Perceptual channel information acquisition method and device, model training method and device and readable medium

Abstract

The invention provides a perception channel information acquisition method, a perception channel information model training method, perception channel information acquisition equipment, perception channel information model training equipment and a readable medium. The perception channel information acquisition method comprises the steps that network equipment receives communication channel state information from terminal equipment; and the network equipment acquires the sensing channel information based on a preset sensing channel information acquisition model according to the communication channel state information. Through extrapolation of the communication channel state information which is easy to obtain, the sensing channel information which is difficult to obtain is obtained, the information obtaining difficulty is reduced, and the working efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, and in particular to a sensing channel information acquisition method and device, and a readable medium. BACKGROUND

[0002] This section is intended to provide background information to facilitate a better understanding of embodiments of the present application recited in the claims. The description herein does not constitute admission of prior art.

[0003] Integrated Sensing and Communication (ISAC) is gradually becoming a core technology for the next generation of mobile networks. This innovative approach fuses communication and sensing functions into one system by sharing radio resources, hardware, and waveforms. ISAC not only enables traditional communication functions, but also provides precise positioning, human activity recognition, and target sensing, among other functions. By integrating these functions, ISAC opens up new application possibilities for mobile operators in multiple vertical industries such as Industrial Internet of Things (IIoT), immersive digital experiences, and smart homes.

[0004] In the field of ISAC, parameter extraction plays a key role in converting radio echoes into important information about the surrounding environment and its objects. Although parameter extraction algorithms have been developed for decades in the field of radar sensing, such as Multiple Signal Classification (MUSIC), Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT), and Compressive Sensing (CS).

[0005] However, the above-mentioned parameter extraction algorithms have the following disadvantages: 1) the MUSIC algorithm requires the number of signal sources to be obtained in advance, otherwise the parameter estimation will be inaccurate; 2) the ESPRIT algorithm requires the signal arrangement to be ordered, otherwise the signal sources will be too close or the arrangement will be disordered, resulting in inaccurate estimation results; 3) the CS algorithm requires the sparsity of the signal in a specific transform domain, and if the signal is not sparse or the degree of sparsity is insufficient, the algorithm cannot extract the signal; at the same time, the CS algorithm has limited application scenarios and is not universal.

[0006] Based on this, the present application provides a sensing channel information acquisition method and device, and a readable medium, which can solve the above technical problems. SUMMARY

[0007] Aspects of the present application provide a sensing channel information acquisition method and device, and a readable medium, which can achieve the extrapolation of sensing channel information according to communication channel state information.

[0008] In an aspect of the present application, a sensing channel information acquisition method is provided, which is applied to a network device, and the method comprises:

[0009] receiving communication channel state information from a terminal device;

[0010] According to the communication channel state information, the sensing channel information is acquired.

[0011] According to another aspect of the present application, a sensing channel information acquisition method is provided, which is applied to a terminal device, and the method comprises:

[0012] Acquiring communication channel state information;

[0013] According to the communication channel state information, the sensing channel information is acquired.

[0014] Further, the method further comprises: the terminal device sending the sensing channel information to a network device.

[0015] According to another aspect of the present application, a sensing channel information acquisition method is provided, which is applied to a terminal device and a network device, and the method comprises:

[0016] The terminal device sends communication channel state information to the network device;

[0017] The network device acquires sensing channel information according to the communication channel state information.

[0018] According to another aspect of the present application, a sensing channel information acquisition method is provided, which is applied to a terminal device and a network device, and the method comprises:

[0019] The terminal device sends sensing channel information acquired according to communication channel state information to the network device;

[0020] The network device receives the sensing channel information sent by the terminal device.

[0021] Based on the above aspects, a sensing channel information acquisition method is provided, and further, the method further comprises: acquiring communication sensing integrated information; and acquiring the communication channel state information based on the communication sensing integrated information.

[0022] Further, the communication channel state information comprises: communication channel state information acquired by parsing the communication sensing integrated information, and / or communication channel state information not acquired by parsing the communication sensing integrated information.

[0023] Further, acquiring the sensing channel information according to the communication channel state information comprises: sending the communication channel state information to a preset sensing channel information acquisition model to acquire the sensing channel information.

[0024] Further, the sending the communication channel state information to the preset perception channel information acquisition model to acquire the perception channel information comprises: sending the communication channel state information to the preset perception channel information acquisition model; the preset perception channel information acquisition model acquires a characteristic element of the communication channel state information according to the communication channel state information; and the characteristic element of the communication channel state information is used to acquire the perception channel information.

[0025] Further, the characteristic element comprises channel gain, phase information, power time delay spread, received signal arrival time and arrival angle.

[0026] In another aspect of the present application, a perception channel information acquisition model training method is provided for implementing the perception channel information acquisition method described above, and the method comprises:

[0027] The first communication device receives second indication information from the second communication device, the second indication information comprising communication channel state information and perception channel information, and the second indication information is used to instruct the first communication device to train an initial perception channel information acquisition model according to the communication channel state information and the perception channel information.

[0028] The first communication device trains the initial perception channel information acquisition model according to the perception channel information and the communication channel state information until the initial perception channel information model meets a preset condition, and a perception channel information acquisition model is obtained.

[0029] Further, the training of the initial perception channel information acquisition model by the first communication device according to the perception channel information and the communication channel state information comprises: when the communication channel state information does not contain a spatial position feature, the first communication device sends a control signal to the initial perception channel information acquisition model; the control signal comprises spatial position coding, and the control signal is used to instruct the initial perception channel information acquisition model to generate the spatial position feature of the communication channel state information according to the spatial position coding.

[0030] Further, before the training of the initial perception channel information acquisition model by the first communication device according to the perception channel information and the communication channel state information, the method further comprises: performing normalization processing on the perception channel information and the communication channel state information.

[0031] Further, the preset condition comprises: a loss value of the initial perception channel information model is not greater than a first threshold value.

[0032] In another aspect of the present application, a perception channel information acquisition device is provided, and the device comprises:

[0033] receiving a communication channel state information from a terminal device;

[0034] obtaining the sensing channel information according to the communication channel state information.

[0035] In yet another aspect of the present application, a sensing channel information obtaining device is provided, which comprises:

[0036] a first obtaining module for obtaining a communication channel state information;

[0037] a second obtaining module for obtaining a sensing channel information based on the communication channel state information.

[0038] In yet another aspect of the present application, a sensing channel information obtaining device is provided, which comprises:

[0039] a sending module for sending a communication channel state information to the network device;

[0040] an obtaining module for obtaining the sensing channel information according to the communication channel state information.

[0041] In yet another aspect of the present application, a sensing channel information obtaining device is provided, which comprises:

[0042] a sending module for sending a sensing channel information obtained according to a communication channel state information to the network device;

[0043] a receiving module for receiving the sensing channel information sent from the terminal device.

[0044] In yet another aspect of the present application, a sensing channel information obtaining model training device is provided, which comprises:

[0045] a receiving module for receiving a first indication information from a second communication device, the first indication information comprising a communication channel state information and a sensing channel information, the first indication information being used for instructing the first communication device to train an initial sensing channel information obtaining model according to the communication channel state information and the sensing channel information;

[0046] a training module for training the initial sensing channel information obtaining model according to the sensing channel information and the communication channel state information until a preset sensing channel information model meets a preset condition, thereby obtaining a trained sensing channel information obtaining model.

[0047] In another aspect of the present application, an electronic device is provided, which comprises:

[0048] at least one processor; and

[0049] a memory in communication connection with the at least one processor; wherein

[0050] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the perception channel information acquisition method and the perception channel information acquisition model training method.

[0051] In another aspect of the present application, a computer readable storage medium is provided, which stores computer program instructions executable by a processor to implement the perception channel information acquisition method and the perception channel information acquisition model training method.

[0052] In another aspect of the present application, a computer program product is provided, which includes a computer program executable by a processor to implement the perception channel information acquisition method and the perception channel information model training method.

[0053] The perception channel information acquisition method proposed in the present application obtains the difficult-to-obtain perception channel information by extrapolation based on the easily-obtained communication channel state information, thereby reducing the difficulty of obtaining the perception channel information and improving the efficiency of obtaining the perception channel information.

[0054] Further, the perception channel information acquisition model training method proposed in the present application generates a training set based on the communication channel state information and the perception channel information, trains an initial perception channel information acquisition model, learns the complex spatio-temporal correlation between the communication channel state information and the perception channel information, and achieves the beneficial effect that the perception channel information can be extrapolated based on the communication channel state information based on the trained perception channel information acquisition model.

[0055] Further, the present application also creatively proposes that when the communication channel state information does not contain spatial position features, spatial position encoding is set to generate spatial position features of the communication channel state information, so that the model can better learn the relationship and correlation between the communication channel state information and the perception channel information; on the contrary, when the communication channel state information contains spatial position features, spatial position encoding is not set to reduce the additional interference of the position space encoding on the model and the training burden of the model, and further achieve the beneficial effects of reducing the training time of the model, improving the training efficiency of the model, and improving the performance of the model.

[0056] Further, the application further creatively proposes that before training the initial perception channel information model according to the perception channel information and the communication channel state information, the perception channel information and the communication channel state information are normalized, so that the perception channel information and the communication channel state information are in accordance with the standard normal distribution, the dimensional difference between the perception channel information and the communication channel state information is eliminated, and the convergence speed of the model is improved. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0058] Other features, objects and advantages of the present application will become more apparent from the following detailed description of the non-limiting embodiments, with reference to the accompanying drawings:

[0059] Figure 1 The flowchart of the perception channel information acquisition method provided by an embodiment of the present application is shown in the figure;

[0060] Figure 2 The flowchart of the perception channel information acquisition method provided by another embodiment of the present application is shown in the figure;

[0061] Figure 3 The structural diagram of the perception channel information acquisition method provided by another embodiment of the present application is shown in the figure;

[0062] Figure 4 The structural diagram of the perception channel information acquisition method provided by another embodiment of the present application is shown in the figure;

[0063] Figure 5 The flowchart of the perception channel information acquisition model training method provided by another embodiment of the present application is shown in the figure;

[0064] Figure 6 The measured communication channel state information in an embodiment of the present application is shown in the figure;

[0065] Figure 7 The measured perception channel information in an embodiment of the present application is shown in the figure;

[0066] Figure 8 The simulation result diagram of the improved Transformer model and the standard Transformer model respectively extrapolating the perception channel information from the communication channel state information in the LOS scene and the NLOS scene in an embodiment of the present application is shown in the figure;

[0067] Figure 9 A structure diagram of a sensing channel information acquisition device provided by an embodiment of the present application is shown in FIG. 1;

[0068] Figure 10 A structure diagram of a sensing channel information acquisition device provided by another embodiment of the present application is shown in FIG. 2;

[0069] Figure 11 A structure diagram of a sensing channel information acquisition device provided by yet another embodiment of the present application is shown in FIG. 3;

[0070] Figure 12 A structure diagram of a sensing channel information acquisition model training device provided by an embodiment of the present application is shown in FIG. 4;

[0071] Figure 13 A structure diagram of an electronic device suitable for implementing the scheme in the embodiments of the present application is shown in FIG. 5;

[0072] The same or similar reference signs in the drawings represent the same or similar components. DETAILED DESCRIPTION

[0073] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0074] In a typical configuration of the present application, the devices of the terminal and the service network each include one or more processors (CPU), input / output interfaces, network interfaces and memories.

[0075] The memory can include a non-permanent memory in a computer readable medium, random access memory (RAM) and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of the computer readable medium.

[0076] Computer-readable media includes permanent and non-permanent, removable and non-removable media implemented by any method or technology for information storage. Information can be computer program instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape disk storage or other magnetic storage device, or any other non-transmission medium that can be used to store information accessible by a computing device.

[0077] The embodiment of the application provides a kind of perception channel information acquisition method, the method comprises:

[0078] The network device receives the communication channel state information from terminal device;

[0079] The network device obtains the perception channel information based on the preset perception channel information acquisition model according to the communication channel state information.

[0080] In actual scene, the execution subject of the method can be user equipment, or the device integrated by user equipment and network equipment through network, or also can be application program running on the above device, the user equipment includes but is not limited to computer, mobile phone, tablet computer, smart watch, bracelet and various terminal devices, and the network equipment includes but is not limited to network host, single network server, multiple network server set or cloud computing based computer set, which can be used to realize part of processing function when setting alarm clock. Here, cloud is composed of a large number of hosts or network servers based on cloud computing (Cloud Computing), wherein, cloud computing is a kind of distributed computing, which is composed of a virtual computer of a group of loosely coupled computer set.

[0081] Embodiment one

[0082] Figure 1 The processing flow of the perception channel information acquisition method provided by the embodiment of the application is shown, and the method at least includes the following processing steps:

[0083] Step S101, the network device receives the communication channel state information from terminal device;

[0084] Step S102, the network device obtains the perception channel information according to the communication channel state information.

[0085] In one of the embodiments, the network device acquires the sensing channel information according to the communication channel state information, which includes that the network device sends the communication channel state information to a preset sensing channel information acquisition model to acquire the sensing channel information. The sensing channel information acquisition model is arranged in the network device.

[0086] In one of the embodiments, the communication channel state information is acquired based on the communication-sensing integrated information.

[0087] In one of the embodiments, the preset sensing channel information acquisition model includes, but is not limited to, a standard Transformer model and an improved Transformer model trained in Embodiment Five, and an autoencoder model trained in Embodiment Six.

[0088] In one of the embodiments, the network device sends the communication channel state information to the preset sensing channel information acquisition model to acquire the sensing channel information, which includes that the network device sends the communication channel state information to the preset sensing channel information acquisition model to acquire the feature elements of the communication channel state information; and the preset sensing channel information acquisition model acquires the sensing channel information according to the feature elements of the communication channel state information.

[0089] In one of the embodiments, the communication channel state information includes communication channel state information acquired from the communication-sensing integrated information, and / or communication channel state information not acquired from the communication-sensing integrated information.

[0090] In one of the embodiments, the network device sends the communication channel state information to the standard Transformer model or the improved Transformer model to acquire the sensing channel information, which includes that a communication channel state information sequence with a dimension of [batch_size, input_step, input_dim] is input into an encoder after being processed by an Embedding layer and spatial position coding, and a zero vector with a dimension of [batch_size, output_step, output_dim] is input into a decoder after being processed by the same. The encoder extracts the potential features of the communication channel state information sequence and sends them to the decoder. The decoder outputs a sensing channel information sequence according to the correlation between the communication channel state information and the sensing channel information learned during training, so as to realize the extrapolation of unknown sensing channel information from known communication channel state information.

[0091] In one of the embodiments, the terminal device sends the communication channel state information to the autoencoder model, which includes that the actually measured communication channel state information is input, and the decoder outputs the predicted sensing channel information.

[0092] In one of the embodiments, the characteristic elements include channel gain, phase information, power delay spread, received signal time of arrival and angle of arrival.

[0093] In one of the embodiments, the terminal device can be a smart phone, a vehicle-mounted terminal, a drone, an Internet of Things terminal device (such as a smart security camera, a smart door lock), etc.; and the network device can be a base station, a road side unit, a WiFi router, a gateway, etc.

[0094] It should be understood that in the prior art, the transmission mechanism of the communication channel state information (the terminal device transmits the communication channel state information to the network device) is already very complete, and the transmission mechanism of the perception channel information (the terminal device transmits the perception channel information to the network device) has not been established / has not been completely established for application. Therefore, if the terminal device sends the communication channel state information to the network device, and then the network device analyzes and obtains the perception channel information, there is no need to re-establish a new transmission mechanism of the perception channel information.

[0095] Embodiment Two

[0096] Figure 2 A processing flow of a method for obtaining perception channel information provided by an embodiment of the present application is shown, which includes at least the following processing steps:

[0097] Step S201: A terminal device obtains communication channel state information.

[0098] Step S202: The terminal device obtains perception channel information according to the communication channel state information.

[0099] In one of the embodiments, the method further includes: the terminal device sends the perception channel information to a network device.

[0100] In one of the embodiments, obtaining the communication channel state information includes: obtaining communication-perception integrated information; and obtaining the communication channel state information based on the communication-perception integrated information.

[0101] In one of the embodiments, the terminal device obtains the perception channel information according to the communication channel state information includes: the terminal device sends the communication channel state information to a preset perception channel information obtaining model to obtain the perception channel information. The preset perception channel information obtaining model is arranged in the terminal device.

[0102] In one of the embodiments, the preset perception channel information obtaining model includes but is not limited to a standard Transformer model and an improved Transformer model trained based on Embodiment Five, and an auto-encoder model trained based on Embodiment Six.

[0103] In one of the embodiments, the terminal device sends the communication channel state information to a preset perception channel information acquisition model to acquire the perception channel information, including: the terminal device sends the communication channel state information to a preset perception channel information acquisition model to acquire the feature elements of the communication channel state information; the preset perception channel information acquisition model acquires the perception channel information according to the feature elements of the communication channel state information.

[0104] In one of the embodiments, the communication channel state information includes: the communication channel state information acquired from the communication-perception integrated information analysis, and / or the communication channel state information not acquired from the communication-perception integrated information analysis.

[0105] In one of the embodiments, the terminal device sends the communication channel state information to a standard Transformer model or an improved Transformer model to acquire the perception channel information, including: inputting the communication channel state information sequence with the dimension of [batch_size, input_step, input_dim] after Embedding layer and spatial position coding processing into the encoder, and inputting the all-zero vector with the dimension of [batch_size, output_step, output_dim] after the same processing into the decoder. The encoder extracts the potential features in the communication channel state information sequence and sends them to the decoder, the decoder outputs the perception channel information sequence according to the correlation between the communication channel state information and the perception channel information learned during training, to realize the extrapolation of unknown perception channel information from known communication channel state information.

[0106] In one of the embodiments, the terminal device sends the communication channel state information to the auto-encoder model, including: the measured communication channel state information as input, and the decoder output as the predicted perception channel information.

[0107] In one of the embodiments, the terminal device can be a smart phone, a vehicle-mounted terminal, a drone, an Internet of Things terminal device (such as a smart security camera, a smart door lock), etc.; the network device can be a base station, a roadside unit, a WiFi router, a gateway, etc.

[0108] It needs to be understood that if the terminal device sends the communication channel state information to the network device, and the network device obtains the sensing channel information based on the communication channel state information, the distortion of the sensing channel information will be aggravated. The reason is that in the process of transmitting the communication channel state information from the terminal device to the network device, the communication channel state information will be compressed, so that some information points in the communication channel state information will be discarded. Then, when the sensing channel information is obtained based on the communication channel state information with insufficient integrity, the accuracy will inevitably be reduced. However, when the sensing channel information is obtained directly based on the obtained communication channel state information at the terminal device side, and then transmitted to the network device, the above problem will not exist.

[0109] Embodiment three

[0110] Figure 3 A processing flow of a sensing channel information obtaining method provided by an embodiment of the present application is shown, and the method comprises at least the following processing steps:

[0111] Step S301, the terminal device sends the communication channel state information to the network device.

[0112] Step S302, the network device obtains the sensing channel information according to the communication channel state information.

[0113] In one of the embodiments, the communication channel state information is obtained by the network device from the communication-sensing integrated information.

[0114] In one of the embodiments, the network device obtaining the sensing channel information according to the communication channel state information comprises: the network device obtaining the sensing channel information according to the communication channel state information based on the indication information.

[0115] Further, the network device obtaining the sensing channel information according to the communication information comprises: the network device sending the communication channel state information to a preset sensing channel information obtaining model to obtain the sensing channel information.

[0116] In one of the embodiments, the preset sensing channel information obtaining model can be arranged in the network device.

[0117] Further, the network device sending the communication channel state information to the preset sensing channel information obtaining model to obtain the sensing channel information comprises: the network device sending the communication channel state information to the preset sensing channel information obtaining model to obtain the characteristic elements of the communication channel state information; and the preset sensing channel information obtaining model obtaining the sensing channel information according to the characteristic elements of the communication channel state information.

[0118] Further, the communication channel state information includes: communication channel state information obtained from the communication perception integration information analysis, and / or communication channel state information not obtained from the communication perception integration information analysis.

[0119] Further, the characteristic elements include channel gain, phase information, power delay spread, received signal arrival time and angle of arrival.

[0120] In one embodiment, the preset perception channel information acquisition model includes, but is not limited to, the standard Transformer model and the improved Transformer model trained based on example five, and the autoencoder model trained based on example six.

[0121] In one embodiment, the network device sends the communication channel state information to the standard Transformer model or the improved Transformer model to obtain the perception channel information, including: inputting the communication channel state information sequence with the dimension of [batch_size, input_step, input_dim] after Embedding layer and spatial position coding processing into the encoder, and inputting the all-zero vector with the dimension of [batch_size, output_step, output_dim] after the same processing into the decoder. The encoder extracts the potential features in the communication channel state information sequence and sends them to the decoder. The decoder outputs the perception channel information sequence according to the correlation between the communication channel state information and the perception channel information learned during training, to realize the extrapolation of unknown perception channel information from known communication channel state information.

[0122] In one embodiment, the network device sends the communication channel state information to the autoencoder model, including: inputting the measured communication channel state information as input, and outputting the predicted perception channel information from the decoder.

[0123] In one embodiment, the terminal device can be a smartphone, a vehicle-mounted terminal, a drone, an Internet of Things terminal device (such as a smart security camera, a smart door lock), etc.; the network device can be a base station, a road side unit, a WiFi router, a gateway, etc.

[0124] It should be understood that in the prior art, the transmission mechanism of the communication channel state information (the terminal device transmits the communication channel state information to the network device) is already very complete, while the transmission mechanism of the perception channel information (the terminal device transmits the perception channel information to the network device) has not been established / has not been completely established for application. Therefore, if the terminal device sends the communication channel state information to the network device, and then the network device analyzes and obtains the perception channel information, there is no need to establish a new transmission mechanism for the perception channel information.

[0125] Embodiment Four

[0126] Figure 4 A method for obtaining perception channel information is shown, which comprises the following steps:

[0127] In step S401, the terminal device sends the perception channel information obtained according to the communication channel state information to the network device.

[0128] In step S402, the network device receives the perception channel information sent by the terminal device.

[0129] In one embodiment, the terminal device obtains the communication channel state information, which comprises the following steps: the terminal device obtains the communication-perception integrated information; and the terminal device obtains the communication channel state information according to the communication-perception integrated information.

[0130] In one embodiment, the terminal device obtains the perception channel information according to the communication channel state information, which comprises the following steps: the terminal device sends the communication channel state information to a preset perception channel information obtaining model to obtain the perception channel information. The preset perception channel information obtaining model is set in the terminal device.

[0131] In one embodiment, the preset perception channel information obtaining model includes, but is not limited to, a standard Transformer model and an improved Transformer model trained according to Embodiment Five, and an auto-encoder model trained according to Embodiment Six.

[0132] In one embodiment, the terminal device sends the communication channel state information to the preset perception channel information obtaining model to obtain the perception channel information, which comprises the following steps: the terminal device sends the communication channel state information to the preset perception channel information obtaining model to obtain the feature elements of the communication channel state information; and the preset perception channel information obtaining model obtains the perception channel information according to the feature elements of the communication channel state information.

[0133] In one embodiment, the communication channel state information includes the communication channel state information obtained from the communication-perception integrated information, and / or the communication channel state information not obtained from the communication-perception integrated information.

[0134] In one of the embodiments, the terminal device sends the communication channel state information to the standard Transformer model or the improved Transformer model to obtain the sensing channel information, including: inputting the communication channel state information sequence with the dimension of [batch_size, input_step, input_dim] after Embedding layer and spatial position coding processing into the encoder, and inputting the zero vector with the dimension of [batch_size, output_step, output_dim] after the same processing into the decoder. The encoder extracts the potential features in the communication channel state information sequence and sends them to the decoder. The decoder outputs the sensing channel information sequence according to the correlation between the communication channel state information and the sensing channel information learned during training, to realize the extrapolation of unknown sensing channel information from known communication channel state information.

[0135] In one of the embodiments, the terminal device sends the communication channel state information to the autoencoder model, including: taking the measured communication channel state information as input, and taking the predicted sensing channel information as output of the decoder.

[0136] In one of the embodiments, the terminal device can be a smart phone, a vehicle-mounted terminal, a drone, an Internet of Things terminal device (such as a smart security camera, a smart door lock), etc.; the network device can be a base station, a road side unit, a WiFi router, a gateway, etc.

[0137] It should be understood that if the communication channel state information is sent to the network device to obtain the sensing channel information based on the communication channel state information by the network device, the distortion of the sensing channel information will be aggravated; the reason is that in the process of transmitting the communication channel state information from the terminal device to the network device, the communication channel state information will be compressed, so that some information points in the communication channel state information will be discarded; obtaining the sensing channel information based on the incomplete communication channel state information will inevitably reduce its accuracy. However, if the sensing channel information is obtained directly based on the obtained communication channel state information at the terminal device side and then sent to the network device, the above technical problems will not exist.

[0138] Embodiment five

[0139] Figure 5 A processing flow of a sensing channel information acquisition model training method provided by an embodiment of the present application is shown, and the method includes:

[0140] Step S501, the first communication device receives second indication information from the second communication device, the second indication information including communication channel state information and sensing channel information, the second indication information being used to instruct the first communication device to train an initial sensing channel information acquisition model according to the communication channel state information and the sensing channel information;

[0141] Step S502, the first communication device trains the initial sensing channel information acquisition model according to the sensing channel information and the communication channel state information until the initial sensing channel information acquisition model meets a preset condition, obtaining a sensing channel information acquisition model.

[0142] In one of the embodiments, the initial sensing channel information acquisition model includes a standard Transformer model and an improved Transformer model.

[0143] In one of the embodiments, the known communication channel state information sequence and the to-be-predicted sensing channel information sequence are respectively input into an Embedding layer with [batch_size, input_step, input_dim] and [batch_size, output_step, output_dim] dimensions. The Embedding layer embeds data into a high-dimensional vector space through a linear transformation, i.e., maps the input communication channel state information and sensing channel information sequences to [batch_size, input_step, d_model] and [batch_size, output_step, d_model] dimensions respectively. Then, spatial position encoding is used to add position information to the communication channel state information and sensing channel information sequences. Because the standard Transformer model learns the features and correlations between data through a self-attention mechanism and cannot perceive the input order of data, spatial position encoding is needed to explicitly add order information to the input. Spatial position encoding can be generated by a fixed function or learned parameters, which maps the position information into a vector that is added to the communication channel state information and sensing channel information sequences to introduce the position information into the representation of the sequences. This process does not change the dimension of the sequences. Therefore, the dimension of the communication channel state information sequence is [batch_size, input_step, d_model] and the dimension of the sensing channel information sequence is [batch_size, output_step, d_model]. As shown in FIG. 3, it is an example diagram of the measured communication channel state information. As shown in FIG. 4, it is an example diagram of the measured sensing channel information. Figure 6 Figure 7

[0144] ​​In one embodiment, the standard Transformer model includes an encoder and a decoder.

[0145] In one embodiment, the encoder is stacked by N layers of identical sub-layers, each of which includes two main modules: a Multi-Head Self-Attention module and a Feed-Forward Network module. To prevent the problem of gradient vanishing and information loss in the encoder, a residual connection is used after the two modules of each layer sub-layer, which allows the input to be directly added to the output of the sub-layer, thereby ensuring that the communication channel state information and the perception channel information can be directly passed through the network. In addition, layer normalization is also used to ensure that the outputs of different layers have consistent distributions in different batches of communication channel state information and perception channel information, further improving the stability of training.

[0146] In the self-attention mechanism, for each feature element in the input communication channel state information sequence, different weight matrices are used to calculate its query (Q), key (K) and value (V), respectively. The Q of each feature element is dot multiplied with the K of other feature elements to calculate the similarity score, and the score is normalized by the Softmax function to obtain the attention weight of each feature element. Then, the weighted sum of all V in the communication channel state information sequence is obtained to obtain the output of the attention mechanism. The multi-head attention mechanism performs the attention algorithm h times in parallel and uses different weight matrices each time. Each head captures different dependencies and features, and finally the outputs of all attention heads are concatenated and mapped back to the output space through a linear layer. This module can help the model to focus on different parts of the communication channel state information sequence in different subspaces, thereby better capturing the complex dependencies in the sequence. The output of the multi-head self-attention mechanism is sent to the feed-forward neural network, in which the input communication channel state information sequence is mapped to a higher dimension, i.e. 4*d_model dimension, through linear transformation, then nonlinearly transformed through the ReLU activation function, and finally mapped back to the d_model dimension through a linear layer and output. The self-attention mechanism is essentially linear, while the feed-forward neural network can perform more complex mapping on the input through nonlinear transformation, helping the model to learn complex patterns in the sequence and enhancing the nonlinear representation ability of the model.

[0147] In one of the embodiments, the decoder is also stacked by N identical sub-layers, each of which includes three main modules: Masked Multi-Head Self-Attention module, Encoder-Decoder Attention module and Feed-Forward Neural Network module. Like the encoder, the modules of each sub-layer are also followed by a residual connection and layer normalization, so that the output of each layer is stable and easy to train.

[0148] The Masked Multi-Head Self-Attention module is similar to the Multi-Head Self-Attention module in the encoder, which generates a new representation by calculating the attention weight of each feature element in the input perception channel information sequence. The mask is introduced to prevent the leakage of future information to be predicted, which prevents the decoder from accessing the sequence of future spatial points when generating the perception channel information sequence of the next spatial point, and only allows access to the perception channel information sequence of the current and previous spatial points. The second attention module of the decoder is the cross-attention mechanism, in which Q comes from the output of the previous module of the decoder, containing the perception channel information sequence information, and K and V come from the output of the encoder, containing the communication channel state information sequence information. The cross-attention mechanism uses Q, K, and V to calculate the attention weight and weighted sum, thereby learning the correlation between the communication channel state information sequence and the perception channel information sequence, and using the learned rule to predict the perception channel information sequence based on the communication channel state information sequence output by the encoder. The Feed-Forward Neural Network module is exactly the same as the encoder, which introduces a nonlinear activation function to enhance the nonlinear representation ability of the model.

[0149] After several encoder and decoder stack processing, the model outputs the perception channel information sequence of the position to be predicted, with a dimension of [batch_size, output_step, d_model], which is mapped to a dimension of [batch_size, output_step, output_dim] by a Linear layer. In the training process, the loss function is used to calculate the difference between the output perception channel information sequence and the actual measured perception channel information sequence to observe the training of the model. When the loss function converges to a lower value, it proves that the model has successfully learned the correlation between the communication channel state information and the perception channel information.

[0150] In one of the embodiments, when the communication channel state information does not contain spatial position information, in order to reduce the model training burden and improve the prediction efficiency of the perceived channel information, the standard Transformer model is improved. Compared with the standard Transformer model, the improved Transformer model removes the spatial position encoding.

[0151] Further, in order to improve the training efficiency of the model, that is, to improve the convergence speed of the model, compared with the standard Transformer model, the improved Transformer model also normalizes the perceived channel information and the communication channel state information before training.

[0152] In one of the embodiments, the second communication device can be a network device or a terminal device.

[0153] As Figure 8 shown in the simulation results of the improved Transformer model and the standard Transformer model respectively extrapolating the perceived channel information from the communication channel state information in the LOS scenario and the NLOS scenario. In the figure, the blue solid line represents the measured perceived channel information, and the red dashed line represents the predicted perceived channel information. Figure 8 (a) is the simulation result schematic diagram of the standard Transformer model in the LOS scenario. Figure 8 (b) is the simulation result schematic diagram of the improved Transformer model in the LOS scenario. Figure 8 (c) is the simulation result schematic diagram of the standard Transformer model in the NLOS scenario. Figure 8 (d) is the simulation result schematic diagram of the improved Transformer model in the NLOS scenario. Obviously, in the LoS and NLoS scenarios, the extrapolation error RMSE based on the Transformer model without spatial position encoding (improved Transformer model) is lower than that based on the standard Transformer model, which proves that removing the spatial position encoding can significantly improve the performance of extrapolating the perceived channel information from the communication channel state information, and the model can better predict the peak amplitude and the corresponding delay point.

[0154] In addition, compared with Figure 8 (b) and Figure 8(d) It can be seen that the model extrapolation effect in the LoS scenario is better than that in the NLoS scenario. Although the RMSE of the LoS scenario is higher than that of the NLoS scenario, the fitting degree between the extrapolated perception channel information and the measured perception channel information is higher than that in the NLoS scenario, because the peak value of the perception channel information in the LoS scenario is much higher, so the overall error looks larger. In fact, the propagation characteristics of the LoS channel are relatively simple, and the perception channel information only shows several strong peaks, which is conducive to capturing the main features of the model. In contrast, in the NLoS channel, the signal will undergo multiple reflections, refractions and scattering. This leads to more dispersed energy distribution and weaker peak values of the perception channel information, thereby increasing the difficulty of extrapolation.

[0155] In summary, in Figure 8 (b) and Figure 9 (d), the extrapolated perception channel information is almost consistent with the measured perception channel information, proving that the model can effectively learn the correlation between the communication channel state information and the perception channel information, and use the known communication channel state information to extrapolate the unknown perception channel information.

[0156] In one of the embodiments, the preset condition includes that the loss value of the initial perception channel information model is not greater than a first threshold. Wherein, the loss value of the standard Transformer and the improved Transformer model is determined based on the following formula:

[0157]

[0158] Wherein, N represents the number of samples, P i represents the measured perception channel information value, represents the perception channel information value extrapolated using the communication channel state information.

[0159] Embodiment six

[0160] The processing flow of the perception channel information acquisition model training method provided by the embodiments of the present application, the method comprises:

[0161] The first communication device receives second indication information from the second communication device, the second indication information includes communication channel state information and perception channel information, and the second indication information is used to instruct the first communication device to train an initial perception channel information acquisition model according to the communication channel state information and the perception channel information.

[0162] The first communication device trains the initial perception channel information acquisition model according to the perception channel information and the communication channel state information, until the initial perception channel information acquisition model meets a preset condition, and obtains a perception channel information acquisition model.

[0163] In one embodiment, the chef perception channel information acquisition model comprises a self-encoder model. The model adopts a supervised learning manner, thereby enhancing the expression ability and generalization ability of the model.

[0164] The model comprises an encoder and a decoder. The training process of the model can be briefly described as follows: in the process of training the model, the low-dimensional code generated by the encoder is controlled so that the encoder can convert the perception channel information into predicted channel state information, and the decoder reconstructs the predicted channel state information output by the encoder to obtain predicted perception channel information.

[0165] In actual application, only the decoder is used to extrapolate the perception channel information from the communication channel state information. The measured communication channel state information is used as the input of the decoder, and the perception channel information output by the decoder is analyzed.

[0166] In one embodiment, the encoder is a PDP2CSI module, which can be represented as a nonlinear mapping. The mapping converts the perception channel information into a low-dimensional vector. Ideally, the low-dimensional representation generated by the encoder should be able to sufficiently capture the key features in the input signal (i.e., the perception channel information) and have a high similarity with the target communication channel state information.

[0167] In one embodiment, the decoder is a CSI2PDP module, which functions to recover the original perception channel information from the low-dimensional representation generated by the encoder and can also be regarded as a nonlinear mapping. Ideally, the perception channel information reconstructed by the decoder should be very close to the original input perception channel information. The architecture of the decoder (CSI2PDP module) is basically the same as that of the improved Transformer described in Embodiment 4.

[0168] In one embodiment, the preset condition comprises that a loss value of the initial perception channel information model is not greater than a first threshold. The loss value of the self-encoder model is calculated based on the following formula:

[0169]

[0170] wherein θ E and θ D are parameters of the encoder and the decoder, f D (f E (P i ; θ E ); θ D ) is the output of the decoder, and f E (P i ; θ E ) is the output of the encoder. The first half of the loss function represents the reconstruction error, and the second half represents the intermediate representation z iThe difference between the real communication channel state information C i .

[0171] In one embodiment, the second communication device can be a terminal device or a network device.

[0172] Based on the autoencoder model proposed in this embodiment, compared with the standard Transformer model and the improved Transformer model described in Embodiment Four, the input and output steps (input_step and output_step) N p of the autoencoder model are arbitrary values, i.e., in the inference stage, any continuous N p measurement point communication channel state information can be predicted to any continuous N p measurement point perception channel information. So that the autoencoder model can flexibly convert between different measurement points, and improve the generality and practicability of the model.

[0173] It should be understood that the communication channel state information described in any embodiment of the present application includes but is not limited to CSI (Channel State Information); the perception channel information includes but is not limited to PDP (Power-Delay-Profile), ADPP (Angle-Delay-Power-Profile), as long as it is any form or information that can represent the perception channel, it belongs to the category of the perception channel information described in the present application.

[0174] Embodiment Seven

[0175] Figure 10 A structure diagram of a perception channel information acquisition device provided by an embodiment of the present application is shown, which comprises:

[0176] A receiving module for receiving communication channel state information from a terminal device;

[0177] An acquisition module for acquiring the perception channel information according to the communication channel state information.

[0178] The perception channel information acquisition device described in this embodiment is arranged in a network device.

[0179] Embodiment Eight

[0180] Figure 11 A perception channel information acquisition device provided by an embodiment of the present application is shown, which comprises:

[0181] A first acquisition module for acquiring communication channel state information;

[0182] The second obtaining module is configured to obtain the sensing channel information based on the communication channel state information.

[0183] The sensing channel information obtaining device is arranged in the terminal device.

[0184] Embodiment Nine

[0185] Figure 12 A sensing channel information obtaining device is shown, which comprises:

[0186] The sending module is configured to send the sensing channel information obtained based on the communication channel state information to the network device.

[0187] The receiving module is configured to receive the sensing channel information sent by the network device.

[0188] The sensing channel information obtaining device is arranged in the terminal device.

[0189] Embodiment Ten

[0190] Figure 13 A sensing channel information obtaining model training device is shown, which comprises:

[0191] The receiving module is configured to receive first indication information from a second communication device, the first indication information comprising communication channel state information and sensing channel information, and the first indication information being used to instruct the first communication device to train an initial sensing channel information obtaining model based on the communication channel state information and the sensing channel information.

[0192] The training module is configured to train the initial sensing channel information obtaining model based on the sensing channel information and the communication channel state information until the preset sensing channel information model meets a preset condition, and obtain a trained sensing channel information obtaining model.

[0193] The sensing channel information obtaining model training device can be arranged in a terminal device or a network device. When the second communication device is a terminal device, the sensing channel information obtaining model training device is arranged in the terminal device; when the second communication device is a network device, the sensing channel information obtaining model training device is arranged in the network device.

[0194] Embodiment Eleven

[0195] Based on the same inventive concept, the embodiments of the present application also provide an electronic device, the method corresponding to the electronic device can be the method for sensing channel information acquisition and the method for training sensing channel information acquisition model in the foregoing embodiments, and the problem solving principle thereof is similar to the method. The electronic device provided by the embodiments of the present application comprises at least one processor, and a memory connected in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method and / or technical solution of the foregoing embodiments of the present application.

[0196] The electronic device can be a user device, or a device integrated by a user device and a network device through a network, or can also be an application program running on the above device, the user device includes but is not limited to computers, mobile phones, tablet computers, smart watches, wristbands and various terminal devices, and the network device includes but is not limited to network hosts, single network servers, multiple network server sets or computer sets based on cloud computing, which can be used to realize part of the processing function when setting an alarm. Here, the cloud is composed of a large number of hosts or network servers based on cloud computing, wherein cloud computing is a kind of distributed computing, which is composed of a virtual computer formed by a group of loosely coupled computer sets.

[0197] ​ The structure of an electronic device suitable for implementing the method and / or technical solution in the embodiments of the present application is shown, the device 1300 comprises a central processing unit (CPU) 1301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 1302 or the program loaded from the storage part 1308 to the random access memory (RAM) 1303. In the RAM 1303, various programs and data required for system operation are also stored. The CPU 1301, the ROM 1302 and the RAM 1303 are connected to each other through a bus 1304. An input / output (I / O) interface 1305 is also connected to the bus 1304.

[0198] The following components are connected to the I / O interface 1305: an input part 1306 including a keyboard, a mouse, a touch screen, a microphone, an infrared sensor, and the like; an output part 1307 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), an LED display, an OLED display, and the like, and a speaker, and the like; a storage part 1308 including one or more computer readable media such as a hard disk, an optical disk, a magnetic disk, a semiconductor memory, and the like; and a communication part 1309 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication part 1309 performs communication processing via a network such as the Internet.

[0199] In particular, the methods and / or embodiments in the present application can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program code for executing the methods shown in the flowcharts. When the computer program is executed by a central processing unit (CPU) 1301, the above-mentioned functions defined in the methods of the present application are executed.

[0200] Embodiment Eleven

[0201] Another embodiment of the present application also provides a computer readable storage medium having stored thereon computer program instructions, which can be executed by a processor to implement the method and / or technical solutions of any one or more embodiments of the present application.

[0202] In particular, the embodiments can employ any combination of one or more computer readable media. The computer readable media can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this document, the computer readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0203] A computer readable signal medium can include a propagated data signal with computer executable code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal can take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium can be any computer readable medium that can be involved in

[0204] The code can be transmitted in any coded or modular form, for example, over electrical, optical, or radio frequency signals, or any suitable combination thereof. The code can be transmitted using any media, including but not limited to wireless, wireline, optical, etc., or any suitable combination thereof.

[0205] The computer program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). The computer program code can also be embodied in a computer program product that can comprise a computer readable medium, which can be readable and / or

[0206] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0207] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the system, device and unit described above can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0208] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are merely schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or page components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0209] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0210] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of hardware plus software functional unit.

[0211] The integrated unit realized in the form of software functional unit can be stored in a computer readable storage medium. The software functional unit is stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0212] Finally, it should be noted that the above examples are only used to illustrate the technical solutions of the present application, and are not intended to limit the same; although the present application has been described in detail with reference to the foregoing examples, those of ordinary skill in the art should understand that the technical solutions recorded in the foregoing examples can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

[0213] Furthermore, the word "comprise" does not exclude other elements or steps, and the singular does not exclude the plural. Multiple units or devices recited in a device claim can also be implemented by one unit or device by means of software or hardware. The words first, second, etc. are used to indicate names and not to indicate any particular order.

Claims

1. A method for acquiring sensing channel information, characterized in that, The method is applied to a network device, and the method includes: Receive communication channel status information from the terminal device; Based on the communication channel status information, obtain the sensing channel information.

2. A method for acquiring sensing channel information, characterized in that, The method is applied to a terminal device, and the method includes: Obtain communication channel status information; Based on the communication channel status information, obtain the sensing channel information.

3. The method for acquiring sensing channel information according to claim 2, characterized in that, The method further includes: The terminal device sends the sensing channel information to the network device.

4. A method for acquiring sensing channel information, characterized in that, The method is applied to network devices and terminal devices, and the method includes: The terminal device sends communication channel status information to the network device; The network device acquires sensing channel information based on the communication channel status information.

5. A method for acquiring sensing channel information, characterized in that, The method is applied to network devices and terminal devices, and the method includes: The terminal device sends sensing channel information obtained based on the acquired communication channel status information to the network device; The network device receives sensing channel information sent from the terminal device.

6. The method for acquiring sensing channel information according to any one of claims 1 to 5, characterized in that, The method further includes: acquiring integrated communication sensing information; and acquiring the communication channel status information based on the integrated communication sensing information.

7. The method for acquiring sensing channel information according to claim 6, characterized in that, The communication channel state information includes: communication channel state information that has been obtained from the integrated communication sensing information parsing, and / or, communication channel state information that has not been obtained from the integrated communication sensing information parsing.

8. The method for acquiring sensing channel information according to claim 7, characterized in that, Based on the communication channel state information, the sensing channel information is obtained, including: The communication channel status information is sent to a preset sensing channel information acquisition model to obtain the sensing channel information.

9. The method for acquiring sensing channel information according to claim 8, characterized in that, Sending the communication channel status information to a preset sensing channel information acquisition model to obtain the sensing channel information includes: The communication channel status information is sent to a preset sensing channel information acquisition model; The preset sensing channel information acquisition model obtains the feature elements of the communication channel state information based on the communication channel state information; The sensing channel information is obtained based on the feature elements of the communication channel state information.

10. The method for acquiring sensing channel information according to claim 9, characterized in that, The characteristic elements include channel gain, phase information, power delay spread, received signal arrival time, and angle of arrival.

11. A method for training a sensing channel information acquisition model, characterized in that, For implementing the sensing channel information acquisition method as described in any one of claims 1 to 10, the method includes: The first communication device receives a second instruction information from the second communication device. The second instruction information includes communication channel status information and sensing channel information. The second instruction information is used to instruct the first communication device to train an initial sensing channel information acquisition model based on the communication channel status information and sensing channel information. The first communication device trains the initial sensing channel information acquisition model based on the sensing channel information and the communication channel state information until the initial sensing channel information model meets the preset conditions, thereby obtaining the sensing channel information acquisition model.

12. The method for training a sensing channel information acquisition model according to claim 11, characterized in that, The first communication device trains the initial sensing channel information acquisition model based on the sensing channel information and the communication channel state information, including: When the communication channel state information does not contain spatial location features, the first communication device sends a control signal to the initial sensing channel information acquisition model; The control signal includes a spatial location code, which is used to instruct the initial sensing channel information acquisition model to generate the spatial location features of the communication channel state information based on the spatial location code.

13. The method for training a sensing channel information acquisition model according to claim 12, characterized in that, Before the first communication device trains the initial sensing channel information acquisition model based on the sensing channel information and the communication channel state information, the method further includes: normalizing the sensing channel information and the communication channel state information.

14. The method for training a sensing channel information acquisition model according to claim 13, characterized in that, The preset conditions include: the loss value of the initial sensing channel information model is not greater than a first threshold.

15. A sensing channel information acquisition device, characterized in that, The device includes: The receiving module is used to receive communication channel status information from the terminal device; The acquisition module is used to acquire the sensing channel information based on the communication channel status information.

16. A sensing channel information acquisition device, characterized in that, The device includes: The first acquisition module is used to acquire communication channel status information; The second acquisition module is used to acquire sensing channel information based on the communication channel state information.

17. A sensing channel information acquisition device, characterized in that, The device includes: A transmitting module for transmitting communication channel status information to the network device; An acquisition module is used to acquire the sensing channel information based on the communication channel state information.

18. A sensing channel information acquisition device, characterized in that, The device includes: The transmitting module transmits sensing channel information obtained based on communication channel status information to the network device. A receiving module is used to receive sensing channel information sent from the terminal device.

19. A training device for a sensing channel information acquisition model, characterized in that, The device includes: A receiving module is configured to receive first indication information from a second communication device, the second indication information including communication channel status information and sensing channel information, the second indication information being configured to instruct the first communication device to train an initial sensing channel information acquisition model based on the communication channel status information and sensing channel information; The training module is used to train the initial sensing channel information acquisition model based on the sensing channel information and the communication channel state information until the preset sensing channel information model meets the preset conditions, thereby obtaining the trained sensing channel information acquisition model.

20. An electronic device, the electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 14.

21. A computer-readable medium having stored thereon computer program instructions that can be executed by a processor to implement the method as claimed in any one of claims 1 to 14.

22. A computer program product comprising a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 14.

Images