronization information to the first device.

[0010] The third aspect of the embodiments of the present application further provides a wireless communication device, comprising: a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the method according to any one of the above aspects.

[0011] The fourth aspect of the embodiments of the present application further provides a computer readable storage medium, wherein the computer readable storage medium comprises instructions, and when the instructions are run, the method according to any one of the above aspects is implemented. BRIEF DESCRIPTION OF DRAWINGS

[0012] FIG. 1a is a possible bilateral model framework diagram of CSI compression;

[0013] FIG. 1b is a possible time domain CSI compression flowchart;

[0014] FIG. 2a is a possible information processing method provided by the embodiments of the present application;

[0015] FIG. 2b is a possible information processing method based on base station generated CSI synchronization information provided by the embodiments of the present application;

[0016] FIG. 2c is a flowchart of a possible information processing method for generating CSI synchronization information based on UE, according to an embodiment of the present application;

[0017] FIG. 2d is a flowchart of a possible information processing method for generating CSI synchronization information based on reference CSI, according to an embodiment of the present application;

[0018] FIG. 3 is a storage schematic diagram of a possible wireless communication device, according to an embodiment of the present application. DETAILED DESCRIPTION

[0019] For the convenience of understanding, the related technologies involved in the embodiments of the present application will be described first.

[0020] 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. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0021] It should be understood that the term "and / or" herein is only to describe the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after it.

[0022] For the sake of clear description, some self-defined terms designed in the technical solutions of the present application are explained as follows in Table 1:

[0023] Table 1

[0024] In order to better understand the technical background of the present application, the framework of AI / ML-based CSI compression and the time-domain CSI compression process will be briefly described. Please refer to FIG. 1a, which is a possible bilateral model framework diagram of CSI compression. In the figure, w represents the measured CSI information; represents the recovered CSI information, which is the output of the decoder decoder. Therefore, w and The similarity determines the system performance of the AI / ML-based CSI compression. The higher the similarity, the closer the CSI information after compression and recovery by the AI / ML model to the accurate CSI information measured by the UE, and the better the system performance of the entire CSI compression. Conversely, the system performance of the entire CSI compression is poorer. In the figure, it is assumed that the current time is t-1, the historical measured CSI information at time t-1 is buffered at the UE side, the gNB sends the CSI-RS to the UE, the UE measures the measured CSI information of the current time slot according to the received CSI-RS at time t, and calculates the difference between the measured CSI information of the current time slot and the historical measured CSI information buffered at the UE to obtain the CSI difference information, which is input into the AI / ML model of the CSI generation part at the UE side for compression. The obtained compressed CSI difference information, i.e. Compressed CSI, is used for CSI uplink transmission; while at the gNB side, the buffered historical CSI information is obtained by adding the historical recovered CSI information at time t-2 and the recovered CSI difference information at time t-1 at the gNB side, and the recovered CSI difference information is obtained by recovering the received compressed CSI difference information by the AI / ML model of the CSI reconstruction part. The obtained recovered CSI difference information at time t-1 is added to the historical recovered CSI information to obtain the recovered CSI information at time t.

[0025] Referring again to FIG. 1b, it is a possible time-domain CSI compression flowchart. As shown in the figure, three time slots t=1 (slot#X), t=2 (slot#X+k) and t=3 (slot#X+2k) are equally spaced, and the CSI information measured by the UE based on the received CSI-RS signal at the three time slots is denoted as V1, V2 and V3, respectively. The traditional AI / ML-based CSI compression is to input V1, V2 and V3 into the AI / ML model of the CSI generation part (i.e. CSI Encoder) at the UE side for compression. For time-domain CSI compression, the input of the AI / ML model of the CSI generation part is the difference between the measured CSI information of the current time slot and the accumulated CSI information of the previous time slot buffered at the UE side. Taking time slot t=2 (slot#X+k) as an example, the measured CSI information of the current time slot is V2, and the historical CSI information accumulated at the previous time by the UE is W1, then the input of the AI / ML model of the CSI generation part is AV2=V2-W1.

[0026] Since the CSI information generally has strong time-domain correlation in the time domain, i.e., the CSI measurement information of the current time slot V2 is relatively close to the historical CSI information W1 accumulated by the UE at the last time, the non-zero term of the calculated difference AV2 is relatively small, and the non-zero term coefficient of the difference AV2 is very small for compression and representation of the difference AV2, such as representation by a Rel-16 Type II codebook or a Rel-17 Enhanced Type II codebook, and the amount of data represented by the codebook is small, which reduces the air interface load of CSI compression reporting. For AI / ML-based CSI compression, the input of the AI / ML model of the CSI generation part on the UE side is AV2, and then the compressed CSI information has a smaller data amount and a smaller air interface load after compression by the Encoder. c = Encoder(AV2).

[0027] On the gNB side, after receiving the compressed CSI data AV c , the AI / ML model (i.e., CSI Decoder) of the CSI reconstruction part on the gNB side decompresses and restores the CSI difference information, i.e., r = Decoder(AV c ).

[0028] After decompression, the gNB side obtains the restored CSI difference information at the current time, and the historical CSI information W'1 accumulated on the gNB side is still buffered, and the sum of the two is the restored CSI information of the current time slot, i.e.,

[0029] The restored CSI information of the current time slot becomes the accumulated historical CSI information W'2 of the next time slot for recovery of the CSI compression information of the next time slot.

[0030] The above is a complete process of AI / ML-based CSI compression based on the time domain. As can be seen from the above process, the Encoder and the Decoder in the CSI compression scheme based on the time domain compress and decompress the difference between the CSI information measured by the UE at the current time and the historical CSI information accumulated by the UE. Due to the high correlation of the CSI information in the time domain, the difference information contains less information, i.e., the number of non-zero terms in the CSI information is small, which can effectively reduce the air interface load of CSI information reporting and improve the system capacity of the entire communication system.

[0031] However, there are some urgent problems in time-domain-based AI / ML CSI compression, including: problem one, the accumulation error of time-domain CSI compression causes the failure of CSI information recovery; problem two, the loss of UCI information causes the failure of CSI information recovery.

[0032] Embodiments of the present application will propose targeted solutions for the above problems.

[0033] For the formation of problem one, assume that the starting slot of the measurement of CSI is t, and every interval slots, the UE will perform a UE measurement and reporting action, that is, in slots t, t+At, t+2At, …, t+kAt, …, the UE will receive the CSI-RS signal sent by the gNB and perform the action of CSI measurement and compression reporting. In the traditional time-domain CSI compression scheme, the UE side caches the measured CSI information, and the gNB side caches the recovered CSI information, both sides update the cache independently, and there is a lack of a cache synchronization mechanism to align the CSI information cached by both sides.

[0034] At the starting slot t=0, since there is no historical CSI information in the cache, we need to initialize the cache. At this time, the UE side caches the measured CSI information at slot t=0, and this measured CSI information is processed by the AI / ML model to obtain compressed CSI information, which is reported to the gNB. After receiving the compressed CSI information, the gNB decompresses it using the AI / ML model to obtain the recovered CSI information. In particular, at the starting slot t, the cached CSI information at the UE side and the gNB side is 0, at this time, the input of the CSI generation part (CSI Encoder) at the UE side is the measured CSI information at the current slot, and the input of the CSI reconstruction part (CSI Decoder) at the gNB side is the compressed CSI information at the current slot. Based on the AI / ML model, the CSI compression is lossy compression, so there is a deviation between the recovered CSI information and the measured CSI information, denoted as e0, then,

[0035] Since the measured CSI information at the UE side at slot t and the recovered CSI information at the gNB side are the historical measured / recovered CSI information cached at the UE side and the gNB side at the next slot, that is, W0 and W'0, there is:

[0036] e0=W'0-W0;

[0037] So for the time slot t+Δt, the CSI Encoder on the UE side inputs the measured CSI differential information, and the CSI Decoder on the gNB side outputs the recovered CSI differential information, and there is also a deviation between the CSI compression and recovery, denoted as e1, so that

[0038] And in order to obtain the recovered CSI at the gNB, it is necessary to add to the historical recovered CSI information W'0, that is,

[0039] So the deviation between the measured CSI information and the recovered CSI information at time slot t+Δt is:

[0040] It is shown that the CSI compression / decompression deviation at time slot t is accumulated to time slot t+Δt, which affects the similarity of CSI compression and recovery at time slot t+Δt.

[0041] By analogy, we can obtain the deviation between the CSI information measured by the UE side (before compression) and the CSI information recovered by the gNB side through the AI / ML model of the CSI Decoder at time slot t+KΔt:

[0042] The compression / decompression error of each CSI measurement time slot is accumulated to time slot t+KΔt. We can consider that the compression / decompression error of each CSI measurement time slot is a random variable independent of each other, and the variance of the random variable is denoted as So we can get:

[0043] As can be seen, the variance of the error of CSI compression increases with time, indicating that the deviation between the measured CSI information and the recovered CSI information increases. Once the gap exceeds a certain threshold, the gNB will be unable to correctly recover the CSI information, causing the entire CSI measurement-compression-reporting process to fail.

[0044] As can be seen from the above analysis, the root cause of the CSI recovery failure is that the historical CSI cached by the UE side and the gNB side is not synchronized, resulting in a deviation between the two CSIs. In the technical solution of the present application, a CSI synchronization information is designed, so that the historical CSI cached by the UE side and the gNB side can be mutually compared every certain period of time, so as to keep consistent and avoid the accumulation of quantization and model inference errors at the gNB side.

[0045] For this purpose, please refer to FIG. 2a, which is a flowchart of a possible information processing method provided by the present application, specifically comprising:

[0046] 201a, the second device generates CSI synchronization information;

[0047] 202a, the second device sends the CSI synchronization information to the first device;

[0048] 203a, the first device corrects the cached CSI of the first device;

[0049] 204a, the first device determines the measurement CSI difference information based on the corrected cached CSI and the next measurement CSI, or determines the next recovery CSI based on the corrected cached CSI and the next recovery CSI difference information.

[0050] The second device generates CSI synchronization information and sends the CSI synchronization information to the first device to correct the cached CSI of the first device, so that the corrected cached CSI of the first device is synchronized with the cached CSI of the second device. The cached CSI is the CSI stored at the UE or base station side at a time other than the current time, which can be historical CSI such as the measurement CSI or recovery CSI of the last time slot, or can be a specific value, which is not limited here. The first device determines the measurement CSI difference information based on the corrected cached CSI and the next measurement CSI, or determines the next recovery CSI based on the corrected cached CSI and the next recovery CSI difference information, for example, determines the difference between the next measurement CSI and the corrected cached CSI as the measurement CSI difference information, or determines the sum of the next recovery CSI difference information and the corrected cached CSI as the next recovery CSI.

[0051] It is worth noting that in the embodiments of the present application, after the first device obtains the CSI synchronization information, the cached CSI of the first device is corrected based on the CSI synchronization information. In actual application, this correction step can be indirectly omitted, for example, the value obtained by directly adding the CSI synchronization information and the cached CSI is directly used to determine the measurement CSI difference information with the next measurement CSI, or the value obtained by directly adding the CSI synchronization information and the cached CSI is directly used to determine the next recovery CSI with the next recovery CSI difference information, that is, step 203a is an optional step.

[0052] In addition, the information processing method is applied in an AI / ML architecture, a first part of the architecture is implemented on the UE, and a second part of the architecture is implemented on a base station node such as a gNB of a wireless carrier network. The first device of the present application is deployed with a CSI encoder or decoder of the AI / ML architecture, and the second device is correspondingly deployed with a CSI decoder or encoder. Therefore, A, the first device is a UE, and the second device is a base station; or B, the first device is a base station, and the second device is a UE. The two scenarios will be described separately as follows.

[0053] A. The first device is a UE, and the second device is a base station;

[0054] Please refer to FIG. 2b, which is a flowchart of another possible information processing method provided by the embodiments of the present application.

[0055] 201b. The base station sends a CSI reference signal (channel state information-reference signal, CSI-RS) to the UE;

[0056] 202b. The UE measures the measurement CSI information of the current time slot according to the received CSI-RS;

[0057] 203b. The UE calculates the difference between the measurement CSI information of the current time slot and the historical measurement CSI information cached by the UE to obtain CSI differential information, and inputs the CSI differential information into a CSI encoder to obtain compressed CSI differential information;

[0058] 204b. The UE transmits the compressed CSI differential information to the base station in an uplink manner;

[0059] 205b. The base station inputs the compressed CSI differential information into a CSI decoder to obtain recovered CSI differential information, and adds the recovered CSI differential information to the historical recovered CSI cached by the base station to obtain the recovered CSI of the current time slot;

[0060] It should be noted that steps 201b-205b are prior art, and will not be described here.

[0061] 206b. The base station generates CSI synchronization information;

[0062] After obtaining the recovered CSI, the base station generates CSI synchronization information according to the recovered CSI, and the CSI synchronization information is used to correct the cached CSI on the UE side. How to generate the CSI synchronization information according to the recovered CSI will be described separately below.

[0063] a. Generating the CSI synchronization information according to the recovered CSI;

[0064] The most direct scheme for generating the CSI synchronization information is that the CSI synchronization information transmitted in a downlink manner is the recovered CSI output by the CSI decoder on the base station side. However, the data volume of the original CSI is large, which consumes a large amount of downlink transmission resources and affects the throughput of the communication system. At the same time, since the problem to be solved is the CSI synchronization problem, the data compression method using sub-band / layer / antenna port sampling is not the preferred method, which causes the CSI that is not sampled to be unable to be transmitted, thereby weakening the effect of CSI synchronization. Therefore, we need to perform effective quantization and compression processing on the recovered CSI output by the CSI decoder on the gNB side, and convert the recovered CSI into CSI synchronization information.

[0065] After the UE receives the synchronization information, the UE can correct the buffered CSI on the UE side using the synchronization information. It can be understood that the buffered CSI on the UE side at this moment (or this time) is basically the same as the recovered CSI on the gNB side after correction; at the next moment (or next time), the UE side uses the corrected buffered CSI as a reference to determine the measurement CSI difference and transmits the result to the gNB side, and the gNB side uses the measurement CSI difference information to determine the recovered CSI at the next moment (or next time), and then the recovered CSI is basically the same as the measurement CSI on the UE side. In this way, the accumulated historical error can be eliminated or reduced.

[0066] In the embodiments of the present application, the specific compression mode can include: a1. using recovered CSI to perform downlink CSI synchronization based on a non-codebook mode; a2. using recovered CSI to perform downlink CSI synchronization based on a codebook mode.

[0067] a1. using recovered CSI to perform downlink CSI synchronization based on a non-codebook mode;

[0068] For the downlink CSI synchronization information transmission scenario, the recovered CSI output by the CSI decoder on the base station side is used to construct the CSI synchronization information in a non-codebook mode. The recovered CSI output by the decoder is quantized and compressed in at least one of the following ways:

[0069] Mode 1: The subband recovered CSI of the n-th subband and the recovered CSI difference information of N-1 subbands are compressed according to the precoding mode to obtain the CSI synchronization information.

[0070] The bandwidth occupied by the UE is divided into N parts to form N subbands, each of which corresponds to a subband recovered CSI. In the N subbands, the subband recovered CSI of the n-th subband is respectively subtracted from the recovered CSIs of the other N-1 subbands to obtain corresponding N-1 subband recovered CSI difference information. Specifically, there are N subbands, and their number indexes are 0, 1, …, N-1 respectively. The subband recovered CSI of the n-th subband is subtracted from the subband recovered CSIs of the 1st, …, n-1th, n+1th, …, N-1th subbands to obtain the subband recovered CSI difference information of the 1st, …, n-1th, n+1th, …, N-1th subbands. Since the CSI often has relatively high frequency domain correlation, that is, the recovered CSIs of the subbands have relatively high similarity. Therefore, the obtained subband recovered CSI difference information has many amplitude coefficients of 0 or close to 0, that is, it has sparsity, which is beneficial to transmission by using some lossless compression encoding mode. Since the smaller the amplitude coefficient of the precoding matrix corresponding to the CSI difference information is, the greater the probability of its occurrence is, and vice versa. Therefore, the traditional equal-length coding (that is, all symbols are represented by equal-length bit strings) scheme will cause waste of bit information.

[0071] In the embodiments of the present application, a variable-length coding mode is adopted, in which low-quantized amplitude values are represented by short bit strings, and high-quantized amplitude values are represented by long bit strings. In this way, the shortest code word of the amplitude coefficient of the precoding matrix corresponding to the subband recovered CSI difference information can be constructed, and the data amount of the transmitted information is reduced. Optionally, Huffman coding, a typical variable-length coding, is adopted to construct the code word with the shortest average length of the header according to the probability of the character occurrence, so that optimal coding can be achieved. For example, assuming that the amplitude coefficient is quantized by 3 bits, the mapping relationship between the quantized value k of the amplitude coefficient and the amplitude coefficient p, and the probability of occurrence are shown in Table 2. ij

[0072] Table 2

[0073] According to the 3-bit quantization mode described in Table 2, the normalized precoding matrix amplitude coefficient is quantized into 8 amplitude levels. Then, after Huffman coding, the average number of bits represented by each quantized amplitude value is:

[0074] 1*0.75+2*0.12+3*0.07+5*0.02+(5+5+6+6)*0.01=1.52bit

[0075] That is, only 1.52 bits are needed to indicate an average quantized amplitude value, which is 49% higher than the average quantization bit number of 3 bits in the traditional equal-length coding mode, greatly improving the quantization and compression efficiency of the CSI synchronization message.​

[0076] Method 2, set part of the sub-band recovered CSI difference information to 0 and then perform compression processing;

[0077] On the basis of method 1, in order to further increase the sparsity of CSI, before the compression processing of the N-1 sub-band recovered CSI difference information according to the precoding mode, part of the sub-band recovered CSI difference information can be set to 0.

[0078] Specifically, since the CSI often has relatively high frequency domain correlation, that is, the CSI of each sub-band has relatively high similarity. Therefore, the obtained sub-band recovered CSI difference information has many amplitude coefficients of 0 or small values close to 0, that is, it has sparsity. For the small values close to 0, in the embodiments of the present application, they can be set to 0 by setting a threshold, thereby further increasing the sparsity. For the processed sub-band recovered CSI difference information, a variable-length coding method similar to method 1 can be used to quantize and compress the sub-band recovered CSI and the sub-band recovered CSI difference information. It should be noted that the threshold can be configured to the UE by the base station through RRC signaling.

[0079] Method 3, in the N-1 sub-band recovered CSI difference information, target sub-band recovered CSI difference information is selected for compression processing;

[0080] Due to the high time domain and frequency domain correlation of the CSI, the amplitude coefficients of the sub-band recovered CSI and the sub-band recovered CSI difference information of many sub-bands are 0 or small values close to 0. The non-zero term amplitude coefficient in the sub-band CSI is the most, which means that the time domain correlation of the CSI of the sub-band is the weakest. As a result, after the compression of the CSI encoder on the UE side and the recovery of the CSI decoder on the gNB side, the sub-band recovered CSI difference information tends to have the largest deviation in amplitude, which has the greatest impact on performance.

[0081] Therefore, in the embodiments of the present application, on the basis of method 1, the target sub-band recovered CSI difference information corresponding to the K target sub-bands with the most non-zero terms in the sub-band recovered CSI difference information is selected to generate the CSI synchronization information. Therefore, by synchronizing the target sub-band recovered CSI difference information to the UE, the deviation between the measurement CSI cached on the UE side and the recovered CSI cached on the gNB side can be corrected to the greatest extent. Further, the K target sub-band recovered CSI difference information can be determined by the following method:

[0082] Step 1, the base station configures the number of K, and the UE selects the sub-band according to the number of non-zero terms of the sub-band recovered CSI difference information;

[0083] Step 2, the UE selects, according to a protocol rule, for example, a gNB-configured threshold T_1 and a number M, when the number of non-zero terms in the amplitude coefficients of the CSI difference information recovered by a certain subband is greater than M, the absolute value of which is greater than T_1, the CSI difference information recovered by the subband is selected as the target subband-recovered CSI difference information.

[0084] Or, step 3, the UE selects, according to a protocol rule, for example, a gNB-configured threshold T_2, when the sum of the absolute values of the amplitude coefficients of the CSI difference information recovered by a certain subband is greater than T_2, the CSI difference information recovered by the subband is selected as the target subband-recovered CSI difference information.

[0085] After the target subband-recovered CSI difference information is selected according to the preset protocol rule, the subband-recovered CSI and the subband-recovered CSI difference information can be quantized and compressed in a variable-length coding manner similar to mode 1.

[0086] Mode 4, transmitting part or all of the subband-recovered CSI.

[0087] Optionally, when the subband-recovered CSI of part of the subbands is transmitted, the number of subband-recovered CSIs reported can also be screened by using the methods of steps 1-3 in mode 3 described above.

[0088] In summary, in the scenario of transmitting downlink CSI synchronization information based on non-codebook transmission, the base station can quantize and compress the recovered CSI output by the decoder to obtain CSI synchronization information by using at least one of modes 1-4, and then transmit the CSI synchronization information to the UE. The CSI synchronization information can include two parts:

[0089] The first part: an indicator of quantized and compressed subband-recovered CSI and / or subband-recovered CSI difference information, including but not limited to at least one of the following indicators: a quantization precision indicator, for example, 3 bits or 4 bits; an encoding mode indicator, for example, variable-length or fixed-length; and a second part data size, that is, the data size of the second part data is indicated in an explicit or implicit manner. It should be noted that when selecting part of the subbands, the selected subbands also need to be indicated, which can be achieved by bitmap / index.

[0090] The second part: the processed subband-recovered CSI information and subband-recovered CSI difference information.

[0091] It should be noted that in actual application, the data of the first part and the second part can be transmitted separately or together, which is not limited here. The data of the first part and the second part can be transmitted through downlink control information (DCI) / RRC / medium access control-control element (MAC-CE) signaling.

[0092] a2. Downlink CSI synchronization based on codebook mode using recovered CSI.

[0093] The recovered CSI output by the CSI decoder on the base station side is used to generate CSI synchronization information in a codebook mode. Similar to the non-codebook scenario, the recovered CSI output by the decoder can be quantized and compressed in any of the above four ways, which will not be repeated here.

[0094] It should be noted that, unlike the non-codebook scenario, the quantized and compressed downlink CSI synchronization information based on codebook transmission is transmitted to the UE. The CSI synchronization information includes two parts:

[0095] The first part: the indicator of quantized and compressed subband recovered CSI and / or subband recovered CSI differential information, including at least one of the following indicators: codebook type indication, used to indicate the codebook type of the precoding matrix indicator (PMI) data of the quantized and compressed subband recovered CSI and / or subband recovered CSI differential information based on codebook, such as Rel-15 Type II codebook, Rel-16 eType II codebook and Rel-17 FeType II codebook, etc.; spatial domain basis number indication, used to indicate the configuration of the selection number L of spatial domain basis in Rel-15 Type II codebook, Rel-16 eType II codebook and Rel-17 FeType II codebook, where for Rel-15 Type II codebook, L∈{2,3,4}, for Rel-16 eType II codebook and Rel-17 FeType II codebook, L∈{2,4,6}, so there are 2 / 3 / 4 / 6 combinations of the selection number of spatial domain basis, which can be indicated by 2 bits; frequency domain basis selection number ratio indication, used to indicate the configuration of the selection number M v of frequency domain basis in Rel-16 eType II codebook, while the selection number of frequency domain basis is proportional to the number N3 of subbands, and the conversion formula is where parameter p vThe ratio represents the number of frequency domain basis selection. In the parameter combination table TS38.214 Table 5.2.2.2.5-1, There are three values, which can be indicated by 2 bits; the second part data size is used to indicate the number of PMI data bits of the second part in a display or implicit manner.

[0096] The second part: the processed PMI.

[0097] It should be noted that the data of the first part and the second part can be transmitted separately or together, and the specific transmission is not limited here. The data of the first part and the second part can be transmitted by DCI / RRC / MAC-CE signaling.

[0098] 207b, the base station sends CSI synchronization information to the UE;

[0099] 208b, the UE corrects the cached CSI on the UE side;

[0100] 209b, the UE determines the measurement CSI difference information based on the corrected cached CSI and the next measurement CSI.

[0101] After the base station generates the CSI synchronization information, the base station sends the CSI synchronization information to the UE. The UE corrects the cached CSI on the UE side based on the CSI synchronization information, for example, adds the CSI synchronization information and the historical measurement CSI to obtain the corrected cached CSI. Then, the next measurement CSI and the corrected cached CSI are subtracted to determine the measurement CSI difference information.

[0102] In summary, the CSI synchronization information in the embodiment can be generated based on the recovered CSI or the recovered CSI difference information, and the compression method in the codebook transmission and non-codebook transmission scenarios is provided respectively.

[0103] B, the first device is a base station, and the second device is a UE.

[0104] Please refer to FIG. 2c, which is a flowchart of another possible information processing method provided by the embodiment of the application.

[0105] 201c, the base station sends CSI-RS to the UE;

[0106] 202c, the UE measures the measurement CSI of the current time slot according to the received CSI-RS;

[0107] 203c, the UE calculates the difference between the measurement CSI of the current time slot and the historical measurement CSI cached by the UE to obtain the CSI difference information, and inputs the CSI difference information into the CSI encoder to obtain the compressed CSI difference information;

[0108] 204c, the UE transmits the compressed CSI differential information to the base station in uplink;

[0109] 205c, the base station inputs the compressed CSI differential information into a CSI decoder to obtain recovered CSI differential information, and adds the recovered CSI differential information to the historical recovered CSI stored at the base station side to obtain recovered CSI of the current time slot;

[0110] It should be noted that steps 201c-205c are prior art, and will not be described here.

[0111] 206c, the UE generates CSI synchronization information;

[0112] After obtaining the measured CSI information, the UE generates CSI synchronization information according to the measured CSI, and the CSI synchronization information is used to correct the buffered CSI at the base station side.

[0113] After the gNB receives the synchronization information, the gNB can correct the buffered CSI at the gNB side using the synchronization information. It can be understood that the buffered CSI at the gNB side at this moment (or this time) is basically the same as the measured CSI at the UE side after correction; at the next moment (or next time), the gNB uses the corrected buffered CSI as a reference to determine the recovered CSI by combining the measured CSI differential information transmitted by the UE side, and then the recovered CSI is basically the same as the measured CSI at the UE side. In this way, the accumulated historical error can be eliminated or reduced.

[0114] The following will describe how to generate CSI synchronization information according to measured CSI.

[0115] c. generating CSI synchronization information according to measured CSI;

[0116] Similar to a at the base station side, the measured CSI at the UE side also needs to be effectively quantized and compressed to convert it into CSI synchronization information. In the embodiments of the present application, the specific compression method can include: c1. using measured CSI to perform uplink CSI synchronization based on a non-codebook mode; and c2. using measured CSI to perform uplink CSI synchronization based on a codebook mode.

[0117] c1. using measured CSI to perform uplink CSI synchronization based on a non-codebook mode;

[0118] For the uplink CSI synchronization information transmission scenario, the measured CSI is constructed into CSI synchronization information in a non-codebook mode. The measured CSI output by the encoder is quantized and compressed in at least one of the following ways:

[0119] Manner 1, according to the precoding manner, the subband measurement CSI corresponding to the n-th subband and the N-1 subband measurement CSI difference information are compressed to obtain the CSI synchronization information, and the subband measurement CSI difference information is the difference value of the subband measurement CSI corresponding to two subbands at the same time;

[0120] The bandwidth occupied by the UE is divided into N parts to form N subbands, and each subband corresponds to a subband measurement CSI. In the N subbands, the subband measurement CSI of the n-th subband is subtracted from the subband measurement CSI of the other N-1 subbands to obtain the corresponding N-1 subband measurement CSI difference information. Specifically, there are N subbands, and their number indexes are 0, 1, …, N-1 respectively. The subband measurement CSI of the n-th subband is transmitted, and the subband measurement CSI of the 1st, …, n-1th, n+1th, …, N-1th subband is subtracted from the subband measurement CSI of the n-th subband to obtain the subband measurement CSI difference information of the 1st, …, n-1th, n+1th, …, N-1th subband. Since the CSI often has a relatively high frequency domain correlation, that is, the measurement CSIs of the subbands have a relatively high similarity. Therefore, the obtained subband measurement CSI difference information has a large number of amplitude coefficients of 0 or a minimum value close to 0, that is, it has sparsity, which is beneficial to transmission by using some lossless compression encoding manner. Since the smaller the amplitude coefficient of the precoding matrix corresponding to the subband measurement CSI difference information is, the greater the probability of its occurrence is, and vice versa. Therefore, the traditional equal-length coding scheme (that is, all symbols are represented by equal-length bit strings) will cause waste of bit information.

[0121] Similar to the a1 of the base station side described above, in this scheme, the similar precoding manner is also used to compress the N-1 subband measurement CSI difference information corresponding to the n-th subband to obtain the CSI synchronization information, and details are not described here.

[0122] Manner 2, setting part of the subband measurement CSI difference information to 0 and then compressing;

[0123] In the c1, on the basis of the manner 1, to further increase the sparsity, part of the subband measurement CSI difference information can be set to 0 before the N-1 subband measurement CSI difference information is compressed according to the precoding manner.

[0124] Specifically, for the minimum value close to 0, in the embodiments of the present application, it can be set to zero by setting a threshold, so as to further increase the sparsity. For the processed subband measurement CSI difference information, the variable-length coding manner similar to the manner 1 in the c1 can be used to quantize and compress the subband measurement CSI and the subband measurement CSI difference information. It should be noted that the threshold can be configured to the UE by the base station through the RRC signaling.

[0125] Mode 3, among the N-1 subband measurement CSI differential information, select the target subband measurement CSI differential information for compression processing;

[0126] Due to the high time domain and frequency domain correlation of CSI, the amplitude coefficients of many subband measurement CSIs and subband measurement CSI differential information are 0 or close to 0. The most non-zero item amplitude coefficient in the subband CSI means that the time domain correlation of the subband CSI is the weakest, and as a result, after the compression of the CSI encoder on the UE side and the recovery of the CSI decoder on the gNB side, the subband measurement CSI differential information tends to have the largest deviation in amplitude, which has the greatest impact on performance.

[0127] Therefore, in the embodiments of the present application, based on mode 1 in c1, the target subband measurement CSI differential information corresponding to the K2 target subbands with the most non-zero items in the subband measurement CSI differential information is selected to generate CSI synchronization information, so that the target subband measurement CSI differential information is synchronized to the base station, which can correct the deviation between the recovered CSI cached on the base station side and the measurement CSI cached on the UE side to the greatest extent. Further, the K2 target subband measurement CSI differential information can be determined by the following method:

[0128] Step 1, the base station configures the number K2, and the UE selects the subband according to the number of non-zero items in the subband measurement CSI differential information;

[0129] Step 2, the UE selects according to the protocol rules, for example, the gNB configures a threshold T_5 and a number M2, when the number of non-zero items with an absolute value greater than T_5 in the amplitude coefficient of a certain subband measurement CSI differential information is greater than M2, the subband measurement CSI differential information is selected as the target subband measurement CSI differential information;

[0130] Or, step 3, the UE selects according to the protocol rules, for example, the gNB configures a threshold T_6, when the sum of the absolute values of the amplitude coefficients of a certain subband measurement CSI differential information is greater than T_4, the subband measurement CSI differential information is selected as the target subband measurement CSI differential information.

[0131] After selecting the target subband measurement CSI differential information according to the preset protocol rules, the subband measurement CSI and the subband measurement CSI differential information can still be quantized and compressed by using the variable length coding mode similar to mode c1.

[0132] Mode 4, transmit part or all of the subband measurement CSIs.

[0133] Optionally, in the embodiments of the present application, when the subband measurement CSI of the transmission part of the subband is measured, the number of subband measurement CSI reported can also be screened by using the method of steps 1-3 in the above-mentioned mode 3 of c1.

[0134] In summary, in the scenario of uplink CSI synchronization information based on non-codebook transmission, the UE can use at least one of modes 1-4 in c1 to quantize and compress the measurement CSI output by the encoder to obtain CSI synchronization information, and uplink the CSI synchronization information to the base station. The CSI synchronization information can include two parts:

[0135] The first part: the indicator of quantization and compression of subband measurement CSI and / or subband measurement CSI difference information, including but not limited to at least one of the following indicators: quantization mode indicator, used to indicate quantization accuracy; encoding mode indicator, used to indicate encoding mode; second part data size, i.e. using explicit or implicit way to indicate the data size of the second part data. It should be noted that when selecting part of the subband, the indication of the selected subband is also needed, which can be realized by bitmap / index.

[0136] The second part: the processed subband measurement CSI and subband measurement CSI difference information.

[0137] It should be noted that in actual application, the data of the first part and the second part can be transmitted separately or together, which is not limited here. DCI / RRC / MAC-CE signaling transmission.

[0138] c2. Uplink CSI synchronization based on codebook mode using measurement CSI.

[0139] The measurement CSI output by the CSI encoder on the UE side is used to generate CSI synchronization information in the codebook mode. Similar to the non-codebook scenario, the measurement CSI can be quantized and compressed in any of the four ways provided in c2 above, which will not be repeated here.

[0140] It should be noted that unlike the non-codebook scenario, the uplink CSI synchronization information based on codebook transmission is quantized and compressed, and the CSI synchronization information is uplinked to the base station. The CSI synchronization information includes two parts:

[0141] The first part: an indicator for quantization compression of sub-band measured CSI and / or sub-band measured CSI differential information, wherein at least one of the following indicators is included: codebook type indication, used to indicate the codebook type of PMI data after codebook-based quantization compression of sub-band measured CSI and / or sub-band measured CSI differential information; spatial domain basis number indication, used for configuration of the selection number L of spatial domain basis in Rel-15 Type II codebook, Rel-16 eType II codebook and Rel-17 FeType II codebook, wherein for Rel-15 Type II codebook, L∈{2,3,4}, for Rel-16 eType II codebook and Rel-17 FeType II codebook, L∈{2,4,6}, so there are four combinations of the selection number of spatial domain basis, 2 / 3 / 4 / 6, which can be indicated by 2 bits; frequency domain basis selection number ratio indication, used for configuration of the selection number M of frequency domain basis in Rel-16 eType II codebook, and the selection number of frequency domain basis is proportional to the number N3 of sub-bands, and the conversion formula is v wherein the parameter p v represents the ratio of the selection number of frequency domain basis. In the parameter combination table TS38.214 Table 5.2.2.2.5-1, there are three values, which can be indicated by 2 bits; second part data size, used to indicate the PMI data bit number of the second part in a display or implicit manner.

[0142] The second part: processed PMI.

[0143] It should be noted that the data of the first part and the second part can be transmitted separately or together, and the specific transmission is not limited here. The data of the first part and the second part can be transmitted by DCI / RRC / MAC-CE signaling.

[0144] 207c, the UE sends CSI synchronization information to the base station;

[0145] 208c, the base station corrects the buffered CSI on the base station side;

[0146] 209c, the base station determines the next recovery CSI based on the corrected buffered CSI and the next recovery CSI differential information.

[0147] After the UE generates the CSI synchronization information, the UE sends the CSI synchronization information to the base station. The base station corrects the buffered CSI on the base station side based on the CSI synchronization information, for example, adds the CSI synchronization information and the historical recovery CSI to obtain the corrected buffered CSI. Then, the next recovery CSI differential information and the corrected buffered CSI are added to determine the next recovery CSI.​

[0148] The above flowcharts shown in FIGs. 2b and 2c detail the way in which the base station can generate CSI synchronization information by restoring CSI and the way in which the UE can generate CSI synchronization information by measuring CSI. In the embodiments of the present application, the base station or the UE can also generate CSI synchronization information by using a known reference CSI to send to the opposite side. Referring to FIG. 2d, a flowchart of an information processing method for generating CSI synchronization information based on a reference CSI according to an embodiment of the present application is provided, which includes the following steps:

[0149] 201d, the second device acquires configuration information of the reference CSI;

[0150] 202d, the second device generates CSI synchronization information based on the reference CSI;

[0151] 203d, the second device sends the CSI synchronization information to the first device;

[0152] 204d, the first device corrects the cached CSI of the first device;

[0153] 205d, the first device determines measurement CSI difference information based on the corrected cached CSI and next measurement CSI, or determines next restored CSI based on the corrected cached CSI and next restored CSI difference information.

[0154] The reference CSI can be information known to both the first device and the second device. In this embodiment, the CSI synchronization information is generated based on the reference CSI, for example, the difference between the current measurement / restored CSI and the reference CSI can be used to generate the CSI synchronization information.

[0155] Taking the generation of the synchronization information by the gNB side as an example, after the UE receives the synchronization information, the UE can use the synchronization information and the reference CSI to sum up to correct the cached CSI on the UE side. It can be understood that the cached CSI on the UE side at this moment (or this time) is basically the same as the restored CSI on the gNB side after correction; at the next moment (or next time), the UE uses the corrected cached CSI as a reference to determine the measurement CSI difference and transmits the result to the gNB side, and the gNB side uses the measurement CSI difference information to determine the restored CSI at the next moment (or next time), then the restored CSI is basically the same as the measurement CSI on the UE side. In this way, the accumulated historical error can be eliminated or reduced.

[0156] Taking the generation of the synchronization information at the UE side as an example, after the gNB receives the synchronization information, the gNB can use the synchronization information and the reference CSI to sum up to correct the buffered CSI at the gNB side. It can be understood that the buffered CSI at the gNB side at this moment (or this time) is basically the same as the measured CSI at the UE side after correction; at the next moment (or next time), the gNB uses the corrected buffered CSI as a reference to determine the recovered CSI in combination with the measurement CSI difference information transmitted by the UE side, and then the recovered CSI is basically the same as the measured CSI at the UE side. In this way, the accumulated historical error can be eliminated or reduced.

[0157] The configuration information of the reference CSI can be configured for the base station and then transmitted to the UE, and the configuration information includes but is not limited to the specific form or generation rule of the reference CSI, the period of CSI synchronization with the reference CSI, and the like.

[0158] When the first device is a base station and the second device is a UE, the UE generates the reference CSI according to the configuration information of the reference CSI, and the difference between the measured CSI of the current time slot and the reference CSI is used as the UE-side CSI difference information. The UE-side CSI difference information is compressed to obtain the CSI synchronization information. The specific compression processing manner is similar to the compression processing of the measured CSI difference information in the above 2c, and details are not repeated here. The UE transmits the CSI synchronization information to the base station to correct the buffered CSI of the base station, so that the base station determines the next recovered CSI based on the corrected buffered CSI and the next recovered CSI difference information.

[0159] When the first device is a UE and the second device is a base station, the base station generates the reference CSI according to the configuration information of the reference CSI, and the difference between the current time slot and the reference CSI is used as the base station-side CSI difference information. The base station-side CSI difference information is compressed to obtain the CSI synchronization information. The specific compression processing manner is similar to the compression processing of the recovered CSI in the above 2a, and details are not repeated here. The base station transmits the CSI synchronization information to the UE to correct the buffered CSI of the UE, so that the UE determines the measured CSI difference information based on the corrected buffered CSI and the next measured CSI.

[0160] It should be noted that in the technical solutions of the present application involved in FIGS. 2a-2d, the CSI synchronization message is generated on one side and sent to the other side to achieve the purpose of synchronizing the buffered CSI on the UE side and the base station side. In the embodiments of the present application, the purpose of synchronizing the buffered CSI on the UE side and the base station side can also be achieved by configuring information. Specifically, based on first configuration information, the buffered CSI of the first device is periodically reset to a predetermined value, the first configuration information includes an interval period and a predetermined value, and the predetermined value is 0 or other values, wherein the first configuration information is shared by the first device and the second device; or, based on second configuration information, the buffered CSI of the first device is reset to a predetermined value, the second configuration information includes a reset trigger condition and a predetermined value, and the second configuration information is shared by the first device and the second device. That is, the first device and the second device can use a fixed period, or use an event indication trigger, to initialize the buffered CSI on both sides at intervals, or reset to a certain predetermined value, which can also achieve the synchronization of the buffered CSI on both sides. In the embodiments of the present application, the predetermined value or the initialized buffered CSI on both sides can be considered as the CSI synchronization information.

[0161] In summary, in the embodiments of the present application, the CSI synchronization information can be: (1) information obtained by the base station through compression / quantization processing to recover the CSI or the base station side CSI difference information; (2) information obtained by the UE through compression / quantization processing to measure the CSI or the UE side CSI difference information; (3) information generated by the base station and the UE side based on configuration information. It should be noted that the CSI synchronization information obtained in the above three cases can be combined into one information, for example, the CSI synchronization information includes two parts, part 1 is the related configuration for compression / quantization processing, and part 2 is the measured CSI or recovered CSI after processing, or the CSI synchronization information can also be generated based on different cases.

[0162] The above explains how to achieve the synchronization of the buffered CSI on the base station side and the UE side through the CSI synchronization message. In actual scenarios, the causes of the out-of-sync of the buffered CSI on both sides mainly include: (1) due to time accumulation error; (2) caused by accidental events, such as UCI loss, etc. Therefore, for these two reasons, the present application also provides a trigger mechanism for CSI synchronization information configuration, that is, what opportunity or what condition will trigger the synchronization of the buffered CSI on the base station side and the UE side, including: cause (1) the buffered CSI on both sides is out of sync due to time accumulation error, which can be triggered by periodic synchronization; cause (2) the buffered CSI on both sides is out of sync due to accidental events, which can be triggered by event-based synchronization. The following will be described in detail:

[0163] That is, on the basis of the flowchart shown in FIG. 2a, the second device sending CSI synchronization information to the first device includes: one, the second device periodically sends CSI synchronization information to the first device based on a preset period; two, the second device sends CSI synchronization information to the first device based on a preset event.

[0164] One, periodically sending CSI synchronization information to the first device;

[0165] For better understanding of the present scheme, taking the first device as UE and the second device as base station as an example for illustration.

[0166] Firstly, due to the sparsity of differential CSI information, the volume of the encoded CSI information is much smaller than that of the original CSI information. Considering that the single-time CSI compression error introduced by the AI model within a single time slot is small, it usually does not cause the CSI information recovery failure at the base station side, and the overhead can be further reduced by reducing the transmission frequency of the CSI synchronization information, that is, the transmission of the CSI synchronization information does not have to be consistent with the frequency of CSI measurement reporting, but the base station can issue a CSI synchronization information to the UE after multiple CSI compression-reporting-decompression operations. This multiple can be configured through RRC or MAC-CE messages, which is applicable to the periodic CSI measurement reporting case.

[0167] Periodic interaction: the base station implements the interaction period of CSI synchronization information through RRC signaling configuration, which can be independently configured or associated with CSI report. For example, it can be specified that every k CSI reports are sent, and then one CSI synchronization information interaction is performed, where k >= 1.

[0168] Therefore, by adopting the periodic transmission of CSI synchronization information, the problem of out-of-sync of the cached CSI at both sides due to time accumulation error can be greatly solved, and the communication overhead can be further reduced.

[0169] Optionally, in order to achieve the purpose of reducing communication overhead, in the embodiments of the present application, the interaction of CSI synchronization information can also be performed in a semi-persistent manner:

[0170] Dynamic interaction: the base station configures a set of reporting configurations (reporting interval / reporting time slot / reporting behavior duration, etc.) for CSI synchronization information interaction through RRC signaling, and optionally, MAC-CE or DCI is used to activate a certain configuration; or directly use RRC to configure a reporting configuration;

[0171] Base station / UE on-demand interaction: If the CSI synchronization information interaction depends on the CSI reporting mechanism, the base station / UE can determine whether to include CSI synchronization information interaction in the CSI report. In this case, the UE needs to inform the base station which CSI report contains CSI synchronization information and which does not. Specifically, there are two ways to indicate:

[0172] - Explicit indication. For example, include a 1-bit indicator in the CSI report, 0 indicates that it contains, and 1 indicates that it does not.

[0173] - Implicit indication. For example, by indicating different formats of resource requests such as resource scheduling (scheduling request, SR) corresponding to the CSI report, or by different scrambling methods for the CSI report.

[0174] In summary, to reduce the overhead of CSI synchronization message interaction, the frequency of CSI synchronization message transmission can be reduced, and three implementation methods of periodicity, dynamic configuration, and on-demand interaction between the two sides can be used to achieve the purpose of reducing overhead.

[0175] II. Sending CSI synchronization information to the first device based on a preset event trigger.

[0176] In the communication process between the base station and the UE, various occasional events may cause the buffered CSI on both sides to be out of step. For example, in a time-domain-based CSI compression scheme, after the AI / ML model of the CSI generation part on the UE side outputs the compressed CSI difference information, it is encapsulated in UCI information and transmitted to the base station through the PUCCH / PUSCH channel. However, this transmission process is not necessarily reliable, such as poor channel quality of the PUCCH / PUSCH uplink channel, inability to establish a connection, or too much resource occupation of the downlink transmission channel, resulting in insufficient resources for the uplink channel to transmit CSI. These reasons can cause UCI information loss, resulting in the compressed CSI data being unable to be transmitted to the base station, and ultimately leading to the failure of the base station's CSI recovery. Therefore, UCI information loss can also cause the buffered CSI on the UE side and the base station side to be out of step. At this time, the UE-side buffered CSI is the measurement CSIW K at time slot t+KΔt, while the base station-side buffered CSI is the recovered CSIW K-1 at time slot t+(K-1)Δt.

[0177] It can be seen that in addition to the cumulative error, there is also a time-domain CSI difference information ΔV k-1 , which is usually much larger than e k-1 . kTherefore, the gap between the UE side and the base station side of the cached CSI will rapidly increase, causing the CSI recovery failure of the base station side at the next moment. In summary, due to the problem caused by the UCI loss event, the time is accidental, and therefore the event-triggered solution is suitable.

[0178] In view of the fact that UCI loss is accidental and cannot be predicted in advance, in the embodiments of the present application, the preset event includes UCI loss as a condition for triggering the sending of CSI synchronization information. The UCI loss can be discussed based on two possibilities of base station side triggering and UE side triggering.

[0179] 1. Base station side triggering:

[0180] That is, the base station side determines UCI loss and its processing mechanism: since the UCI resource is configured by the base station, if the base station does not receive the reported CSI data on the specified time slot and RE resource, or fails to parse the CSI data, it can be considered that the UCI loss event has occurred. Once the UCI loss event occurs, the base station cannot obtain any recovery CSI differential information of the current time slot. When the base station side determines UCI loss, the following two dimensions need to be processed: (1) due to the lack of recovery CSI differential information of the current time slot, the base station can only choose to use the CSI at t-1 time for data precoding in the subsequent physical downlink shared channel (PDSCH) / physical downlink control channel (PDCCH) transmission. (2) Since the UE side has updated the cache, the base station side cannot update the cache, and the base station needs to trigger the cache CSI synchronization mechanism. At this time, the base station can choose to generate the CSI synchronization information by itself and send it through RRC signaling; if the base station chooses to trigger the CSI synchronization by the UE, it needs to send a trigger message to trigger the UE to send the CSI synchronization information. It should be noted that the configuration of the CSI synchronization information can be configured in the trigger message or pre-configured by other messages.

[0181] 2. UE side triggering:

[0182] UCI loss and its processing mechanism: the UE side determines the UCI loss mainly based on whether the CSI reporting resources are sufficient. In the existing protocol, when the transmission resources are insufficient, UCI multiplexing is performed, that is, different UCIs are multiplexed into the same uplink channel for transmission. Further, when different UCIs collide in the same uplink channel, the UCI information needs to be selectively discarded according to the priority of the CSI report. Since the CSI synchronization information has two parts, the first part (an indicator for quantization compression of the recovered CSI subband information and / or the recovered CSI subband differential information) and the second part (processed recovered CSI subband information and recovered CSI subband differential information, or recovered CSI subband information), when UCI multiplexing and resource collision occur, if the priority of the CSI synchronization information is lower than that of other UCI information (such as SR, HARQ ACK, CSI report, etc.), the existing protocol can choose to discard the second part of the information, and activate a special format of the first part of the information, which contains a message to the base station that the CSI synchronization UCI (for periodic reporting) does not contain valid CSI synchronization information. Therefore, when the base station receives this special format of CSI synchronization information, it does not need to update the cached CSI information, and until a new CSI synchronization information is received, the recovered CSI is obtained by using the existing cached CSI and the received compressed CSI differential information to perform subsequent PDSCH / PDCCH transmission.

[0183] Optionally, in the embodiments of the present application, in addition to UCI information loss, the preset event can also be that the model monitoring result does not meet the preset condition, to trigger the cached CSI synchronization. For example, when performing time domain CSI compression, if the model monitoring result does not meet the expectation, the traditional processing method is to directly perform the model switching process; since the model performance deterioration can also be caused by the mismatch of the input of the double-sided model (that is, the cached CSI is not synchronized), therefore, the CSI cache synchronization can also be introduced before the model switching to improve the existing mechanism.

[0184] Therefore, in the embodiments of the present application, periodic triggering or event indication triggering can be used to trigger the synchronization of the cached CSI on both sides.

[0185] In addition, the base station and the UE how to generate the CSI synchronization information is detailed through FIG. 2b and FIG. 2c, and in the flowchart shown in FIG. 2a, the second device generates the CSI synchronization information and then sends the CSI synchronization information to the first device. Considering that the size of the CSI synchronization information obtained after compression processing is still very large, in order to achieve the effect of balancing the load, the CSI synchronization information can be transmitted in steps. For this purpose, the embodiment of the present application also provides a method for transmitting the CSI synchronization information, that is, the second device sending the CSI synchronization information to the first device in FIG. 2a includes: transmitting the plurality of CSI synchronization sub-information contained in the CSI synchronization information to the second device in steps, and the CSI synchronization sub-information is divided based on a preset division rule. The preset division rule includes the following two kinds:

[0186] 1) Based on the subband number, all subbands corresponding to the CSI synchronization information are divided into M subband groups, each subband group includes k subbands, and the M subband groups correspond to M CSI synchronization sub-information; specifically,

[0187] Step 1, based on the subband number, all subbands are equally divided into M groups, each group including k subbands. The M subband groups correspond to the subband measurement CSI or subband recovery CSI of the M subband groups;

[0188] Step 2, using the subband measurement CSI or subband recovery CSI of the M subband groups to generate M CSI synchronization sub-information, and the M CSI synchronization sub-information is taken as a CSI synchronization information group;

[0189] Step 3, the M CSI synchronization sub-information is respectively sent in M time slots, and the sending interval is Q time slots configured by the gNB.

[0190] Step 4, in the mth time slot of the M time slots, the opposite end receives the CSI synchronization sub-information corresponding to the subband group m, and then updates the historical measurement CSI or historical recovery CSI corresponding to the subband group m in the buffer. When the M CSI synchronization sub-information corresponding to all M subband groups is sent, the comprehensive CSI buffer information synchronization is completed.

[0191] It should be noted that in the above CSI synchronization sub-information, in order for the opposite side to know which subband needs to be updated each time, the current subband group indication needs to be included in the CSI synchronization sub-information, for example: subband group index / bitmap, subband index / bitmap. It should be pointed out that the size of M can be configured to the UE by the base station through DCI or RRC signaling, and the size of M is related to the currently configured frame structure, and the size of M is related to the number of available resources.

[0192] For time division duplexing (TDD) scenarios, due to the non-uniformity of uplink and downlink resource allocation, the transmission of M CSI synchronization sub-information may present a non-uniform state, i.e., for a CSI synchronization information group, the time interval between each CSI synchronization sub-information is non-uniform. The time interval is configured by the base station and is related to the current frame structure configuration. The configuration of the time interval can be a specific slot / symbol index method, a bitmap method, or a configuration time window form, which is not limited in the present application.

[0193] 2) In all subbands, select K subbands corresponding to the CSI synchronization sub-information according to a first predetermined rule for local synchronization, and select all subbands corresponding to the CSI synchronization sub-information according to a second predetermined rule for global synchronization; specifically,

[0194] Local CSI synchronization:

[0195] Step 1, select part of the subbands by the first predetermined rule, and obtain the corresponding measurement CSI or recovered CSI according to the selected subbands, wherein the first predetermined rule can be: (1) random selection, for example, randomly selecting K subband CSI information for synchronization in each transmission, which makes each subband have the same probability of being synchronized in the long term; (2) selection according to the SINR / CQI calculated by the subband CSI, for example, selecting the K subbands with the largest SINR / CQI, because these subbands have the greatest impact on communication performance, improving them helps to improve the overall spectral efficiency; (3) selection according to the SINR / CQI calculated by the subband CSI, for example, selecting the K subbands with the smallest SINR / CQI to improve the performance of poor subbands and promote fairness;

[0196] Step 2, generate CSI synchronization information using the measurement CSI or recovered CSI obtained from the selected subbands;

[0197] Step 3, transmit the CSI synchronization information to the communication opposite end;

[0198] Global CSI synchronization:

[0199] Step 4, after a certain number of local synchronizations, trigger a global synchronization, i.e., select all subbands, generate CSI synchronization information and transmit, to realize the update of the historical measurement CSI or historical recovered CSI of all subbands in the cache.

[0200] It should be noted that the trigger of the global synchronization can be periodic, for example, the global synchronization is triggered once every M times of the local synchronization; or triggered by the base station or the UE through the downlink or uplink channel, at this time, it can be triggered through RRC / MAC-CE / DCI / UCI, etc. In addition, in the above CSI synchronization information, in order to make the opposite side know which subband CSI needs to be updated each time, the CSI synchronization information also needs to contain the indication of the current subband group, for example: subband group index / bitmap, subband index / bitmap, etc.

[0201] Therefore, by transmitting the CSI synchronization information which needs to be synchronized in the time dimension multiple times, the size of the CSI synchronization information transmitted each time is reduced, and the load balancing effect is achieved.

[0202] The above figure details the information processing method provided by the embodiments of the present application. Please refer to FIG. 3, which is a storage schematic diagram of a wireless communication device in the embodiments of the present application. The wireless communication device includes a processor and a memory, the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the method provided by any embodiment of the information processing method and any non-conflicting combination. The storage medium 20 of the wireless communication device in the embodiments of the present application stores instruction / program data 21, which is executed to realize the method provided by any embodiment of the information processing method and any non-conflicting combination. Among them, the instruction / program data 21 can form a program file and be stored in the above-mentioned storage medium 20 in the form of a software product, so that a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor executes all or part of the steps of the method of each embodiment of the present application. The aforementioned storage medium 20 includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes, or a computer, a server, a mobile phone, a tablet, and other terminal devices.

[0203] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the described device embodiment is merely illustrative. For example, the division of the units is only a logical function division. There can be another division manner for the actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0204] In addition, each function unit in the various embodiments of the present application can be integrated into a processing unit, or each unit can exist physically independently, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware, or in the form of a software function unit.

[0205] The above is merely an implementation of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent flow transformation made by using the content of the present application specification and drawings, or directly or indirectly applied to other related technical fields, is also included in the patent protection scope of the present application.

[0206] The above embodiments can be implemented, in whole or in part, by software, hardware (such as a circuit), firmware or any combination thereof. When implemented by software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the flow or function described in the embodiments of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through a wired (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. containing one or more available medium collections. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state disk.

[0207] It should be understood that the term "and / or" in this document is merely used to describe associated relationship, and it can mean three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. In addition, the character " / " in this document generally means that the associated objects before and after the " / " are in an "or" relationship, but can also mean an "and / or" relationship, which can be understood according to the context before and after.

[0208] In this application, "at least one" means one or more, and "multiple" means two or more. "At least one of the following" or the like means any combination of the items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be singular or plural.

[0209] It should be understood that the size of the sequence number of the above processes in various embodiments of the present application does not mean the order of execution, and the execution order of the processes should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0210] Those skilled in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0211] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0212] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be realized by other ways. For example, the above-described device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the shown or discussed objects can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0213] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, i.e., may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0214] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit.

[0215] The functions, if realized in the form of software functional units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or the part of the present application that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0216] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An information processing method applied to a first device, wherein the first device is equipped with a Channel State Information (CSI) encoder or decoder using an Artificial Intelligence (AI) / Machine Learning (ML) model, and a second device is correspondingly equipped with a CSI decoder or encoder using an AI / ML model, characterized in that, The method comprises: correcting the cached CSI of the first device; and determining measurement CSI difference information based on the corrected cached CSI and next measurement CSI; or determining next recovery CSI based on the corrected cached CSI and next recovery CSI difference information. The correction of the cached CSI of the first device comprises:

2. The method of claim 1, wherein, obtaining CSI synchronization information to correct the cached CSI based on the CSI synchronization information. The obtaining of the CSI synchronization information comprises: receiving the CSI synchronization information sent by the second device.

3. The method of claim 2, wherein, The correction of the cached CSI based on the CSI synchronization information comprises: adding the CSI synchronization information to historical recovery CSI to obtain the corrected cached CSI; or adding the CSI synchronization information to historical measurement CSI to obtain the corrected cached CSI.

4. The method of claim 3, wherein, The correction of the cached CSI of the first device comprises: periodically resetting the cached CSI of the first device to a predetermined value based on first configuration information, the first configuration information comprising an interval period and the predetermined value; or resetting the cached CSI of the first device to the predetermined value based on second configuration information, the second configuration information comprising a reset trigger condition and the predetermined value. The method comprises:

5. The method of claim 1, wherein, generating CSI synchronization information; sending the CSI synchronization information to the first device. The first device is a user equipment (UE), and the second device is a base station. The generation of the CSI synchronization information comprises:

6. An information processing method applied to a second device, wherein the second device is deployed with a channel state information, CSI, decoder or encoder applying an AI / ML model, and wherein the first device corresponds to be deployed with a CSI encoder or decoder applying an AI / ML model, the method comprising: receiving, by the second device, a CSI report from the first device; determining, by the second device, a CSI decoding / encoding scheme based on the AI / ML model; and performing, by the second device, CSI decoding / encoding based on the determined CSI decoding / encoding scheme. generating the CSI synchronization information according to recovery CSI. The generation of the CSI synchronization information according to the recovery CSI comprises: generating the CSI synchronization information according to the recovery CSI for non-codebook downlink transmission.

7. The method of claim 6, wherein, The generation of the CSI synchronization information according to the recovery CSI for non-codebook downlink transmission comprises:

8. The method of claim 7, wherein, determining subband recovery CSI difference information of other N-1 subbands according to a subband recovery CSI corresponding to an n-th subband, the N subbands being subbands obtained by dividing a bandwidth occupied by the UE into N parts, the subband recovery CSI difference information being a difference between subband recovery CSIs corresponding to two subbands at the same time; and performing compression processing on the subband recovery CSI of the n-th subband and the N-1 subband recovery CSI difference information according to a precoding manner to obtain the CSI synchronization information.

9. The method of claim 8, wherein, Before the compression processing on the subband recovery CSI of the n-th subband and the N-1 subband recovery CSI difference information according to the precoding manner, the method further comprises: setting, to 0, an amplitude coefficient having an absolute value less than a preset value among amplitude coefficients of the N-1 subband recovery CSI difference information.

10. The method of claim 9, wherein, Before the compression processing on the subband recovery CSI of the n-th subband and the N-1 subband recovery CSI difference information according to the precoding manner, the method further comprises: determining target subband recovery CSI difference information from the N-1 subband recovery CSI difference information according to amplitude coefficients of the subband recovery CSI difference information for compression processing. ​ 11. The method of claim 10, wherein, ​ ​ 12. The method of claim 10, wherein, ​ ​ 13. The method of claim 12, wherein, The determining the target sub-band recovery CSI differential information comprises: The sub-band recovery CSI differential information with the number of target amplitude coefficients greater than the second preset value is determined as the target sub-band recovery CSI differential information, and the absolute value of the target amplitude coefficient is greater than the third preset value.

14. The method of claim 12, wherein, The determining the target sub-band recovery CSI differential information comprises: The sub-band recovery CSI differential information with the sum of absolute values of each amplitude coefficient greater than the fourth preset value is determined as the target sub-band recovery CSI differential information.

15. The method of claim 9, wherein, The generating the CSI synchronization information according to the recovery CSI for non-codebook downlink transmission comprises: In the sub-band recovery CSI corresponding to the N sub-bands, a target sub-band recovery CSI is determined according to a preset rule to generate the CSI synchronization information.

16. The method of claim 9, wherein, The generating the CSI synchronization information according to the recovery CSI comprises: The CSI synchronization information is generated according to the recovery CSI for codebook downlink transmission.

17. The method of claim 16, wherein, The method further comprises: determining a precoding matrix indicator (PMI) configuration parameter corresponding to the generation of the PMI, the PMI configuration parameter comprising but not limited to the following information: codebook type indication information, spatial domain basis selection number indication information, and frequency domain basis selection number proportion indication information; sending the PMI configuration parameter to the UE.

18. The method of claim 16, wherein, The generating the CSI synchronization information according to the recovery CSI for codebook downlink transmission comprises: In the N sub-bands, each sub-band corresponds to a sub-band recovery CSI, and sub-band recovery CSI differential information of an n-th sub-band and other N-1 sub-bands is determined, the N sub-bands being obtained by dividing the bandwidth occupied by the UE into N parts, and the sub-band recovery CSI differential information being a difference value of the recovery CSIs corresponding to two sub-bands at the same time; performing compression processing on the sub-band recovery CSI of the n-th sub-band and the N-1 sub-band recovery CSI differential information according to the PMI to obtain the CSI synchronization information.

19. The method of claim 18, wherein, Before the compression processing on the sub-band recovery CSI of the n-th sub-band and the N-1 sub-band recovery CSI differential information according to the PMI, the method further comprises: setting an amplitude coefficient with an absolute value less than a preset value to 0 among the amplitude coefficients possessed by the N-1 sub-band recovery CSI differential information.

20. The method of claim 18, wherein, Before the compression processing on the sub-band recovery CSI of the n-th sub-band and the N-1 sub-band recovery CSI differential information according to the PMI, the method further comprises: determining target sub-band recovery CSI differential information from the N-1 sub-band recovery CSI differential information according to the amplitude coefficients of the sub-band recovery CSI differential information for compression processing.

21. The method of claim 20, wherein, The determining the target sub-band recovery CSI differential information comprises: The sub-band recovery CSI differential information with the number of target amplitude coefficients greater than the second preset value is determined as the target sub-band recovery CSI differential information, and the absolute value of the target amplitude coefficient is greater than the third preset value.

22. The method of claim 19, wherein, The determining the target sub-band recovery CSI differential information comprises: The sub-band recovery CSI differential information with the sum of absolute values of each amplitude coefficient greater than the fourth preset value is determined as the target sub-band recovery CSI differential information.

23. The method of claim 16, wherein, The generating the CSI synchronization information according to the recovery CSI comprises: In the subband recovery CSI corresponding to the N subbands, a target subband recovery CSI is determined according to a preset rule to generate the CSI synchronization information, and the CSI synchronization information further comprises a target subband indication parameter.

24. The method of claim 7, wherein, The generating the CSI synchronization information comprises: generating a reference CSI based on reference CSI configuration information; taking a difference between the recovery CSI and the reference CSI as base station side CSI difference information; generating the CSI synchronization information according to the base station side CSI difference information.

25. The method of claim 10, wherein, The reference CSI configuration information comprises one or more of the following information: a generation rule, a synchronization period, or a generation trigger condition, wherein the synchronization period is used to indicate that the reference CSI is generated periodically.

26. The method of claim 24, wherein, The generating the CSI synchronization information according to the base station side CSI difference information comprises: generating the CSI synchronization information according to the base station side CSI difference information for non-codebook downlink transmission.

27. The method of claim 24, wherein, The generating the CSI synchronization information according to the base station side CSI difference information for non-codebook downlink transmission comprises: determining a subband base station side CSI difference value between the n-th subband and other N-1 subbands, wherein the N subbands are obtained by dividing a bandwidth occupied by the UE into N parts, and the subband base station side CSI difference value is a difference between subband base station side CSI difference information corresponding to two subbands at the same time; performing compression processing on the subband base station side CSI difference information of the n-th subband and the N-1 subband base station side CSI difference values according to a precoding mode to obtain the CSI synchronization information.

28. The method of claim 6, wherein, The first device is a base station, and the second device is a user equipment (UE).

29. The method of claim 27, wherein, The generating the CSI synchronization information comprises: generating the CSI synchronization information according to the measurement CSI.

30. The method of claim 29, wherein, The generating the CSI synchronization information according to the measurement CSI comprises: generating the CSI synchronization information according to the measurement CSI for non-codebook uplink transmission.

31. The method of claim 30, wherein, The generating the CSI synchronization information according to the measurement CSI for non-codebook uplink transmission comprises: determining a measurement CSI subband difference information between the n-th subband and other N-1 subbands, wherein the N subbands are obtained by dividing a bandwidth occupied by the UE into N parts, and the measurement CSI subband difference information is a difference between subband measurement CSIs corresponding to two subbands at the same time; performing compression processing on the subband measurement CSI of the n-th subband and the N-1 measurement CSI subband difference information according to a precoding mode to obtain the CSI synchronization information.

32. The method of claim 31, wherein, Before the performing compression processing on the subband measurement CSI of the n-th subband and the N-1 measurement CSI subband difference information according to the precoding mode, the method further comprises: setting an amplitude coefficient with an absolute value less than a preset value to 0 among amplitude coefficients possessed by the N-1 subband measurement CSI difference information.

33. The method of claim 31, wherein, Before the subband measurement CSI of the n-th subband and the N-1 subband measurement CSI differential information are compressed according to the precoding mode, the method further comprises: According to the amplitude coefficient of each subband measurement CSI differential information, the target subband measurement CSI differential information is determined from the N-1 subband measurement CSI differential information for compression processing.

34. The method of claim 33, wherein, The determination of the target subband measurement CSI differential information comprises: The subband measurement CSI differential information with the number of target amplitude coefficients greater than the second preset value is determined as the target subband measurement CSI differential information, and the absolute value of the target amplitude coefficient is greater than the third preset value.

35. The method of claim 33, wherein, The determination of the target measurement CSI subband differential information comprises: The measurement CSI subband differential information with the sum of the absolute values of each amplitude coefficient greater than the fourth preset value is determined as the target subband measurement CSI differential information.

36. The method of claim 30, wherein, The generation of the CSI synchronization information according to the measurement CSI for non-codebook uplink transmission comprises: In the subband measurement CSI corresponding to the N subbands, the target subband measurement CSI is determined according to the preset rule to generate the CSI synchronization information, and the CSI synchronization information further comprises a target subband indication parameter.

37. The method of claim 29, wherein, The generation of the CSI synchronization information according to the measurement CSI comprises: The generation of the CSI synchronization information according to the measurement CSI for codebook uplink transmission.

38. The method of claim 28, wherein, The generation of the CSI synchronization information comprises: Generating a reference CSI based on reference CSI configuration information; The difference between the measurement CSI and the reference CSI is used as UE-side CSI differential information; The CSI synchronization information is generated according to the UE-side CSI differential information.

39. The method of claim 6, wherein, The sending of the CSI synchronization information to the first device comprises: Based on a preset trigger condition, the CSI synchronization information is sent to the first device, and the preset trigger condition comprises a preset period or a preset event.

40. The method of claim 39, wherein, The preset trigger condition is configured by transmission configuration information, and the transmission configuration information is carried in at least one of RRC signaling, medium access control-control element MAC-CE, downlink control information DCI, system message block SIB and random access response RAR.

41. The method of claim 39, wherein, The preset period is associated with a CSI report message, which is used to indicate that the CSI synchronization information is transmitted once every k times of CSI reporting, and k is an integer greater than 1.

42. The method of claim 40, wherein, The transmission configuration information comprises at least one of the following information: transmission interval, transmission time slot, transmission duration and transmission available window.

43. The method of claim 40, wherein, The transmission configuration information comprises carrying indication and indication mode, and the carrying indication is used to indicate whether the CSI synchronization information is carried in the CSI reporting message.

44. The method of claim 39, wherein, The preset condition comprises but is not limited to the following conditions: uplink control information UCI loss, model monitoring result not meeting the preset condition.

45. The method of claim 39, wherein, Before the CSI synchronization information is sent to the first device, the method further comprises: The UCI information loss is determined.

46. The method of claim 40, wherein, When the second device is a base station, the determination of the UCI information loss comprises: determining that the UCI information is lost when a CSI reporting message is not received on a designated time slot or resource element (RE) resource; or, determining that the UCI information is lost when a CSI decoder on the base station side in the AI / ML architecture fails to decode.

47. The method of claim 46, wherein, The sending of the CSI synchronization information to the first device includes: issuing the CSI synchronization information through RRC signaling; or, sending indication information to the UE, instructing the UE to send the CSI synchronization information.

48. The method of claim 45, wherein, When the second device is a UE, the determination that the UCI information is lost includes: determining that the UCI information is lost when a CSI reporting resource does not meet transmission requirements.

49. The method of claim 6, wherein, The sending of the CSI synchronization information to the first device includes: transmitting multiple CSI synchronization sub-information included in the CSI synchronization information to the second device in batches, the CSI synchronization sub-information being divided from the CSI synchronization information based on a preset division rule.

50. The method of claim 49, wherein, The preset division rule includes: dividing all subbands corresponding to the CSI synchronization information into M subband groups based on subband numbers, each of the subband groups including k subbands, and the M subband groups corresponding to M CSI synchronization sub-information.

51. The method of claim 50, wherein, The transmission to the second device in batches includes: transmitting the M CSI synchronization sub-information in M time slots at a configured transmission interval, each of the CSI synchronization sub-information including subband group indication information.

52. The method of claim 51, wherein, The value of M is related to a currently configured frame structure and the number of available resources.

53. The method of claim 49, wherein, The preset division rule includes: selecting CSI synchronization sub-information corresponding to K subbands in the all subbands according to a first predetermined rule to perform local synchronization; and selecting CSI synchronization sub-information corresponding to the all subbands according to a second predetermined rule to perform global synchronization.

54. The method of claim 53, wherein, The first predetermined rule includes, but is not limited to, at least one of the following rules: random selection, selection according to a signal to interference plus noise ratio (SINR) / channel quality indicator (CQI) value from large to small, and selection according to the SINR / CQI value from small to large.

55. The method of claim 54, wherein, The second predetermined rule includes, but is not limited to, at least one of the following rules: periodic triggering and instruction triggering, the periodic triggering being used to indicate that global synchronization is triggered once every m times of local synchronization, and the instruction triggering being used to indicate that global synchronization is performed through a received global triggering instruction.

56. A wireless communication device, comprising: A processor and a memory, the memory being used to store a computer program, and the processor being used to invoke and run the computer program stored in the memory to execute the method in any one of claims 1 to 55.

57. A readable storage medium characterized by, The computer readable storage medium includes instructions that, when executed, cause the method in any one of claims 1 to 55 to be implemented.

Images