Determination method and device of decoder, equipment, storage medium and program product

Abstract

The invention relates to a decoder determination method and device, equipment, a storage medium and a program product. The decoder determination method comprises the following steps: acquiring at least one decoder obtained by training; and determining a first decoder in the at least one decoder according to the decoding performance of the at least one decoder for the plurality of encoders, the decoding performance of the first decoder for the plurality of encoders satisfying a preset requirement. According to the embodiment of the invention, the decoder capable of accurately decoding data coded by different encoders can be provided.

Description

Decoder determination method, apparatus, device, storage medium and program product TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of communication, and particularly relates to a decoder determination method, apparatus, device, storage medium and program product. BACKGROUND

[0002] In a wireless communication system, a transmitter can encode channel state information (CSI), and then send the encoded CSI to a receiver. The receiver can receive the encoded CSI sent by the transmitter, and decode the encoded CSI to reconstruct the CSI.

[0003] SUMMARY

[0004] Embodiments of the present disclosure provide a decoder determination method, apparatus, device, storage medium and program product, which are used to determine a decoder that can accurately decode data encoded by multiple encoders.

[0005] According to a first aspect of embodiments of the present disclosure, a decoder determination method is provided, comprising:

[0006] obtaining at least one decoder trained;

[0007] determining a first decoder from the at least one decoder according to decoding performance of the at least one decoder on multiple encoders, wherein the decoding performance of the first decoder on the multiple encoders meets a preset requirement.

[0008] In embodiments of the present disclosure, the decoding performance of the first decoder on the multiple encoders meets the preset requirement, that is, the first decoder can accurately decode data encoded by multiple encoders. Through the embodiments of the present disclosure, the purpose of determining a decoder that can accurately decode data encoded by multiple encoders is achieved.

[0009] According to a second aspect of embodiments of the present disclosure, a decoder determination apparatus is provided, comprising:

[0010] an obtaining module configured to obtain at least one decoder trained;

[0011] a determining module configured to determine a first decoder from the at least one decoder according to decoding performance of the at least one decoder on multiple encoders, wherein the decoding performance of the first decoder on the multiple encoders meets a preset requirement.

[0012] According to a third aspect of embodiments of the present disclosure, a decoder determination device is provided, comprising one or more processors; wherein the processor is configured to execute the decoder determination method of the first aspect.

[0013] According to a fourth aspect of the embodiments of the present disclosure, a storage medium is provided, which stores instructions, when the instructions are executed on an electronic device, causing the electronic device to perform the determination method of the decoder according to the first aspect.

[0014] According to a fifth aspect of the embodiments of the present disclosure, a program product is provided, which includes a program and / or instructions, when the program and / or instructions are executed by an electronic device, causing the electronic device to perform the determination method of the decoder according to the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following describes the drawings required for the embodiments, and the following drawings are only some embodiments of the present disclosure, and do not specifically limit the protection scope of the present disclosure.

[0016] FIG. 1a is an exemplary schematic diagram of an architecture of a communication system related to an embodiment of the present disclosure.

[0017] FIG. 1b is an exemplary schematic diagram of an architecture of a communication system related to an embodiment of the present disclosure.

[0018] FIG. 2a is an exemplary flowchart of a determination method of a decoder according to an embodiment of the present disclosure.

[0019] FIG. 2b is an exemplary flowchart of a determination method of a decoder according to an embodiment of the present disclosure.

[0020] FIG. 2c is an exemplary flowchart of a determination method of a decoder according to an embodiment of the present disclosure.

[0021] FIG. 2d is an exemplary flowchart of a determination method of a decoder according to an embodiment of the present disclosure.

[0022] FIG. 2e is an exemplary flowchart of a determination method of a decoder according to an embodiment of the present disclosure.

[0023] FIG. 2f is an exemplary flowchart of a determination method of a decoder according to an embodiment of the present disclosure.

[0024] FIG. 2g is an exemplary flowchart of a determination method of a decoder according to an embodiment of the present disclosure.

[0025] FIG. 2h is an exemplary schematic diagram of a CNN-based decoder according to an embodiment of the present disclosure.

[0026] FIG. 3 is an exemplary structural schematic diagram of a determination apparatus of a decoder according to an embodiment of the present disclosure.

[0027] Figure 4a is an exemplary structural diagram of an electronic device provided according to an embodiment of the present disclosure.

[0028] Figure 4b is an exemplary structural diagram of a chip provided according to an embodiment of the present disclosure. Detailed Implementation

[0029] This disclosure provides a method, apparatus, device, storage medium, and program product for determining a decoder that can accurately decode data encoded by a variety of encoders.

[0030] In a first aspect, embodiments of this disclosure provide a method for determining a decoder, comprising:

[0031] Obtain at least one decoder obtained from training;

[0032] Based on the decoding performance of at least one decoder on multiple encoders, a first decoder is determined among at least one decoder, and the decoding performance of the first decoder on multiple encoders meets preset requirements.

[0033] In this embodiment, the decoding performance of the first decoder for multiple encoders meets preset requirements; that is, the first decoder can accurately decode data encoded by multiple encoders. Through this embodiment, the objective of determining a decoder capable of accurately decoding data encoded by multiple encoders is achieved.

[0034] In conjunction with some embodiments of the first aspect, in some embodiments, at least one decoder is trained on a dataset including channel features.

[0035] In the above embodiments, at least one decoder can be obtained by training the dataset, thus achieving the goal of obtaining at least one decoder.

[0036] In conjunction with some embodiments of the first aspect, in some embodiments, the channel features include at least one of the following:

[0037] Number of devices: This refers to the number of devices that send Channel State Information (CSI) when generating the dataset.

[0038] Number of transmission ports: This refers to the number of ports on the device that transmit CSI data.

[0039] Subband quantity, which refers to the number of subbands occupied by the device when transmitting CSI;

[0040] Number of in-phase and quadrature IQs;

[0041] Number of time slots, which refers to the number of time slots occupied by the device when transmitting CSI; or...

[0042] Random number of times, the random number of times is used to indicate the random number of times of randomly generating a data set.

[0043] In the above embodiment, the channel feature can include at least one of the number of devices, the number of transmission ports, the number of subbands, the number of I / Qs, the number of time slots, or the random number of times, achieving the purpose of determining the channel feature.

[0044] In combination with some embodiments of the first aspect, in some embodiments, the at least one decoder is obtained by training on the same data set, and the same data set is a set of the plurality of data sets.

[0045] In the above embodiment, the at least one decoder can be obtained by training on the same data set, achieving the purpose of obtaining the at least one decoder.

[0046] In combination with some embodiments of the first aspect, in some embodiments, the at least one decoder is obtained by training on different data sets, and the similarity between the different data sets is greater than or equal to a first threshold.

[0047] In the above embodiment, the at least one decoder can be obtained by training on at least one data set with a similarity greater than or equal to the first threshold, achieving the purpose of obtaining the at least one decoder.

[0048] In combination with some embodiments of the first aspect, in some embodiments, the decoding performance of the first decoder on the plurality of encoders satisfies a preset requirement, including at least one of the following:

[0049] The decoding performance of the first decoder is greater than or equal to a second threshold;

[0050] In order of decoding performance from high to low, the first decoder is the top K decoders in the at least one decoder, K being an integer greater than or equal to 1; or,

[0051] The decoding performance difference of the first decoder on the plurality of encoders is less than a third threshold.

[0052] In the above embodiment, the first decoder can be determined in the at least one decoder, achieving the purpose of determining a decoder that can accurately decode data encoded by a plurality of encoders.

[0053] In combination with some embodiments of the first aspect, in some embodiments, the at least one decoder is obtained by training on at least one data set.

[0054] In the above embodiment, the at least one decoder can be obtained by training on the data set, achieving the purpose of obtaining the at least one decoder.

[0055] In some embodiments of the first aspect, the first decoder is determined from the decoding performance of the at least one decoder on the plurality of encoders according to at least one of the following:

[0056] The first decoder is determined from the decoding performance of the at least one decoder on the plurality of encoders according to at least one of the following when the similarity between the at least one data set is greater than or equal to the first threshold value:

[0057] The performance similarity of the at least one decoder is determined when the similarity between the at least one data set is less than the first threshold value, and the first decoder is determined according to the performance similarity.

[0058] In the above embodiments, the first decoder can be determined from the at least one decoder according to the similarity between the at least one data set, achieving the purpose of determining the first decoder.

[0059] In some embodiments of the first aspect, the similarity between the at least one data set is a cosine similarity between the at least one data set.

[0060] In the above embodiments, the cosine similarity between the at least one data set can be determined as the similarity between the at least one data set, achieving the purpose of determining the similarity between the at least one data set.

[0061] In some embodiments of the first aspect, the method further comprises:

[0062] Obtaining decoding performance of a preset decoder on the encoded data of the at least one data set;

[0063] If the difference between the decoding performance corresponding to the at least one data set is less than a fourth threshold value, it is determined that the similarity between the at least one data set is greater than or equal to the first threshold value.

[0064] In the above embodiments, the similarity between the at least one data set can be determined according to the difference between the decoding performance corresponding to the at least one data set, achieving the purpose of determining whether the similarity between the at least one data set is greater than or equal to the first threshold value.

[0065] In some embodiments of the first aspect, the first decoder is determined from the decoding performance of the at least one decoder on the plurality of encoders according to at least one of the following:

[0066] The decoder in the at least one decoder that satisfies at least one of the following conditions is determined as the first decoder:

[0067] The decoding performance is greater than or equal to a second threshold value;

[0068] The first decoders are the top K decoders in the at least one decoder in descending order of decoding performance, K being an integer greater than or equal to 1; or

[0069] The decoding performance difference of the plurality of encoders is less than a third threshold.

[0070] In the above embodiment, the first decoders can be determined in the at least one decoder, and the purpose of determining the decoders that can accurately decode the data encoded by the plurality of encoders is achieved.

[0071] In combination with some embodiments of the first aspect, in some embodiments, determining the performance similarity of the at least one decoder comprises:

[0072] Determining the first performance of each decoder by the plurality of first data;

[0073] Determining the performance similarity of the at least one decoder according to the first performance of each decoder.

[0074] In the above embodiment, the first performance of each decoder can be determined, and the performance similarity of the at least one decoder can be determined according to the first performance of each decoder, and the purpose of determining the performance similarity of the at least one decoder is achieved.

[0075] In combination with some embodiments of the first aspect, in some embodiments, for any one decoder; determining the first performance of the decoder by the plurality of first data comprises:

[0076] Encoding each first data by the encoder corresponding to the decoder to obtain second data corresponding to each first data;

[0077] Decoding each second data corresponding to each first data by the decoder to obtain third data corresponding to each first data;

[0078] Determining the first performance of the decoder according to each first data and the third data corresponding to the first data.

[0079] In the above embodiment, the first performance of the decoder can be determined by the plurality of first data and the encoder corresponding to the decoder, and the purpose of determining the first performance of the decoder is achieved.

[0080] In combination with some embodiments of the first aspect, in some embodiments, determining the first performance of the decoder according to each first data and the third data corresponding to the first data comprises:

[0081] Determining a data similarity between each first data and the corresponding third data to obtain a plurality of first similarities;

[0082] Determining the first performance according to the plurality of first similarities.

[0083] In the above embodiment, the first performance of the decoder can be determined by the plurality of first similarities corresponding to the plurality of first data, and the purpose of determining the first performance of the decoder is achieved.

[0084] In some embodiments of the first aspect, the first decoder is determined according to the performance similarity, including:

[0085] When the performance similarity is greater than or equal to the preset performance threshold, the decoder that meets at least one of the following conditions in the at least one decoder is determined as the first decoder: the decoding performance is greater than or equal to the second threshold; the first decoder is the top K decoders in the at least one decoder in descending order of decoding performance, K is an integer greater than or equal to 1; or the decoding performance difference of the plurality of encoders is less than the third threshold;

[0086] Or,

[0087] When the performance similarity is less than the preset performance threshold, the first decoder is obtained by training the decoder by using the at least one data set.

[0088] In the above embodiment, the first decoder can be determined in the at least one decoder when the performance similarity is greater than or equal to the preset performance threshold, and the purpose of determining the first decoder is achieved.

[0089] In some embodiments of the first aspect, the at least one decoder obtained by training includes:

[0090] The at least one second device sends at least one decoder.

[0091] In the above embodiment, the at least one decoder can be obtained from the second device, and the purpose of obtaining the at least one decoder is achieved.

[0092] In some embodiments of the first aspect, the method further includes:

[0093] The method is performed by the first device; the number of the at least one decoder is M, M is an integer greater than 1; and the at least one decoder obtained by training includes:

[0094] The first device receives N decoders sent by the at least one second device;

[0095] Wherein, N is an integer greater than 1 and less than M; and M-N decoders in the M decoders except the N decoders are obtained by training of the first device.

[0096] In the above embodiment, the at least one decoder includes M decoders, the first device can obtain N decoders from the second device, and M-N decoders can be trained, so that the purpose of obtaining the at least one decoder is achieved.

[0097] In combination with some embodiments of the first aspect, in some embodiments, the method is performed by the first device; and the at least one decoder is trained by the first device.

[0098] In the above embodiment, the at least one decoder can be trained, so that the purpose of obtaining the at least one decoder is achieved.

[0099] In combination with some embodiments of the first aspect, in some embodiments, the method further includes:

[0100] The decoding performance of each decoder is determined by the at least one encoder.

[0101] In the above embodiment, the decoding performance of each decoder can be determined by the encoder, so that the purpose of determining the decoding performance of the decoder is achieved.

[0102] In combination with some embodiments of the first aspect, in some embodiments, for any one decoder; the decoding performance of the decoder is determined by the at least one encoder, including:

[0103] The second performance of the decoder is obtained by the at least one encoder respectively performing performance detection on the decoder.

[0104] The decoding performance is determined according to the at least one second performance.

[0105] In the above embodiment, the second performance of the decoder can be detected by the encoder, and the decoding performance of the decoder can be determined according to the second performance, so that the purpose of determining the decoding performance of the decoder is achieved.

[0106] In combination with some embodiments of the first aspect, in some embodiments, for any one encoder; the second performance of the decoder is obtained by the encoder performing performance detection on the decoder, including:

[0107] The fifth data is obtained by the encoder performing encoding processing on the fourth data.

[0108] The sixth data is obtained by the decoder performing decoding processing on the fifth data.

[0109] The second performance is determined according to the similarity between the fourth data and the sixth data.

[0110] In the above embodiment, the second performance of the decoder can be determined by the encoder and the fourth data, so that the purpose of determining the second performance of the decoder is achieved.

[0111] In some embodiments of the first aspect, in some embodiments, the similarity between the fourth data and the sixth data is a cosine similarity between the fourth data and the sixth data.

[0112] In the above embodiments, the cosine similarity between the fourth data and the sixth data can be determined as the similarity between the fourth data and the sixth data, achieving the purpose of determining the similarity between the fourth data and the sixth data.

[0113] In some embodiments of the first aspect, in some embodiments,

[0114] The decoding performance is an average value of the at least one second performance; or

[0115] The decoding performance is a maximum value in the at least one second performance; or

[0116] The decoding performance is a minimum value in the at least one second performance.

[0117] In the above embodiments, the decoding performance of the decoder can be determined according to the at least one second performance, achieving the purpose of determining the decoding performance.

[0118] In a second aspect, the embodiments of the present disclosure provide a determination apparatus of a decoder, comprising:

[0119] An acquisition module configured to acquire at least one decoder trained;

[0120] A determination module configured to determine a first decoder from the at least one decoder according to decoding performances of a plurality of encoders on the at least one decoder, the decoding performance of the first decoder on the plurality of encoders meeting a preset requirement.

[0121] In a third aspect, the embodiments of the present disclosure provide a determination device of a decoder, comprising:

[0122] One or more processors;

[0123] The processor is configured to execute the first aspect and the optional implementation manners of the first aspect.

[0124] In a fourth aspect, the embodiments of the present disclosure provide a storage medium, the storage medium stores instructions, when the instructions are executed on an electronic device, the electronic device executes the method described in the first aspect and the optional implementation manners of the first aspect.

[0125] In a fifth aspect, the embodiments of the present disclosure provide a program product, when the program product is executed by an electronic device, the electronic device executes the method described in the first aspect and the optional implementation manners of the first aspect.

[0126] In a sixth aspect, the embodiments of the present disclosure provide a computer program which, when running on a computer, causes the computer to perform the method described in the first aspect and the optional implementation manners of the first aspect.

[0127] In a seventh aspect, the embodiments of the present disclosure provide a chip or a chip system. The chip or the chip system includes processing circuitry configured to perform the method described in the first aspect and the optional implementation manners of the first aspect.

[0128] It can be understood that the determination apparatus of the decoder, the determination device of the decoder, the storage medium, the program product, the computer program, the chip or the chip system are all used to perform the method proposed in the embodiments of the present disclosure. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method, which will not be described here again.

[0129] The embodiments of the present disclosure propose a determination method, apparatus and device of a decoder, a storage medium and a program product.

[0130] In some embodiments, the determination method of the decoder and the terms such as the information processing method, the data processing method, the resource processing method and the communication method can be replaced with each other, and the determination apparatus of the decoder and the terms such as the information processing apparatus, the data processing apparatus, the resource processing apparatus and the communication apparatus can be replaced with each other.

[0131] The embodiments of the present disclosure are not exhaustive, but only illustrate some embodiments, and are not specific limitations on the protection scope of the present disclosure. In the case of no contradiction, each step in an embodiment can be implemented as an independent embodiment, and the steps can be combined arbitrarily, for example, the scheme after removing some steps in an embodiment can also be implemented as an independent embodiment, and the order of the steps in an embodiment can be exchanged arbitrarily, in addition, the optional implementation manners in an embodiment can be combined arbitrarily; in addition, the embodiments can be combined arbitrarily, for example, the steps of different embodiments or part or all of the steps of different embodiments can be combined arbitrarily, and an embodiment can be combined with the optional implementation manners of other embodiments.

[0132] In the embodiments of the present disclosure, the terms and / or descriptions between the embodiments are consistent if there is no special description and logical conflict, and can be referred to each other, and the technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0133] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments, and not as a limitation on the present disclosure.

[0134] In the embodiments of the present disclosure, an element expressed in singular form, such as "a", "an", "the", "said", "above", "preceding", "this", and the like, can represent "one and only one", or can represent "one or more", "at least one", and the like, unless otherwise specified. For example, in the case of using articles such as "a", "an", "the" in English, the noun after the article can be understood as a singular expression, or can be understood as a plural expression.

[0135] In the embodiments of the present disclosure, "plurality" refers to two or more.

[0136] In some embodiments, the terms "at least one of", "one or more of", "a plurality of", "multiple", and the like can be replaced with each other.

[0137] In some embodiments, the description modes such as "at least one of A, B", "A and / or B", "A in one case and B in another case", "in response to a case A, in response to a case B", and the like can include the following technical solutions according to the case: in some embodiments, A is executed regardless of B; in some embodiments, B is executed regardless of A; in some embodiments, A and B are selectively executed from A and B; in some embodiments, A and B are executed (A and B are both executed). When there are more branches such as A, B, C, and the like, it is similar to the above.

[0138] In some embodiments, the description modes such as "A or B" and the like can include the following technical solutions according to the case: in some embodiments, A is executed regardless of B; in some embodiments, B is executed regardless of A; in some embodiments, A and B are selectively executed from A and B (A and B are selectively executed). When there are more branches such as A, B, C, and the like, it is similar to the above.

[0139] The prefix words of "first", "second" and the like in the embodiments of the present disclosure are merely used to distinguish different description objects, and do not constitute limitation on the position, order, priority, quantity or content of the description objects. The description objects are described in the claims or embodiments, and should not be construed as redundant limitation because of the use of the prefix words. For example, the description object is "field", and the ordinal words before "field" in "first field" and "second field" do not limit the position or order between "fields", and "first" and "second" do not limit whether the "fields" modified thereby are in the same message or not, nor limit the order of "first field" and "second field". For another example, the description object is "level", and the ordinal words before "level" in "first level" and "second level" do not limit the priority between "levels". For another example, the quantity of the description object is not limited by the ordinal words, and can be one or more. For example, "first device", wherein the quantity of "device" can be one or more. In addition, the objects modified by different prefix words can be the same or different, for example, the description object is "device", and "first device" and "second device" can be the same device or different devices, and the types thereof can be the same or different; for another example, the description object is "information", and "first information" and "second information" can be the same information or different information, and the contents thereof can be the same or different.

[0140] In some embodiments, "including A", "containing A", "for indicating A", "carrying A" can be interpreted as directly carrying A, or indirectly indicating A.

[0141] In some embodiments, the terms of "in response to", "in response to determining", "in the case of", "when", "when", "if", "if" and the like can be replaced with each other.

[0142] In some embodiments, the terms of "greater than", "greater than or equal to", "not less than", "more than", "more than or equal to", "not less than", "higher than", "higher than or equal to", "not lower than", "above" and the like can be replaced with each other, and the terms of "less than", "less than or equal to", "not greater than", "less than", "less than or equal to", "not more than", "lower than", "lower than or equal to", "not higher than", "below" and the like can be replaced with each other.

[0143] In some embodiments, apparatuses and devices can be explained as entities, which can also be explained as virtual, and the names thereof are not limited to the names described in the embodiments, and in some cases can also be understood as "equipment", "device", "circuit", "network element", "node", "function", "unit", "section", "system", "network", "chip", "chip system", "entity", "subject", and the like.

[0144] In some embodiments, "network" can be explained as an apparatus included in the network, for example, an access network device, a core network device, and the like.

[0145] In some embodiments, "terminal" or "terminal device" can be referred to as "user equipment (UE)", "user terminal", "mobile station (MS)", "mobile terminal (MT)", subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, and the like.

[0146] In some embodiments, obtaining data, information, and the like can comply with the laws and regulations of the country where the location is.

[0147] In some embodiments, data, information, and the like can be obtained after obtaining the consent of the user.

[0148] In addition, each element, each row, or each column in the table of the embodiments of the present disclosure can be implemented as an independent embodiment, and any combination of any element, any row, or any column can also be implemented as an independent embodiment.

[0149] FIG. 1a is an exemplary schematic diagram of an architecture of a communication system according to an embodiment of the present disclosure. As shown in FIG. 1a, the communication system 100 includes a transmitter 101 and a receiver 102. It should be understood that the number and form of the transmitter and the number and form of the receiver shown in FIG. 1a are only for example and do not constitute a limitation on the embodiments of the present disclosure, and in actual applications, two or more transmitters and two or more receivers can be included. The communication system 100 shown in FIG. 1a is only exemplified as including one transmitter 101 and one receiver 102.

[0150] The transmitter 101 can be understood as a node or device that converts information into radio waves and transmits the radio waves through an antenna. The specific type of the transmitter 101 is not limited in the embodiments of the present disclosure.

[0151] In some embodiments, the transmitter 101 can include, but is not limited to, at least one of a base station, a terminal, a router, a satellite transmitter, a broadcast station transmitter, and a Bluetooth device.

[0152] In some embodiments, the base station can include an International Mobile Telecommunications Base Station (IMT BS), such as a Macrocell Base Station, a Microcell Base Station, a Small Cell Base Station, and the like.

[0153] In some embodiments, the terminal includes at least one of a mobile phone, a wearable device, an Internet of Things device, a communication-capable car, a smart car, a tablet (Pad), a wireless transceiver-equipped computer, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self-driving, a wireless terminal device in remote medical surgery, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, a wireless terminal device in a smart home, and the like, but is not limited thereto.

[0154] The receiver 102 can include components such as an antenna. The receiver 102 can be understood as a node or device responsible for receiving radio waves transmitted through the antenna and converting the radio waves back into the original information form. The specific type of receiver 102 is not limited in the embodiments of the present disclosure.

[0155] In some embodiments, the receiver 102 can include, but is not limited to, at least one of a terminal, a wireless network card, a satellite receiver, a Bluetooth receiver.

[0156] It can be understood that the communication system described in the embodiments of the present disclosure is for more clear illustration of the technical solutions of the present disclosure, and does not constitute a limitation on the technical solutions proposed in the present disclosure. It can be known by those skilled in the art that, as the system architecture evolves and new business scenarios appear, the technical solutions proposed in the present disclosure are also applicable to similar technical problems.

[0157] The following embodiments of the present disclosure can be applied to the communication system 100 shown in FIG. 1a or part of the subject, but are not limited thereto. The subjects shown in FIG. 1a are exemplary, and the communication system can include all or part of the subjects in FIG. 1a, or other subjects other than those in FIG. 1a. The number and form of each subject is arbitrary, each subject can be physical or virtual, the connection relationship between each subject is exemplary, each subject can not be connected or can be connected, the connection can be in any way, can be direct connection or indirect connection, can be wired connection or wireless connection.

[0158] Embodiments of the present disclosure can be applied to Long Term Evolution (LTE), LTE-Advanced (LTE-A), LTE-Beyond (LTE-B), SUPER 3G, IMT-Advanced, 4th generation mobile communication system (4G), 5th generation mobile communication system (5G), 5G new radio (NR), 6th generation communication system (6G), Future Radio Access (FRA), New-Radio Access Technology (RAT), New Radio (NR), New radio access (NX), Future generation radio access (FX), Global System for Mobile communications (GSM (registered trademark)), CDMA2000, Ultra Mobile Broadband (UMB), IEEE 802.11 (Wi-Fi (registered trademark)), IEEE 802.16 (WiMAX (registered trademark)), IEEE 802.20, Ultra-WideBand (UWB), Bluetooth (Bluetooth (registered trademark)), Public Land Mobile Network (PLMN) network, Device-to-Device (D2D) system, Machine to Machine (M2M) system, Internet of Things (IoT) system, Vehicle-to-Everything (V2X), system using other communication methods, next-generation system expanded based on them, and the like. In addition, a plurality of systems can be combined (for example, combination of LTE or LTE-A and 5G, and the like).

[0159] FIG. 1b is one exemplary schematic diagram of an architecture of a communication system to which embodiments of the present disclosure relate. As shown in FIG. 1b, on the basis of FIG. 1a, an encoder can be provided in the transmitter 101 of the communication system 100, and a decoder can be provided in the receiver 102.

[0160] The transmitter 101 can send the CSI to the receiver 102. Specifically, the transmitter 101 can encode the CSI by an encoder, and can send the encoded CSI to the receiver 102. After receiving the encoded CSI, the receiver 102 can decode the encoded CSI by a decoder to reconstruct the CSI. However, the encoding techniques and standards adopted by different encoders can be different. For example, in one scenario, for a decoder from a non-supplier, although the decoder usually reconstructs the CSI data at the network (NW) end, since the encoders can come from various different suppliers, and each supplier has its unique encoding techniques and standards, it is difficult for the decoder to effectively accommodate and decode the CSI data compressed by the encoders of different suppliers, thereby posing a significant obstacle to the interoperability in a cross-supplier environment. Therefore, how to obtain a decoder for accurately decoding the data encoded by different encoders is a technical problem to be solved.

[0161] To solve the above problems, the disclosure provides a method, device, equipment, storage medium and program product for determining a decoder. The decoder determined by the above method can accurately decode the data encoded by multiple encoders.

[0162] The method, device, equipment, storage medium and program product for determining a decoder provided by the disclosure will be described in detail below with reference to the accompanying drawings.

[0163] Referring to FIG. 2a, FIG. 2a is an exemplary flow diagram of a method for determining a decoder according to an embodiment of the disclosure. As shown in FIG. 2a, the method for determining a decoder includes the following steps:

[0164] In step S2101, at least one decoder is obtained.

[0165] In some embodiments, the at least one decoder is obtained by training on the same data set.

[0166] In some embodiments, the same data set is a collection of multiple data sets, i.e., the same data set is obtained by mixing multiple different individual data sets.

[0167] In some embodiments, the data sets come from different suppliers.

[0168] In some embodiments, the method for determining a decoder provided by the embodiment of the disclosure can be performed by a first device.

[0169] In some embodiments, the first device is at least one of a desktop computer, a tablet computer (Pad), a computer with wireless transceiver function, a server, but is not limited thereto.

[0170] In some embodiments, the at least one decoder can be a decoder trained based on a Convolutional Neural Network (CNN) model.

[0171] In some embodiments, when training the decoder based on the CNN model, the input of the CNN model can be a data set.

[0172] In some embodiments, the data set can include channel features, and the number of channel features can be one or more groups.

[0173] In some embodiments, the channel features can include one or more of the following features:

[0174] The number of devices, wherein the number of devices is the number of devices that send CSI when generating the data set;

[0175] The number of transmission ports, wherein the number of transmission ports is the number of ports of the device that send CSI;

[0176] The number of subbands, wherein the number of subbands is the number of subbands occupied by the device sending CSI;

[0177] The number of In-phase and Quadrature (IQ);

[0178] The number of time slots, wherein the number of time slots is the number of time slots occupied by the device sending CSI;

[0179] The number of random times, wherein the number of random times indicates the number of random times for generating the data set.

[0180] In some embodiments, the channel features can exist in the form of a matrix, and the data set can include one or more matrices. Optionally, the dimensions of the matrix can be the number of devices * the number of transmission ports * the number of subbands * the number of IQ * the number of time slots * the number of random times.

[0181] For example, taking the number of devices as 210, the number of transmission ports as 32, the number of subbands as 13, the number of IQ as 2, the number of time slots as 400, and the number of random times as 5, the size of the matrix is 210*32*13*2*400*5=349.44M.

[0182] In some embodiments, in the above step S2101, when obtaining the at least one decoder, the following three cases can be included but are not limited to:

[0183] Case 1: Obtain the at least one decoder from the local of the first device.

[0184] In this case, the first device can train the at least one decoder through the same data set.

[0185] In this case, the first device can obtain the at least one decoder trained locally.

[0186] In some embodiments, "obtain", "get", "receive", "transmit", "bidirectionally transmit", "send and / or receive" can be replaced by each other, which can be interpreted as receiving from other subjects, obtaining from protocols, obtaining from higher layers, processing by itself, implementing autonomously, and other meanings.

[0187] Case 2: Obtain the at least one decoder from the second device.

[0188] In some embodiments, the second device can be a device for training the at least one decoder, and the name is not limited thereto.

[0189] In some embodiments, the second device is at least one of a desktop computer, a tablet computer (Pad), a computer with wireless transceiver function, a server, but is not limited thereto.

[0190] In this case, the second device can train the at least one decoder by using the same data set.

[0191] In some embodiments, the number of second devices can be one, and the one second device can train the at least one decoder by using the data set.

[0192] In some embodiments, the number of second devices can be multiple, and the multiple second devices can train the at least one decoder by using the data set.

[0193] In this case, the first device can obtain the at least one decoder trained by the second device. That is, the first device can receive the at least one decoder transmitted by the second device.

[0194] In some embodiments, the first device can receive the at least one decoder transmitted by the second device through any known or unknown communication means. For example, the first device can receive the at least one decoder transmitted by the second device through a network.

[0195] Case 3: Obtain N decoders from the second device and obtain M-N decoders locally.

[0196] In some embodiments, the at least one decoder described above includes N decoders obtained from the second device and M-N decoders obtained locally.

[0197] Wherein, M is an integer greater than 1, and N is an integer greater than 1 and less than M.

[0198] In this case, the second device can train N decoders by using the data set, and the first device can train M-N decoders by using the data set. The data set used by the first device and the second device for training is the same data set.

[0199] In this case, the first device can obtain the M-N trained decoders from the local device, and can obtain the N decoders from the second device.

[0200] In step S2102, for any one decoder, the performance of the decoder is detected by using at least one encoder, and at least one second performance is obtained.

[0201] In some embodiments, the method of detecting the performance of each decoder by using at least one encoder to obtain at least one second performance corresponding to each decoder is the same. In step S2102, the method of detecting the performance of the decoder by using at least one encoder to obtain at least one second performance is described by taking any one decoder as an example.

[0202] In some embodiments, the at least one second performance can be the performance detection result of the decoder obtained by detecting the performance of the decoder by using at least one encoder. For any one of the at least one encoder, the performance of the decoder can be detected by using the encoder to obtain one second performance.

[0203] In some embodiments, the second performance of the decoder can be determined by the following method: for any one of the at least one encoder, the fourth data can be encoded by using the encoder to obtain the fifth data; the fifth data can be decoded by using the decoder to obtain the sixth data; and the second performance can be determined according to the similarity between the fourth data and the sixth data. The fourth data can be any data set.

[0204] In some embodiments, the similarity between the fourth data and the sixth data can be the cosine similarity between the fourth data and the sixth data, that is, the value of the second performance is the cosine similarity between the fourth data and the sixth data.

[0205] For example, taking the channel feature of the fourth data as vector A and the channel feature of the sixth data as vector B as an example, the cosine similarity between the fourth data and the sixth data can be wherein, may be the dot product of vector A and vector B, may be the norm of vector A, may be the norm of vector B, and the value range of the cosine similarity is [0, 1], that is, the value range of the second performance of the decoder is [0, 1].

[0206] In some embodiments, the second performance can be determined according to the cosine similarity of the fourth data and the sixth data.

[0207] In some embodiments, taking the number of encoders as n for example, for any one of the at least one decoder, the fourth data can be encoded by each of the n encoders respectively to obtain the fifth data corresponding to each encoder; the fifth data corresponding to each encoder can be decoded by the decoder to obtain the sixth data corresponding to each encoder; and the second performance of the decoder in decoding the encoded data of each encoder can be determined according to the similarity of the fourth data and the sixth data corresponding to each encoder.

[0208] For example, for the decoder a, when the second performance of the decoder a is obtained, the fourth data can be encoded by the encoder 1 to obtain the fifth data, the fifth data can be decoded by the decoder a to obtain the sixth data, and the cosine similarity of the fourth data and the sixth data can be determined, and then the second performance of the decoder a detected by the encoder 1 can be determined according to the cosine similarity of the fourth data and the sixth data. That is, the second performance of the decoder a in decoding the encoded data of the encoder 1 can be obtained.

[0209] Further, in the same way, the second performance of each of the at least one decoder in decoding the encoded data of the n encoders can be obtained. That is, the at least one second performance corresponding to the decoder in the above steps.

[0210] In some embodiments, the fourth data corresponding to each encoder can be the same, or the fourth data corresponding to each encoder can also be different.

[0211] In step 2103, for any one of the decoders, the decoding performance of the decoder on the multiple encoders can be determined according to the at least one second performance.

[0212] In some embodiments, the average of the at least one second performance corresponding to the decoder can be determined as the decoding performance of the decoder on the multiple encoders. That is, the decoding performance of the decoder on the at least one encoder is the average of the at least one second performance value obtained when decoding the encoded data of the at least one encoder.

[0213] In some embodiments, the maximum of the at least one second performance corresponding to the decoder can be determined as the decoding performance of the decoder on the multiple encoders. That is, the decoding performance of the decoder on the at least one encoder is the maximum of the at least one second performance obtained when decoding the encoded data of the at least one encoder.

[0214] In some embodiments, the minimum value in the at least one second performance corresponding to the decoder can be determined as the decoding performance of the decoder to the plurality of encoders, i.e., the minimum value in the at least one second performance obtained when the decoding performance of the decoder to the at least one encoder meets the preset requirement.

[0215] In step S2104, the first decoder is determined from the at least one decoder according to the decoding performance of the decoder to the plurality of encoders.

[0216] The first decoder can be a decoder in the at least one decoder whose decoding performance to the plurality of encoders meets the preset requirement.

[0217] The number of the first decoder can be one or more.

[0218] In some embodiments, the decoding performance to the plurality of encoders mentioned above refers to the second performance mentioned above.

[0219] In some embodiments, the decoder whose decoding performance to the plurality of encoders meets the preset requirement includes, but is not limited to, at least one of the following cases:

[0220] Case 1: the decoding performance of the first decoder to the plurality of encoders is greater than or equal to a second threshold value;

[0221] Specifically, the decoder in the at least one decoder whose decoding performance to the plurality of encoders is greater than or equal to the second threshold value can be determined as the first decoder.

[0222] For example, the at least one decoder includes a decoder a, a decoder b, a decoder c and a decoder d, if the decoding performance of the decoder a to the plurality of encoders is x1, the decoding performance of the decoder a to the plurality of encoders is x2, and the decoding performance of the decoder a to the plurality of encoders is x3, wherein if the values of x1 and x2 are both greater than or equal to the second threshold value, and the value of x3 is less than the second threshold value, the decoding performance of the decoder a and the decoder b meets the preset requirement, and the decoder a and the decoder b are determined as the first decoder. Case 2: according to the order from high to low of the decoding performance, the first decoder is the first K decoders in the at least one decoder.

[0223] In some embodiments, K is an integer greater than or equal to 1.

[0224] Specifically, the at least one decoder can be sorted in descending order of decoding performance, and the first decoder can be determined as the first K decoders in the sorted at least one decoder. For example, if the at least one decoder includes a decoder a, a decoder b, a decoder c, and a decoder d, and the value of K is 2, and the decoding performance of the decoders in descending order is the decoder b, the decoder c, the decoder a, and the decoder d, the decoder b and the decoder c can be determined as the first decoder.

[0225] Case 3: The decoding performance difference of the first decoder to the plurality of encoders is less than a third threshold.

[0226] Specifically, the decoder in the at least one decoder, whose decoding performance difference to the plurality of encoders is less than the third threshold, can be determined as the first decoder. For example, if the decoding performance of the decoder a to the encoder 1 is x1, the decoding performance of the decoder a to the encoder 2 is x2, and the decoding performance of the decoder a to the encoder 3 is x3, and the difference between any two of x1, x2, and x3 is less than the third threshold, it indicates that the decoding performance of the decoder a to the encoder 1, the encoder 2, and the encoder 3 is close, and the decoder a can be determined as the first decoder.

[0227] In some embodiments, the terms “preset”, “set”, “certain”, “preseted”, “indicated”, “a certain”, “any”, “first” and the like can be replaced with each other, and “certain A”, “preset A”, “preset A”, “set A”, “indicated A”, “a certain A”, “any A”, “first A” can be interpreted as A predefined in a protocol or the like, or A obtained by setting, configuration, or indication, or certain A, a certain A, any A, or first A, but not limited thereto.

[0228] In some embodiments, the second threshold and the third threshold can be set according to actual needs, and the embodiments of the present disclosure are not limited thereto. In some embodiments, “preset threshold”, “threshold”, “preset value”, “certain value”, “a certain value”, “set value”, “indicated value”, “preset value”, “set value” and the like can be replaced with each other. “Preset threshold A”, “threshold A”, “preset value A”, “certain value A”, “a certain value A”, “set value A”, “indicated value A”, “preset value A”, “set value A” can be interpreted as A predefined in a protocol or the like, or A obtained by setting, configuration, or indication, or certain A, a certain A, any A, or first A, but not limited thereto.

[0229] In the embodiments of the present disclosure, the first device can obtain at least one decoder, the at least one decoder being obtained by training on the same data set; for any one of the decoders, the performance of the decoder can be detected by at least one encoder respectively, and at least one second performance can be obtained; for any one of the decoders, the decoding performance of the decoder can be determined according to the at least one second performance; and the first decoder can be determined from the at least one decoder according to the decoding performance of the at least one decoder. The first decoder determined by the above method has decoding performance on the plurality of encoders that meets a preset requirement, that is, the first decoder can be used to accurately decode data encoded by the plurality of encoders.

[0230] Referring to FIG. 2b, FIG. 2b is an exemplary flow diagram of a method for determining a decoder according to an embodiment of the present disclosure. As shown in FIG. 2b, the method for determining a decoder includes the following steps:

[0231] In step S2201, at least one decoder is obtained.

[0232] In some embodiments, the at least one decoder is obtained by training on at least one data set.

[0233] In some embodiments, the similarity between the at least one data set is greater than or equal to a first threshold.

[0234] In some embodiments, the first threshold can be set according to actual needs, and the embodiments of the present disclosure do not limit this.

[0235] In some embodiments, the at least one decoder corresponds to the at least one data set one by one. That is, a decoder can be obtained by training on each data set. For example, decoder 1 can be obtained by training on data set A, decoder 2 can be obtained by training on data set B, and so on.

[0236] In some embodiments, obtaining the at least one decoder includes at least the following three cases:

[0237] Case 1: obtaining the at least one decoder from the local device.

[0238] In this case, the first device can obtain the at least one decoder by training on at least one data set with a similarity greater than or equal to a first threshold. The first device can obtain the at least one decoder trained from the local device.

[0239] Case 2: obtaining the at least one decoder from a second device.

[0240] In this case, the second device can obtain the at least one decoder by training on at least one data set with a similarity greater than or equal to a first threshold. The first device can obtain the at least one decoder trained from the second device.

[0241] Case 3: obtaining N decoders from the second device, obtaining M-N decoders locally, and the at least one decoder comprises the N decoders obtained from the second device and the M-N decoders obtained locally.

[0242] In this case, the second device can train the N decoders by using the N data sets, and the first device can train the M-N decoders by using the M-N data sets. The similarity between the N data sets and the M-N data sets is greater than or equal to the first threshold.

[0243] In this case, the first device can obtain the M-N decoders trained locally, and can obtain the N decoders from the second device.

[0244] Step S2202: For any one decoder, the performance of the decoder is detected by using the at least one encoder respectively, and at least one second performance is obtained.

[0245] Step S2203: For any one decoder, the decoding performance of the decoder is determined according to the at least one second performance.

[0246] Step S2204: The first decoder is determined from the at least one decoder according to the decoding performance of the at least one decoder.

[0247] The optional implementation manner of steps S2201-S2204 can refer to the optional implementation manner of steps S2101-S2104 of FIG. 2a and other associated parts in the embodiments involved in FIG. 2a, which will not be described herein.

[0248] In the embodiments of the present disclosure, the first device can obtain at least one decoder, the at least one decoder is obtained by training at least one data set with a similarity greater than or equal to a first threshold; for any one decoder, the performance of the decoder can be detected by using the at least one encoder respectively, and at least one second performance is obtained; for any one decoder, the decoding performance of the decoder can be determined according to the at least one second performance; and the first decoder can be determined from the at least one decoder according to the decoding performance of the at least one decoder. The decoding performance of the first decoder determined by the above method to the multiple encoders meets the preset requirement, that is, the first decoder can be used to accurately decode the data encoded by the multiple encoders.

[0249] Referring to FIG. 2c, FIG. 2c is an exemplary flow diagram of a method for determining a decoder according to an embodiment of the present disclosure. As shown in FIG. 2c, the method for determining a decoder comprises the following steps:

[0250] Step S2301: obtaining at least one decoder and at least one data set.

[0251] In some embodiments, the at least one decoder is trained by the at least one data set.

[0252] In some embodiments, the at least one decoder corresponds to the at least one data set one by one. That is, a decoder can be trained by each data set. For example, a decoder a can be trained by a data set A, a decoder b can be trained by a data set B, and so on.

[0253] In some embodiments, obtaining the at least one decoder and the at least one data set at least includes the following three cases:

[0254] Case 1: obtaining the at least one decoder and the at least one data set from the local.

[0255] In this case, the first device can train the at least one decoder by the at least one data set. The first device can obtain the trained at least one decoder and the at least one data set used for training the at least one decoder from the local.

[0256] Case 2: obtaining the at least one decoder and the at least one data set from the second device.

[0257] In this case, the second device can train the at least one decoder by the at least one data set. The first device can obtain the trained at least one decoder and the at least one data set used for training the at least one decoder from the second device.

[0258] Case 3: obtaining N decoders and N data sets from the second device, and obtaining M-N decoders and M-N data sets from the local.

[0259] In this case, the second device can train N decoders by N data sets, and the first device can train M-N decoders by M-N data sets.

[0260] In this case, the first device can obtain the trained M-N decoders and the M-N data sets used for training the M-N decoders from the local, and can obtain the N decoders and the N data sets used for training the N decoders from the second device.

[0261] Step S2302: determining whether the similarity between the at least one data set is greater than or equal to a first threshold.

[0262] If the similarity between the at least one data set is greater than or equal to the first threshold, step S2303 is performed.

[0263] If the similarity between the at least one data set is less than the first threshold, step S2304 is performed.

[0264] In some embodiments, the similarity between the at least one data set can be a cosine similarity between the at least one data set.

[0265] In some embodiments, the first threshold value can be a similarity baseline determined according to two identical data sets.

[0266] In some embodiments, the first threshold value can be determined by the following steps S1-S3:

[0267] Step S1, determining two identical data sets.

[0268] In some embodiments, assuming that the two identical data sets are data set C and data set D respectively, the data set C can include P vectors, and the data set D can include Q vectors. Wherein, P and Q can be integers greater than or equal to 1.

[0269] Step S2, determining the average cosine similarity of the two data sets in step S1.

[0270] In some embodiments, the average cosine similarity of the two data sets in step S1 can be determined by the following formula:

[0271] Wherein, Can be the dot product of vector C and vector D, Can be the norm of vector C, Can be the norm of vector D, and the value range of the average cosine similarity is [0, 1].

[0272] Step S3, determining the average cosine similarity in step S2 as the first threshold value.

[0273] In some embodiments, there are at least three ways to determine whether the similarity between the at least one data set is greater than or equal to the first threshold value:

[0274] Method one: according to the cosine similarity between the at least one data set, determine whether the similarity between the at least one data set is greater than or equal to the first threshold value, if the similarity is greater than or equal to the first threshold value, it means that the at least one data set is similar, if the similarity is less than the first threshold value, it means that the at least one data set is not similar.

[0275] In this way, the cosine similarity between any two data sets in the at least one data set can be determined; if the cosine similarity is greater than or equal to the first threshold value, it is determined that the two data sets are similar; if the cosine similarity between the two data sets is less than the first threshold value, it is determined that the two data sets are not similar.

[0276] The second mode is to determine whether the at least one data set is similar through a preset decoder.

[0277] In this mode, the decoding performance of the preset decoder on the encoded data of the at least one data set can be obtained; if the difference between the decoding performances corresponding to the at least one data set is less than a fourth threshold, it is determined that the at least one data set is similar; if the difference between the decoding performances corresponding to the at least one data set is greater than the fourth threshold, it is determined that the at least one data set is not similar.

[0278] In some embodiments, the preset decoder can be any decoder. For example, the preset decoder can be a decoder that has been pre-trained.

[0279] In some embodiments, the fourth threshold can be set according to actual needs, and the embodiments of the present disclosure do not limit this.

[0280] In some embodiments, obtaining the decoding performance of the preset decoder on the encoded data of the at least one data set can include the following steps S4-S7:

[0281] Step S4, encoding the at least one data set through a preset encoder to obtain the encoded data corresponding to each data set.

[0282] In some embodiments, the preset encoder can be an encoder corresponding to the preset decoder. Alternatively, the preset encoder can be an encoder used when the preset decoder is trained.

[0283] For any one data set in the at least one data set, the data set can be encoded through the preset encoder to obtain the encoded data corresponding to the data set.

[0284] Step S5, decoding the encoded data corresponding to each data set through the preset decoder to obtain the decoded data corresponding to each data set.

[0285] For the encoded data corresponding to any one data set, the encoded data can be decoded through the preset decoder to obtain the decoded data corresponding to the data set.

[0286] Step S6, determining the similarity between the decoded data and the encoded data corresponding to each data set.

[0287] In some embodiments, for any one data set, the cosine similarity between the decoded data and the encoded data corresponding to the data set can be determined as the similarity between the decoded data and the encoded data corresponding to the data set.

[0288] Step S7, for any one data set in the at least one data set, the similarity corresponding to the data set is determined as the decoding performance of the preset decoder on the encoded data of the data set.

[0289] In some embodiments, the decoding performance of the preset decoder on the encoded data of the data set, which can also be referred to as "the decoding performance corresponding to the data set", can be determined.

[0290] In some embodiments, the difference between the decoding performance corresponding to the at least one data set can be less than a fourth threshold value, which can include that, for any two data sets in the at least one data set, the difference between the decoding performance corresponding to the two data sets is less than the fourth threshold value.

[0291] Option 3: The at least one data set can be input into the preset model, and the indication information can be obtained from the preset model. The indication information is used to indicate whether the similarity between the at least one data set is greater than or equal to a preset similarity.

[0292] In step S2303, for any one decoder, the performance of the decoder is detected by using the at least one encoder, and at least one second performance is obtained.

[0293] In step S2304, for any one decoder, the decoding performance of the decoder is determined according to the at least one second performance.

[0294] In step S2305, according to the decoding performance of the at least one decoder on the plurality of encoders, a first decoder is determined from the at least one decoder.

[0295] The optional implementation of steps S2303-S2305 can refer to the optional implementation of steps S2102-S2104 of FIG. 2a and other related parts in the embodiments involved in FIG. 2a, which will not be described here.

[0296] In step S2306, for any one decoder, each first data is encoded by using the encoder corresponding to the decoder, and second data corresponding to each first data is obtained.

[0297] It should be noted that, in the process of training the decoder, the decoder is trained by using the data set and the encoder. For any one decoder, the encoder used for training the decoder and the data set are set.

[0298] In some embodiments, the encoder corresponding to the decoder can be the encoder used for training the decoder.

[0299] In some embodiments, the number of first data can be a plurality.

[0300] In some embodiments, each first data can exist in the form of a matrix. For example, each first data can be CSI represented in the form of a matrix.

[0301] In some embodiments, the second data can be data obtained by encoding the first data by using the encoder corresponding to the decoder.

[0302] In some embodiments, the plurality of first data can be encoded by the encoder corresponding to the decoder to obtain second data corresponding to each first data.

[0303] It should be understood that the second data corresponds to the first data one by one, and the number of the second data is the same as the number of the first data.

[0304] In step S2307, for any one decoder, the second data corresponding to each first data is decoded by the decoder to obtain third data corresponding to each first data.

[0305] In some embodiments, after the plurality of first data is encoded by the encoder corresponding to the decoder to obtain second data corresponding to each first data, the second data corresponding to each first data can be decoded by the decoder to obtain third data corresponding to each first data. Specifically, for any one first data, the first data can be encoded by the encoder corresponding to the decoder to obtain second data corresponding to the first data, and then the second data can be decoded by the decoder to obtain third data corresponding to the first data.

[0306] It should be understood that the third data corresponds to the first data one by one, and the number of the third data is the same as the number of the first data.

[0307] In step S2308, for any one decoder, the first performance of the decoder is determined according to each first data and third data corresponding to the first data.

[0308] In some embodiments, the data similarity between each first data and corresponding third data can be determined to obtain a plurality of first similarities, and the first performance can be determined according to the plurality of first similarities.

[0309] In some embodiments, the data similarity between each first data and corresponding third data can be the cosine similarity between the first data and the corresponding third data. For example, for any one first data, the cosine similarity between the first data and the corresponding third data can be the first similarity corresponding to the first data.

[0310] In some embodiments, the cosine similarity between each first data and corresponding third data can be determined to obtain the first similarity corresponding to each first data. In this way, a plurality of first similarities can be obtained.

[0311] In some embodiments, the average of the plurality of first similarities can be determined as the first performance, or the maximum of the plurality of first similarities can be determined as the first performance, or the minimum of the plurality of first similarities can be determined as the first performance.

[0312] Step S2309: determining the performance similarity of the at least one decoder according to the first performance of the at least one decoder.

[0313] It should be noted that the first performance of each of the at least one decoder can be determined through steps S2306-S2309.

[0314] In some embodiments, the performance similarity of the at least one decoder can be the similarity between the first performance of the at least one decoder. Specifically, the similarity between the first performance of the at least one decoder can be determined, and the similarity between the first performance of the at least one decoder can be determined as the performance similarity of the at least one decoder.

[0315] In some embodiments, the similarity between the first performance of the at least one decoder can be the cosine similarity between the first performance of the at least one decoder.

[0316] Step S2310: determining whether the performance similarity of the at least one decoder is greater than or equal to a preset performance threshold.

[0317] If the performance similarity of the at least one decoder is greater than or equal to the preset performance threshold, steps S2303-S2305 are performed.

[0318] If the performance similarity of the at least one decoder is less than the preset performance threshold, step S2311 is performed.

[0319] In some embodiments, the preset performance threshold can be set according to actual needs, which is not limited in the embodiments of the present disclosure.

[0320] In some embodiments, if the performance similarity of the at least one decoder is greater than or equal to the preset performance threshold, the decoding performance of the at least one decoder can be determined, and the first decoder can be determined from the at least one decoder according to the decoding performance of the at least one decoder on the plurality of encoders. That is, steps S2303-S2305 can be performed to determine the first decoder from the at least one decoder.

[0321] The optional implementation of steps S2303-S2305 can refer to the optional implementation of steps S2102-S2104 of FIG. 2a and other associated parts in the embodiments involved in FIG. 2a, which will not be repeated here.

[0322] In some embodiments, if the performance similarity of the at least one decoder is less than the preset performance threshold, step S2311 can be performed.

[0323] Step S2311, at least one decoder is trained by the total data set.

[0324] In some embodiments, the total data set can be a combination of the at least one data set obtained in S2301. For example, assuming that the at least one data set obtained in S2301 is data set 1, data set 2 and data set 3 respectively, the total data set can include data set 1, data set 2 and data set 3.

[0325] In some embodiments, the way to train at least one decoder by the total data set is different according to the way to obtain at least one data set in S2301. The way to train at least one decoder by the total data set includes at least the following three cases:

[0326] Case 1, the first device trains at least one decoder by the total data set.

[0327] In this case, the first device can train at least one decoder by the total data set and at least one encoder. The at least one encoder can be at least one encoder used by the first device to train the decoder in S2301. That is, the first device can train at least one decoder by the total data set and at least one encoder used to train the decoder in S2301.

[0328] Case 2, the second device trains at least one decoder by the total data set.

[0329] In this case, the first device can instruct the second device to train at least one decoder by the total data set.

[0330] In this case, the second device can train at least one decoder by the total data set and at least one encoder. The at least one encoder can be at least one encoder used by the second device to train the decoder in S2301. That is, the second device can train at least one decoder by the total data set and at least one encoder used to train the decoder in S2301.

[0331] Case 3, the first device and the second device train at least one decoder by the total data set.

[0332] In this case, the first device can train M-N decoders by the total data set. The second device can train N decoders by the total data set. The present disclosure realizes that at least one decoder can include M-N decoders trained by the first device and N decoders trained by the second device.

[0333] In some embodiments, after the at least one decoder is trained by using the total data set, the decoding performance of the at least one decoder can be determined; and the first decoder can be determined from the at least one decoder according to the decoding performance of the at least one decoder to the plurality of encoders. That is, after the at least one decoder is trained by using the total data set, steps S2303-S2305 can be performed to determine the first decoder from the at least one decoder.

[0334] The optional implementation of steps S2303-S2305 can refer to the optional implementation of steps S2102-S2104 of FIG. 2a and other associated parts in the embodiments involved in FIG. 2a, which will not be described herein again.

[0335] In the embodiments of the present disclosure, the first device can obtain the at least one decoder and the at least one data set, determine whether the similarity between the at least one data set is greater than or equal to a first threshold value, perform performance detection on the decoder by using the at least one encoder for any one decoder to obtain at least one second performance, determine the decoding performance of the decoder according to the at least one second performance, and determine the first decoder from the at least one decoder according to the decoding performance of the at least one decoder to the plurality of encoders, if the similarity between the at least one data set is greater than or equal to the first threshold value; if the similarity between the at least one data set is less than the first threshold value, encode each first data by using the corresponding encoder of the decoder for any one decoder, obtain the second data corresponding to each first data, decode the second data corresponding to each first data by using the decoder to obtain the third data corresponding to each first data, determine the first performance of the decoder according to each first data and the third data corresponding to the first data, determine the performance similarity of the at least one decoder according to the first performance of the at least one decoder, determine whether the performance similarity of the at least one decoder is greater than or equal to a preset performance threshold value, determine the decoding performance of the at least one decoder if the performance similarity of the at least one decoder is greater than or equal to the preset performance threshold value, and determine the first decoder from the at least one decoder according to the decoding performance of the at least one decoder to the plurality of encoders; if the performance similarity of the at least one decoder is less than the preset performance threshold value, the at least one data set can be combined into a total data set, the at least one decoder can be trained by using the total data set, and the first decoder can be determined from the at least one decoder. The decoding performance of the first decoder to the plurality of encoders determined by the above method meets the preset requirement, that is, the first decoder can be used to accurately decode the data encoded by the plurality of encoders.

[0336] Referring to FIG. 2d, FIG. 2d is an exemplary flow diagram of a method for determining a decoder according to an embodiment of the present disclosure. As shown in FIG. 2d, the method for determining a decoder includes the following steps:

[0337] In step S2401, at least one decoder is obtained.

[0338] In some embodiments, the at least one decoder is obtained by training on the same data set.

[0339] In some embodiments, the same data set is obtained by mixing a plurality of different individual data sets.

[0340] In some embodiments, the data sets are from different suppliers.

[0341] Optional implementation of step S2401 can refer to the optional implementation of step S2101 of FIG. 2a and other associated parts of the embodiments involved in FIG. 2a, which will not be repeated here.

[0342] In step S2402, the decoding performance of the at least one decoder on a plurality of encoders is determined.

[0343] In some embodiments, the decoding performance of each decoder can be determined by at least one encoder.

[0344] In some embodiments, the at least one encoder can be the encoder used in training the at least one decoder.

[0345] In step S2403, a first decoder is determined from the at least one decoder according to the decoding performance of the at least one decoder on a plurality of encoders.

[0346] In some embodiments, optional implementation of steps S2402-S22403 can refer to the optional implementation of steps S2102-S2104 of FIG. 2a and other associated parts of the embodiments involved in FIG. 2a, which will not be repeated here.

[0347] In an embodiment of the present disclosure, the first device can obtain at least one decoder obtained by training on the same data set, can determine the decoding performance of the at least one decoder, and can determine a first decoder from the at least one decoder according to the decoding performance of the at least one decoder on a plurality of encoders. The decoding performance of the first decoder determined by the above method on a plurality of encoders meets the preset requirement, that is, the first decoder can be used to accurately decode data encoded by a plurality of encoders.

[0348] Referring to FIG. 2e, FIG. 2e is a schematic diagram of a method for determining a decoder according to an embodiment of the present disclosure. As shown in FIG. 2e, the method for determining a decoder includes the following steps:

[0349] At S2501, at least one decoder is obtained, the at least one decoder being obtained by training on at least one data set.

[0350] The similarity of the at least one data set is greater than or equal to a first threshold.

[0351] Optional implementation of S2501 can refer to optional implementation of S2105 in FIG. 2b and other associated parts in the embodiments involved in FIG. 2b, which will not be repeated here.

[0352] At S2502, decoding performance of the at least one decoder is determined.

[0353] In some embodiments, the decoding performance of each decoder can be determined by the at least one encoder.

[0354] In some embodiments, the at least one encoder can be the encoder used in the training of the at least one.

[0355] Optional implementation of S2502 can refer to optional implementation of S2202-S2203 in FIG. 2b and other associated parts in the embodiments involved in FIG. 2b, which will not be repeated here.

[0356] At S2503, a first decoder is determined from the at least one decoder according to decoding performance of the at least one decoder on the plurality of encoders.

[0357] Optional implementation of S2503 can refer to optional implementation of S2204 in FIG. 2b and other associated parts in the embodiments involved in FIG. 2b, which will not be repeated here.

[0358] In the embodiments of the present disclosure, the first device can obtain at least one decoder obtained by training on at least one data set, can determine decoding performance of the at least one decoder, and can determine a first decoder from the at least one decoder according to decoding performance of the at least one decoder on the plurality of encoders. The decoding performance of the first decoder determined by the above method on the plurality of encoders meets the preset requirement, that is, the first decoder can be used to accurately decode data encoded by the plurality of encoders.

[0359] Referring to FIG. 2f, FIG. 2f is an exemplary flow diagram of a method for determining a decoder according to an embodiment of the present disclosure. As shown in FIG. 2f, the method for determining a decoder includes the following steps:

[0360] At S2601, at least one decoder and at least one data set are obtained, the at least one decoder being obtained by training on the at least one data set.

[0361] Step S2602, determining whether the similarity between the at least one data set is greater than or equal to a first threshold value.

[0362] If the similarity between the at least one data set is greater than or equal to the first threshold value, step S2603 is performed.

[0363] If the similarity between the at least one data set is less than the first threshold value, step S2605 is performed.

[0364] The optional implementation of steps S2601-S2602 can refer to the optional implementation of steps S2301-S2302 of FIG. 2c and other associated parts of the embodiments involved in FIG. 2c, which will not be repeated here.

[0365] Step S2603, determining the decoding performance of the at least one decoder.

[0366] In some embodiments, the decoding performance of the at least one decoder can be determined by the at least one encoder.

[0367] In some embodiments, the at least one encoder can be the at least one used encoder.

[0368] The optional implementation of step S2603 can refer to the optional implementation of steps S2304-S2304 of FIG. 2c and other associated parts of the embodiments involved in FIG. 2c, which will not be repeated here.

[0369] Step S2604, determining a first decoder in the at least one decoder according to the decoding performance of the plurality of encoders by the at least one decoder.

[0370] The optional implementation of step S2604 can refer to the optional implementation of step S2305 of FIG. 2c and other associated parts of the embodiments involved in FIG. 2c, which will not be repeated here.

[0371] Step S2605, determining the performance similarity of the at least one decoder.

[0372] In some embodiments, when determining the performance similarity of the at least one decoder, the first performance of each decoder can be determined by the plurality of first data, and the performance similarity of the at least one decoder can be determined according to the first performance of each decoder.

[0373] The implementation of determining the performance similarity of the at least one decoder can refer to steps S2306-S2309 of FIG. 2c, which will not be repeated here.

[0374] Step S2606, determining whether the performance similarity of the at least one decoder is greater than or equal to a preset performance threshold value.

[0375] If the performance similarity of the at least one decoder is greater than or equal to the preset performance threshold, step S2603-step S2604 are performed.

[0376] If the performance similarity of the at least one decoder is less than the preset performance threshold, step S2607 is performed.

[0377] In step S2607, the at least one data set is combined into a total data set, and the at least one decoder is trained by using the total data set.

[0378] The optional implementation of step S2606-step S2607 can refer to the optional implementation of step S2310-step S2311 in FIG. 2c and other associated parts in the embodiments involved in FIG. 2c, which will not be described here.

[0379] It should be noted that after step S2607 is performed, step S2603-step S2604 can be continuously performed to determine the first decoder in the at least one decoder.

[0380] In the embodiments of the present disclosure, the first device can obtain the at least one decoder and the at least one data set, determine whether the similarity between the at least one data set is greater than or equal to a first threshold, determine the decoding performance of the at least one decoder if the similarity between the at least one data set is greater than or equal to the first threshold, determine the first decoder in the at least one decoder according to the decoding performance of the at least one decoder to the multiple encoders, determine the performance similarity of the at least one decoder if the similarity between the at least one data set is less than the first threshold, determine the decoding performance of the at least one decoder if the performance similarity of the at least one decoder is greater than or equal to a preset performance threshold, determine the first decoder in the at least one decoder according to the decoding performance of the at least one decoder to the multiple encoders, combine the at least one data set into a total data set and train the at least one decoder by using the total data set if the performance similarity of the at least one decoder is less than the preset performance threshold, and determine the first decoder in the at least one decoder trained. The decoding performance of the first decoder determined by the above method meets the preset requirement, that is, the first decoder can be used to accurately decode the data encoded by the multiple encoders.

[0381] Referring to FIG. 2g, FIG. 2g is an exemplary flow diagram of a method for determining a decoder according to an embodiment of the present disclosure. As shown in FIG. 2g, the method for determining a decoder includes the following steps:

[0382] In step S2701, at least one trained decoder is obtained.

[0383] The optional implementation of step S2701 can refer to the optional implementation of step S2101 in FIG. 2a, and other associated parts in the embodiments involved, which will not be repeated here. Alternatively, the optional implementation of step S2701 can refer to the optional implementation of step S2201 in FIG. 2b, and other associated parts in the embodiments involved, which will not be repeated here. Alternatively, the optional implementation of step S2701 can refer to the optional implementation of step S2301 in FIG. 2c, and other associated parts in the embodiments involved, which will not be repeated here.

[0384] In step S2702, a first decoder is determined from the at least one decoder according to the decoding performance of the plurality of encoders on the at least one decoder.

[0385] The decoding performance of the first decoder on the plurality of encoders meets a preset requirement.

[0386] The optional implementation of step S2702 can refer to the optional implementation of steps S2102-S2104 in FIG. 2a, and other associated parts in the embodiments involved, which will not be repeated here. Alternatively, the optional implementation of step S2702 can refer to the optional implementation of steps S2202-S2204 in FIG. 2b, and other associated parts in the embodiments involved, which will not be repeated here. Alternatively, the optional implementation of step S2702 can refer to the optional implementation of steps S2302-S2311 in FIG. 2c, and other associated parts in the embodiments involved, which will not be repeated here.

[0387] In the embodiments of the present disclosure, part or all of the steps, and their optional implementations, can be combined with part or all of the steps in other embodiments, or combined with optional implementations of other embodiments.

[0388] FIG. 2h is a schematic diagram of a CNN-based decoder according to an embodiment of the present disclosure. As shown in FIG. 2h, the decoder includes a neuron A, a fully connected layer (FC), a first convolutional layer, a first batch normalization module, a resource network block, and a second convolutional layer.

[0389] In some embodiments, the neuron A is used for scalar dequantization of input data, converting quantized integer values back to floating-point numbers for further calculation or analysis. The bit width of the neuron A is not limited. Optionally, the bit width of the neuron A can be 32.

[0390] The size of the input data is not limited in the embodiments of the present disclosure. Optionally, taking 64-bit input data as an example, scalar dequantization processing by the neuron A with a bit width of 32 can obtain a 32-bit floating-point value.

[0391] In some embodiments, the parameters of the full connection layer are not limited by the embodiments of the present disclosure. Optionally, taking the full connection layer with 2 13 32 as an example, it is indicated that there are 2 independent full connection layers, or it is indicated that the input data has 2 channels, and the length of the input feature vector of the full connection layer is 13, and 13 features of each channel are mapped to 32 output neurons.

[0392] In some embodiments, the parameters of the first convolution layer are not limited by the embodiments of the present disclosure. Optionally, the size of the convolution kernel of the first convolution layer can be 3*3, and the stride of the convolution layer can be 1.

[0393] In some embodiments, the first batch normalization module is used to standardize the output of the convolution layer and introduce a nonlinear activation to improve the training stability and performance of the model. Optionally, the batch normalization module is used to perform a 2D batch normalization operation on the two-dimensional feature map of the output of the first convolution layer to perform batch normalization, and an activation function ReLU (Rectified Linear Unit) is used on the batch normalized data, thereby alleviating the gradient vanishing problem and accelerating the training process.

[0394] In some embodiments, the resource network block can include a third convolution layer, a second batch normalization module, a fourth convolution layer, a third batch normalization module, and an activation function ReLU.

[0395] In some embodiments, the parameters of the second convolution layer are not limited by the embodiments of the present disclosure. Optionally, the size of the convolution kernel of the second convolution layer can be 1*1, and the stride of the convolution layer can be 1.

[0396] In some embodiments, through the above-mentioned modules, the output of the CNN-based decoder is large decoding data with a size of 2*13*32.

[0397] The embodiments of the present disclosure also propose a device for implementing any of the above methods, for example, a device is proposed, and the above device includes units or modules for implementing each step in any of the above methods.

[0398] It should be understood that the division of each unit or module in the above apparatus is only a logical function division, and all or part of them can be integrated into a physical entity or physically separated in actual implementation. In addition, the units or modules in the apparatus can be implemented in the form of processor calling software: for example, the apparatus includes a processor, the processor is connected with a memory, the memory stores instructions, and the processor calls the instructions stored in the memory to realize the functions of any of the above methods or the units or modules of the above apparatus, wherein the processor is a general processor such as a central processing unit (CPU) or a microprocessor, and the memory is a memory in the apparatus or a memory outside the apparatus. Alternatively, the units or modules in the apparatus can be implemented in the form of hardware circuit, and the functions of part or all of the units or modules can be realized by the design of the hardware circuit. The above hardware circuit can be understood as one or more processors; for example, in one implementation, the above hardware circuit is an application-specific integrated circuit (ASIC), and the functions of part or all of the units or modules are realized by the design of the logical relationship between the elements in the circuit; for another example, in another implementation, the above hardware circuit is a programmable logic device (PLD), and a field programmable gate array (FPGA) is taken as an example, which can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by a configuration file, so as to realize the functions of part or all of the units or modules. All units or modules of the above apparatus can be all implemented in the form of processor calling software, or all implemented in the form of hardware circuit, or part implemented in the form of processor calling software and the remaining part implemented in the form of hardware circuit.

[0399] In the embodiments of the present disclosure, the processor is a circuit with signal processing capability. In one implementation, the processor can be a circuit with instruction reading and running capability, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), a digital signal processor (DSP), or the like. In another implementation, the processor can implement certain functions through a logical relationship of a hardware circuit, and the logical relationship of the hardware circuit is fixed or can be reconfigured. For example, the processor is a hardware circuit implemented by an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In the reconfigurable hardware circuit, the processor loads a configuration document to implement the configuration of the hardware circuit. It can be understood that the processor loads instructions to implement the functions of the above part or all units or modules. In addition, the hardware circuit can also be designed for artificial intelligence, which can be understood as an ASIC, such as a neural network processing unit (NPU), a tensor processing unit (TPU), a deep learning processing unit (DPU), or the like.

[0400] FIG. 3 is an example structure diagram of the determination apparatus of the decoder provided by the embodiments of the present disclosure. As shown in FIG. 3, the determination apparatus 3100 of the decoder can include at least one of an acquisition module 3101, a determination module 3102, and the like.

[0401] In some embodiments, the acquisition module 3101 is configured to acquire at least one decoder trained.

[0402] In some embodiments, the determination module 3102 is configured to determine a first decoder from the at least one decoder according to decoding performance of the plurality of encoders on the at least one decoder, and the first decoder has decoding performance of the plurality of encoders satisfying a preset requirement.

[0403] In some embodiments, the at least one decoder is obtained by training a data set, and the data set includes channel features.

[0404] In some embodiments, the channel features include at least one of the following:

[0405] A number of devices, the number of devices being a number of devices that generate the data set, transmit channel state information (CSI);

[0406] A number of transmission ports, the number of transmission ports being a number of ports through which the device transmits the CSI;

[0407] A number of subbands, the number of subbands being a number of subbands occupied by the device when transmitting the CSI;

[0408] A number of in-phase and quadrature (IQ) pairs;

[0409] A number of slots, the number of slots being a number of slots occupied by the device when transmitting the CSI; or

[0410] A number of random times, the number of random times being a number of times of randomly generating the data set.

[0411] In some embodiments, the at least one decoder is trained on a same data set, the same data set being a set of the plurality of data sets.

[0412] In some embodiments, the at least one decoder is trained on different data sets, a similarity between the different data sets being greater than or equal to a first threshold.

[0413] In some embodiments, the first decoder satisfies a preset requirement on decoding performance of the plurality of encoders, including at least one of the following:

[0414] The decoding performance of the first decoder is greater than or equal to a second threshold;

[0415] In an order from high to low of the decoding performance, the first decoder is a top K decoder in the at least one decoder, K being an integer greater than or equal to 1; or

[0416] The difference in decoding performance of the first decoder on the plurality of encoders is less than a third threshold.

[0417] In some embodiments, the at least one decoder is trained on at least one data set.

[0418] In some embodiments, the determining module 3102 is specifically configured to: when a similarity between the at least one data set is greater than or equal to the first threshold, determine the first decoder according to decoding performance of the at least one decoder on the plurality of encoders;

[0419] When the similarity between the at least one data set is less than the first threshold, determine a performance similarity of the at least one decoder, and determine the first decoder according to the performance similarity.

[0420] In some embodiments, the similarity between the at least one data set is a cosine similarity between the at least one data set.

[0421] In some embodiments, the obtaining module 3101 is further configured to:

[0422] obtain decoding performance of the preset decoder on the encoded data of the at least one data set;

[0423] The determining module 3102 is further configured to: if the difference between the decoding performances corresponding to the at least one data set is less than a fourth threshold, determine that the similarity between the at least one data set is greater than or equal to the first threshold.

[0424] In some embodiments, the determining module 3102 is specifically configured to: determine, as the first decoder, a decoder in the at least one decoder that satisfies at least one of the following conditions:

[0425] the decoding performance is greater than or equal to a second threshold;

[0426] in descending order of the decoding performance, the first decoder is the top K decoders in the at least one decoder, K being an integer greater than or equal to 1; or

[0427] the difference between the decoding performances of the plurality of encoders is less than a third threshold.

[0428] In some embodiments, the obtaining module 3101 is specifically configured to:

[0429] determine the first performance of each decoder through the plurality of first data;

[0430] determine the performance similarity of the at least one decoder according to the first performance of each decoder.

[0431] In some embodiments, for any one decoder; the determining module 3102 is specifically configured to:

[0432] encode each first data through the corresponding encoder of the decoder to obtain second data corresponding to each first data;

[0433] decode the second data corresponding to each first data through the decoder to obtain third data corresponding to each first data;

[0434] determine the first performance of the decoder according to each first data and the third data corresponding to the first data.

[0435] In some embodiments, the determining module 3102 is specifically configured to:

[0436] determine the data similarity between each first data and the corresponding third data to obtain a plurality of first similarities;

[0437] determine the first performance according to the plurality of first similarities.

[0438] In some embodiments, the determining module 3102 is specifically configured to:

[0439] When the performance similarity is greater than or equal to the preset performance threshold, the decoder satisfying at least one of the following conditions in the at least one decoder is determined as the first decoder: the decoding performance is greater than or equal to a second threshold; in descending order of decoding performance, the first decoder is the first K decoders in the at least one decoder, K is an integer greater than or equal to 1; or, the decoding performance difference of the plurality of encoders is less than a third threshold;

[0440] Or, when the performance similarity is less than the preset performance threshold, the decoder training is performed through the at least one data set to obtain the first decoder.

[0441] In some embodiments, the obtaining module 3101 is specifically configured to:

[0442] Receive at least one decoder sent by at least one second device.

[0443] In some embodiments, the method is performed by the first device; the number of at least one decoder is M, M is an integer greater than 1; the obtaining module 3101 is specifically configured to:

[0444] Receive N decoders sent by at least one second device;

[0445] Wherein, N is an integer greater than 1 and less than M; M-N decoders in the M decoders except the N decoders are obtained by the first device.

[0446] In some embodiments, the method of the present embodiment is performed by the first device; the at least one decoder is obtained by the first device.

[0447] In some embodiments, the determining module 3102 is further configured to:

[0448] Determine the decoding performance of each decoder through at least one encoder.

[0449] In some embodiments, for any one decoder; the determining module 3102 is specifically configured to:

[0450] Respectively detect the performance of the decoder through at least one encoder to obtain at least one second performance;

[0451] Determine the decoding performance according to the at least one second performance.

[0452] In some embodiments, for any one encoder; the determining module 3102 is specifically configured to:

[0453] Encode the fourth data through the encoder to obtain the fifth data;

[0454] The fifth data is decoded by the decoder to obtain sixth data.

[0455] The second performance is determined according to a similarity between the fourth data and the sixth data.

[0456] In some embodiments, the similarity between the fourth data and the sixth data is a cosine similarity between the fourth data and the sixth data.

[0457] In some embodiments, the decoding performance is an average of the at least one second performance; or,

[0458] The decoding performance is a maximum value in the at least one second performance; or,

[0459] The decoding performance is a minimum value in the at least one second performance.

[0460] Optionally, the obtaining module 3101 is configured to perform at least one of the communication steps (for example, steps S2101, S2201, S2301, S2401, S2501, S2601, S2701, but not limited thereto) in any of the above methods, and details are not described herein again. Optionally, the determining module 3102 is configured to perform at least one of the other steps (for example, steps S2102-S2104, S2202-S2204, S2302-S2311, S2402-S2403, S2602-S2407, S2702, but not limited thereto) in any of the above methods, and details are not described herein again.

[0461] FIG. 4a is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure. The electronic device can be a determination device of a decoder, or a chip, chip system, or processor, etc. supporting the determination device of the decoder to implement any of the above methods. The electronic device 4100 can be used to implement the methods described in the above method embodiments, and details can be referred to the descriptions in the above method embodiments.

[0462] As shown in FIG. 4a, the electronic device 4100 includes one or more processors 4101. The processor 4101 can be a general-purpose processor or a special-purpose processor, etc., for example, a baseband processor or a central processor. The baseband processor can be used to process communication protocols and communication data, and the central processor can be used to control a communication device (such as a base station, a baseband chip, a terminal device, a terminal device chip, a DU or a CU, etc.), execute a program, and process data of the program. The electronic device 4100 is configured to implement any of the above methods.

[0463] In some embodiments, the electronic device 4100 further includes one or more memories 4102 for storing instructions. Optionally, all or part of the memories 4102 can also be outside the electronic device 4100.

[0464] In some embodiments, the electronic device 4100 further includes one or more transceivers 4103. When the communication device 4100 includes one or more transceivers 4103, the transceiver 4103 performs at least one of the communication steps (for example, steps S2101, S2201, S2301, S2401, S2501, S2601, S2701, but not limited to) in the above-described methods, such as transmission and / or reception. Here, no further description is given. Optionally, the determination module 3102 is configured to perform at least one of the other steps (for example, steps S2102-S2104, S2202-S2204, S2302-S2311, S2402-S2403, S2602-S2407, S2702, but not limited to) in any of the above methods.

[0465] The processor 4101 performs at least one of the other steps (for example, steps S2102-S2104, S2202-S2204, S2302-S2311, S2402-S2403, S2602-S2407, S2702, but not limited to).

[0466] In some embodiments, the transceiver can include a receiver and / or a transmitter, which can be separate or integrated together. Optionally, the terms transceiver, transceiving unit, transceiver, transceiving circuit, etc. can be replaced with each other, the terms transmitter, transmitting unit, transmitter, transmitting circuit, etc. can be replaced with each other, and the terms receiver, receiving unit, receiver, receiving circuit, etc. can be replaced with each other.

[0467] In some embodiments, the electronic device 4100 can include one or more interface circuits 4104. Optionally, the interface circuit 4104 is connected to the memory 4102, and the interface circuit 4104 can be used to receive signals from the memory 4102 or other devices, and can be used to send signals to the memory 4102 or other devices. For example, the interface circuit 4104 can read instructions stored in the memory 4102 and send the instructions to the processor 4101.

[0468] The electronic device 4100 in the above embodiment description can be a determining device of a decoder, but the scope of the electronic device 4100 described in the present disclosure is not limited thereto, and the structure of the electronic device 4100 can not be limited by FIG. 4. The electronic device can be a standalone device or can be part of a larger device. For example, the electronic device can be: 1) a standalone integrated circuit (IC), or a chip, or a chip system or subsystem; (2) a set of one or more ICs, which can optionally also include storage components for storing data, programs; (3) an ASIC, such as a modem; (4) a module that can be embedded in other devices; (5) a receiver, a terminal device, a smart terminal device, a cellular phone, a wireless device, a handset, a mobile unit, a car-mounted device, a network device, a cloud device, an artificial intelligence device, and the like; (6) other devices, and the like.

[0469] FIG. 4b is a schematic structural diagram of a chip according to an embodiment of the present disclosure. For the case where the electronic device can be a chip or a chip system, the structure of the chip 4200 can be seen from the schematic structural diagram of the chip 4200 shown in FIG. 4b, but is not limited thereto.

[0470] The chip 4200 includes one or more processors 4201, and the chip 4200 is configured to execute any of the above methods.

[0471] In some embodiments, the chip 4200 further includes one or more interface circuits 4202. Optionally, the interface circuit 4202 is connected with the memory 4203, and the interface circuit 4202 can be configured to receive signals from the memory 4203 or other devices, and the interface circuit 4202 can be configured to send signals to the memory 4203 or other devices. For example, the interface circuit 4202 can read instructions stored in the memory 4203 and send the instructions to the processor 4201.

[0472] In some embodiments, the interface circuit 4202 performs at least one of the communication steps such as sending and / or receiving in the above methods, and the processor 4201 performs at least one of the other steps (for example, steps S2121, S2122, S2124, S2125, but not limited thereto).

[0473] In some embodiments, the terms of interface circuit, interface, transceiver pin, transceiver, and the like can be replaced with each other.

[0474] In some embodiments, the chip 4200 further includes one or more memories 4203 for storing instructions. Optionally, all or part of the memory 4203 can be outside the chip 4200.

[0475] The modules and / or devices described in each embodiment of the virtual device, the physical device, the chip, etc. can be combined or separated as appropriate. Alternatively, some or all of the steps can also be performed by a plurality of modules and / or devices in cooperation, which is not limited here.

[0476] The disclosure also provides a storage medium, and the storage medium stores instructions, which, when executed on the electronic device 4100, cause the electronic device 4100 to perform any of the above methods. Alternatively, the storage medium is an electronic storage medium. Alternatively, the storage medium is a computer-readable storage medium, but is not limited to this, and it can also be a storage medium readable by other devices. Alternatively, the storage medium can be a non-transitory storage medium, but is not limited to this, and it can also be a transitory storage medium.

[0477] The disclosure also provides a program product, which includes a program and / or instructions, and the program and / or instructions, when executed by the electronic device 4100, cause the electronic device 4100 to perform any of the above methods. Alternatively, the program product is a computer program product.

[0478] The disclosure also provides a computer program, which, when executed on a computer, causes the computer to perform any of the above methods.

[0479] Those skilled in the art can appreciate that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the disclosure.

[0480] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0481] The above is only a specific implementation of the disclosure, but the protection scope of the disclosure is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the disclosure, which should be covered within the protection scope of the disclosure.

Claims

1. A method of determining for a decoder, characterized in that, The method comprises: obtaining at least one decoder trained; determining a first decoder from the at least one decoder according to decoding performance of the at least one decoder on a plurality of encoders, wherein the decoding performance of the first decoder on the plurality of encoders meets a preset requirement.

2. The method of claim 1, wherein, The at least one decoder is obtained by training on a data set, and the data set comprises channel characteristics.

3. The method of claim 2, wherein, The channel characteristics comprise at least one of the following: a number of devices, wherein the number of devices is a number of devices sending channel state information (CSI) when generating the data set; a number of transmission ports, wherein the number of transmission ports is a number of ports through which the device sends the CSI; a number of subbands, wherein the number of subbands is a number of subbands occupied by the device when sending the CSI; a number of in-phase and quadrature (IQ) components; a number of time slots, wherein the number of time slots is a number of time slots occupied by the device when sending the CSI; or a number of random times, wherein the number of random times is a number of times of randomly generating the data set.

4. The method according to claim 2 or 3, characterized in that, The at least one decoder is obtained by training on a same data set, and the same data set is a collection of a plurality of data sets.

5. The method according to claim 2 or 3, characterized in that, The at least one decoder is obtained by training on different data sets, and a similarity between the different data sets is greater than or equal to a first threshold.

6. The method according to claim 4 or 5, characterized in that, The decoding performance of the first decoder on the plurality of encoders meets the preset requirement, and comprises at least one of the following: the decoding performance of the first decoder is greater than or equal to a second threshold; in descending order of decoding performance, the first decoder is a first K decoder in the at least one decoder, wherein K is an integer greater than or equal to 1; or a difference in decoding performance of the first decoder on the plurality of encoders is less than a third threshold.

7. The method of claim 2 or 3, wherein, The at least one decoder is obtained by training on at least one data set.

8. The method of claim 7, wherein, The first decoder is determined from the at least one decoder according to decoding performance of the at least one decoder on a plurality of encoders, and comprises: when a similarity between the at least one data set is greater than or equal to a first threshold, the first decoder is determined according to the decoding performance of the at least one decoder on the plurality of encoders; when the similarity between the at least one data set is less than the first threshold, a performance similarity of the at least one decoder is determined, and the first decoder is determined according to the performance similarity.

9. The method of claim 8, wherein, The similarity between the at least one data set is a cosine similarity between the at least one data set.

10. The method according to claim 8 or 9, characterized in that, The method further comprises: obtaining decoding performance of a preset decoder on encoded data of the at least one data set; if a difference between decoding performance corresponding to the at least one data set is less than a fourth threshold, it is determined that a similarity between the at least one data set is greater than or equal to a first threshold.

11. The method according to claim 8 or 9, characterized in that, The first decoder is determined according to the decoding performance of the at least one decoder on a plurality of encoders, and comprises: decoders in the at least one decoder that meet at least one of the following conditions are determined as the first decoder: the decoding performance of the at least one decoder on the plurality of encoders is greater than or equal to a second threshold; in descending order of decoding performance, the first decoder is a first K decoder in the at least one decoder, wherein K is an integer greater than or equal to 1; or The decoding performance difference of the plurality of encoders is less than a third threshold value.

12. The method of claim 8 or 9, wherein, The performance similarity of the at least one decoder is determined, including: The first performance of each decoder is determined through a plurality of first data. The performance similarity of the at least one decoder is determined according to the first performance of each decoder.

13. The method of claim 12, wherein, For any one decoder; The first performance of the decoder is determined through a plurality of first data, including: Each first data is encoded by the encoder corresponding to the decoder to obtain second data corresponding to each first data; Each first data is decoded by the decoder to obtain third data corresponding to each first data; The first performance of the decoder is determined according to each first data and the third data corresponding to the first data.

14. The method of claim 13, wherein, The first performance of the decoder is determined according to each first data and the third data corresponding to the first data, including: The data similarity between each first data and the corresponding third data is determined to obtain a plurality of first similarities; The first performance is determined according to the plurality of first similarities.

15. The method according to any one of claims 8-14, characterized in that, The first decoder is determined according to the performance similarity, including: When the performance similarity is greater than or equal to a preset performance threshold value, a decoder in the at least one decoder that meets at least one of the following conditions is determined as the first decoder: the decoding performance of the plurality of encoders is greater than or equal to a second threshold value; in descending order of decoding performance, the first decoder is the first K decoders in the at least one decoder, and K is an integer greater than or equal to 1; or the decoding performance difference of the plurality of encoders is less than a third threshold value; Or, when the performance similarity is less than the preset performance threshold value, a decoder is trained through the at least one data set to obtain the first decoder.

16. The method according to any one of claims 1 to 15, characterized in that, The at least one decoder obtained by training includes: The at least one second device sends the at least one decoder.

17. The method according to any one of claims 1 to 15, characterized in that, The method is performed by a first device; The number of the at least one decoder is M, and M is an integer greater than 1; The at least one decoder obtained by training includes: The N decoders are sent by at least one second device. Wherein, N is an integer greater than 1 and less than M; M-N decoders in the M decoders except the N decoders are obtained by training of the first device.

18. The method according to any one of claims 1 to 15, characterized in that, The method is performed by a first device; and the at least one decoder is obtained by training of the first device.

19. The method according to any one of claims 1 to 18, characterized in that, The method further includes: The decoding performance of each decoder is determined through at least one encoder.

20. The method of claim 19, wherein, For any one decoder; The decoding performance of the decoder is determined through at least one encoder, including: The at least one second performance is obtained by respectively performing performance detection on the decoder through the at least one encoder; The decoding performance is determined according to the at least one second performance.

21. The method of claim 20, wherein, For any one encoder; the second performance is obtained by performing performance detection on the decoder through the encoder, including: The fifth data is obtained by performing encoding processing on the fourth data through the encoder; The fifth data is obtained by performing encoding processing on the fourth data through the encoder; The fifth data is decoded by the decoder to obtain sixth data; The second performance is determined according to the similarity between the fourth data and the sixth data.

22. The method of claim 21, wherein, The similarity between the fourth data and the sixth data is a cosine similarity between the fourth data and the sixth data.

23. The method of any one of claims 20-22, wherein, The decoding performance is an average of the at least one second performance; or, The decoding performance is a maximum of the at least one second performance; or, The decoding performance is a minimum of the at least one second performance.

24. A determining apparatus of a decoder, characterized by, Comprising: An acquisition module configured to acquire at least one decoder trained; A determination module configured to determine a first decoder from the at least one decoder according to decoding performances of the plurality of encoders on the at least one decoder, the decoding performance of the first decoder on the plurality of encoders satisfying a preset requirement.

25. A determining device of a decoder, the determining device comprising: Comprising: One or more processors; The processor is configured to execute the determination method of the decoder according to any one of claims 1-23.

26. A storage medium, the storage medium storing instructions, wherein, When the instructions are executed on the electronic device, the determination method of the decoder according to any one of claims 1-23 is implemented.

27. A computer program product, characterised in that, The program and / or instructions are executed by the electronic device, so that the communication device executes the determination method of the decoder according to any one of claims 1-23.

Images