Model determination method and device, semantic quantification method and device, communication equipment and storage medium

Abstract

The invention relates to the technical field of communication, in particular to a model determination method and device, a semantic quantification method and device, communication equipment and a storage medium, and the model determination method comprises the steps: determining a training sample set; and training the first initial model and the second initial model based on the training sample set, taking the first initial model as a semantic quantification model when training is stopped, and taking the second initial model as a positioning model when training is stopped. According to the method and the device, the prediction semantic quantification data output by the semantic quantification model can be ensured, and the characteristics of the channel measurement data input into the semantic quantification model can be accurately reflected. On the basis, the accuracy of the positioning result of the channel measurement data predicted by the positioning model through the predicted semantic quantization data can be further ensured.

Description

Model determination, semantic quantization method and device, communication device and storage medium TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of communication, in particular to a model determination method, a semantic quantization method, a model determination device, a semantic quantization device, a communication device and a storage medium. BACKGROUND

[0002] To meet the needs of various commercial service scenarios and industrial Internet of Things scenarios for location services, high-precision positioning technology has become a popular research direction, which is helpful to realize indoor navigation, augmented reality, virtual reality, automatic driving and other business services. However, there are still some technical problems in the positioning technology that need to be solved.

[0003] SUMMARY

[0004] Embodiments of the present disclosure provide a model determination method, a semantic quantization method, a model determination device, a semantic quantization device, a communication device and a storage medium to solve the technical problems in the related art.

[0005] According to a first aspect of embodiments of the present disclosure, a model determination method is provided, which includes: determining a training sample set, wherein a sample in the training sample set includes channel measurement data corresponding to a positioning reference signal, and a label of the sample includes semantic quantization data; training a first initial model and a second initial model based on the training sample set, wherein the first initial model at the time of stopping training is used as a semantic quantization model, and the second initial model at the time of stopping training is used as a positioning model; wherein the input of the semantic quantization model includes channel measurement data, the output of the semantic quantization model includes predicted semantic quantization data corresponding to the channel measurement data, the input of the positioning model includes the predicted semantic quantization data, and the output of the positioning model includes a predicted positioning result corresponding to the channel measurement data.

[0006] According to a second aspect of embodiments of the present disclosure, a semantic quantization method is provided, which is performed by a first device, wherein the semantic quantization model is deployed in the first device, and the method includes: determining first channel measurement data by measuring a positioning reference signal; inputting the first channel measurement data into the semantic quantization model to determine first predicted semantic quantization data corresponding to the first channel measurement data.

[0007] According to a third aspect of the embodiments of the present disclosure, a model training method is provided, which is executed by a first training device, and includes: determining a training sample set, wherein a sample in the training sample set includes channel measurement data corresponding to a positioning reference signal, and a label of the sample includes a positioning result corresponding to the channel measurement data; training a first initial model based on the training sample set, and the first initial model at a training stop time is a semantic quantization model; wherein an input of the semantic quantization model includes channel measurement data, and an output of the semantic quantization model includes predicted semantic quantization data corresponding to the channel measurement data.

[0008] According to a fourth aspect of the embodiments of the present disclosure, a model training method is provided, which is executed by a second training device, and includes: receiving predicted semantic quantization data output by a first initial model trained by a first training device; and training a second initial model based on the predicted semantic quantization data; wherein an input of the second initial model includes the predicted semantic quantization data.

[0009] According to a fifth aspect of the embodiments of the present disclosure, a model determination apparatus is provided, which includes: a processing module configured to determine a training sample set, wherein a sample in the training sample set includes channel measurement data corresponding to a positioning reference signal, and a label of the sample includes a positioning result corresponding to the channel measurement data; train a first initial model and a second initial model based on the training sample set, and the first initial model at a training stop time is a semantic quantization model, and the second initial model at the training stop time is a positioning model; wherein an input of the semantic quantization model includes channel measurement data, an output of the semantic quantization model includes predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model includes the predicted semantic quantization data, and an output of the positioning model includes a predicted positioning result corresponding to the channel measurement data.

[0010] According to a sixth aspect of the embodiments of the present disclosure, a model training apparatus is provided, which includes: a processing module configured to determine a training sample set, wherein a sample in the training sample set includes channel measurement data corresponding to a positioning reference signal, and a label of the sample includes a positioning result corresponding to the channel measurement data; train a first initial model based on the training sample set, and the first initial model at a training stop time is a semantic quantization model; wherein an input of the semantic quantization model includes channel measurement data, and an output of the semantic quantization model includes predicted semantic quantization data corresponding to the channel measurement data.

[0011] According to a seventh aspect of the embodiments of the present disclosure, a model training apparatus is provided, comprising: a receiving module configured to receive predicted semantic quantization data output by a first initial model trained by a first training device; a processing module configured to train a second initial model based on the predicted semantic quantization data; wherein an input of the second initial model comprises the predicted semantic quantization data.

[0012] According to an eighth aspect of the embodiments of the present disclosure, a semantic quantization apparatus is provided, comprising: a processing module configured to determine first channel measurement data by measuring a positioning reference signal; and input the first channel measurement data into a semantic quantization model to determine first predicted semantic quantization data corresponding to the first channel measurement data.

[0013] According to a ninth aspect of the embodiments of the present disclosure, a communication device is provided, comprising: one or more processors; wherein the communication device is configured to perform the model training method of the first aspect, and / or the semantic quantization method of the second aspect.

[0014] According to a tenth aspect of the embodiments of the present disclosure, a storage medium is provided, the storage medium stores instructions, when the instructions are executed on a communication device, the communication device is caused to perform the model training method of the first aspect, and / or the semantic quantization method of the second aspect.

[0015] According to an eleventh aspect of the embodiments of the present disclosure, a program product is provided, when the program product is executed by a communication device, the communication device is caused to perform the model training method of the first aspect, and / or the semantic quantization method of the second aspect.

[0016] According to the embodiments of the present disclosure, the semantic quantization model can be obtained by training the first initial model based on a training sample set by using machine learning, deep learning, etc. Since the semantic quantization model is an AI model trained based on machine learning, deep learning, etc., when the semantic quantization model is used to quantize the channel measurement data, the channel measurement data is quantized based on the characteristics of the channel measurement data, rather than quantized according to fixed rules, so that the appropriate semantic quantization result can be predicted for the channel measurement data with different characteristics.

[0017] Accordingly, the predicted semantic quantization data output by the semantic quantization model can accurately reflect the characteristics of the channel measurement data input into the semantic quantization model, which can further ensure the accuracy of the positioning result of the channel measurement data predicted by the positioning model using the predicted semantic quantization data. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0019] FIG. 1A is a schematic diagram of an architecture of a communication system according to an embodiment of the present disclosure.

[0020] FIG. 1B is a schematic diagram of two positioning methods according to an embodiment of the present disclosure.

[0021] FIG. 1C is a schematic diagram of five application modes according to an embodiment of the present disclosure.

[0022] FIG. 2 is a schematic flowchart of a model determination method according to an embodiment of the present disclosure.

[0023] FIG. 3 is a schematic diagram of predicting a positioning result according to an embodiment of the present disclosure.

[0024] FIG. 4 is an interaction schematic diagram of semantic quantization according to an embodiment of the present disclosure.

[0025] FIG. 5 is a schematic block diagram of a model determination apparatus according to an embodiment of the present disclosure.

[0026] FIG. 6 is a schematic block diagram of a semantic quantization apparatus according to an embodiment of the present disclosure.

[0027] FIG. 7A is a schematic diagram of a structure of a communication device according to an embodiment of the present disclosure.

[0028] FIG. 7B is a schematic diagram of a structure of a chip according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0029] Embodiments of the present disclosure provide a model determination method and apparatus, a semantic quantization method and apparatus, a communication device and a storage medium.

[0030] In a first aspect, embodiments of the present disclosure provide a model determination method, which comprises: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises semantic quantization data; training a first initial model and a second initial model based on the training sample set, wherein the first initial model at a training stop time is a semantic quantization model, and the second initial model at the training stop time is a positioning model; wherein an input of the semantic quantization model comprises channel measurement data, an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

[0031] In the above embodiments, the semantic quantization model can be obtained by training the first initial model based on the training sample set by using machine learning, deep learning or the like. Since the semantic quantization model is an AI model trained based on machine learning, deep learning or the like, when performing semantic quantization on the channel measurement data, the semantic quantization model does not quantize the channel measurement data according to fixed rules, but can fully consider the characteristics of the channel measurement data and quantize the channel measurement data based on the characteristics of the channel measurement data, so that appropriate semantic quantization results can be predicted for channel measurement data with different characteristics.

[0032] Accordingly, the predicted semantic quantization data output by the semantic quantization model can accurately reflect the characteristics of the channel measurement data input into the semantic quantization model. On this basis, the accuracy of the predicted semantic quantization data used by the positioning model to predict the positioning result of the channel measurement data can be further ensured.

[0033] In combination with some embodiments of the first aspect. In some embodiments, in the process of training the first initial model and the second initial model based on the training sample set, the difference between the predicted positioning result and the actual positioning result is used as a loss function.

[0034] In combination with some embodiments of the first aspect. In some embodiments, the condition for stopping training comprises at least one of the following: the result of the loss function is within a threshold range; and the number of training cycles is greater than or equal to a number threshold.

[0035] In combination with some embodiments of the first aspect. In some embodiments, the semantic quantization model is deployed on a first device, and / or the positioning model is deployed on a second device; and the first device is configured to determine the channel measurement data by receiving a positioning reference signal, and send the predicted semantic quantization data to the second device.

[0036] In some embodiments of the first aspect. In some embodiments, the channel measurement data comprises at least one of: correlation information of the positioning reference signal; a channel impulse response corresponding to the positioning reference signal; a complex number corresponding to the channel impulse response; power delay profile data corresponding to the channel impulse response; delay profile data corresponding to the channel impulse response; spectrum information of a channel corresponding to the positioning reference signal; point cloud information of the channel corresponding to the positioning reference signal.

[0037] In some embodiments of the first aspect. In some embodiments, the correlation information of the positioning reference signal comprises at least one of: an amplitude of the positioning reference signal; a phase of the positioning reference signal; I-path information of the positioning reference signal; Q-path information of the positioning reference signal.

[0038] In some embodiments of the first aspect. In some embodiments, the spectrum information comprises at least one of: time delay spread spectrum information; Doppler spectrum information; micro-Doppler spectrum information; angle spectrum information; signal strength spectrum information.

[0039] In some embodiments of the first aspect. In some embodiments, the positioning reference signal comprises at least one of: a positioning reference signal sent by a network device; a sounding reference signal sent by a terminal.

[0040] In a second aspect, embodiments of the present disclosure provide a model training method, performed by a first training device, the method comprising: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; training a first initial model based on the training sample set, and the first initial model at a training stop time is taken as a semantic quantization model; wherein an input of the semantic quantization model comprises channel measurement data, and an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data.

[0041] In some embodiments of the second aspect. In some embodiments, the method further comprises: sending, to a second training device, predicted semantic quantization data output by the first initial model, for the second device to train a second initial model; wherein the second initial model at a training stop time is taken as a positioning model, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

[0042] In a third aspect, embodiments of the present disclosure provide a model training method, executed by a second training device, the method comprising: receiving predicted semantic quantization data output by a first initial model trained by a first training device; training a second initial model based on the predicted semantic quantization data; wherein an input of the second initial model comprises the predicted semantic quantization data.

[0043] In a fourth aspect, embodiments of the present disclosure provide a semantic quantization method, executed by a first device, wherein the semantic quantization model is deployed in the first device, the method comprising: determining first channel measurement data by measuring a positioning reference signal; inputting the first channel measurement data into the semantic quantization model to determine first predicted semantic quantization data corresponding to the first channel measurement data.

[0044] In some embodiments in combination with the fourth aspect. In some embodiments, the method further comprises: sending the first predicted semantic quantization data to a second device, wherein the positioning model is deployed in the second device, and the positioning model determines a first predicted positioning result corresponding to the first channel measurement data based on the first predicted semantic quantization data.

[0045] In a fifth aspect, embodiments of the present disclosure provide a model determination apparatus, the apparatus comprising: a processing module configured to determine a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; train a first initial model and a second initial model based on the training sample set, wherein the first initial model at a training stop time is a semantic quantization model, and the second initial model at the training stop time is a positioning model; wherein an input of the semantic quantization model comprises channel measurement data, an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

[0046] In a sixth aspect, embodiments of the present disclosure provide a model training apparatus, the apparatus comprising: a processing module configured to determine a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; train a first initial model based on the training sample set, wherein the first initial model at a training stop time is a semantic quantization model; wherein an input of the semantic quantization model comprises channel measurement data, and an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data.

[0047] In a seventh aspect, embodiments of the present disclosure provide a model training apparatus, the apparatus comprising: a receiving module configured to receive predicted semantic quantization data output by a first initial model trained by a first training device; a processing module configured to train a second initial model based on the predicted semantic quantization data; wherein an input of the second initial model comprises the predicted semantic quantization data.

[0048] In an eighth aspect, embodiments of the present disclosure provide a semantic quantization apparatus, the apparatus comprising: a processing module configured to determine first channel measurement data by measuring a positioning reference signal; input the first channel measurement data into a semantic quantization model to determine first predicted semantic quantization data corresponding to the first channel measurement data.

[0049] In a ninth aspect, embodiments of the present disclosure provide a communication device, comprising: one or more processors; wherein the communication device is configured to perform the model training method of the first aspect, any one of the optional embodiments of the first aspect, and / or the semantic quantization method of the second aspect, any one of the optional embodiments of the second aspect.

[0050] In a tenth aspect, embodiments of the present disclosure provide a storage medium, the storage medium storing instructions that, when executed on a communication device, cause the communication device to perform the model training method of the first aspect, any one of the optional embodiments of the first aspect, and / or the semantic quantization method of the second aspect, any one of the optional embodiments of the second aspect.

[0051] In an eleventh aspect, embodiments of the present disclosure provide a program product, the program product, when executed on a communication device, causes the communication device to perform the model training method of the first aspect, any one of the optional embodiments of the first aspect, and / or the semantic quantization method of the second aspect, any one of the optional embodiments of the second aspect.

[0052] In a twelfth aspect, embodiments of the present disclosure provide a computer program, the computer program, when executed on a computer, causes the computer to perform the model training method of the first aspect, any one of the optional embodiments of the first aspect, and / or the semantic quantization method of the second aspect, any one of the optional embodiments of the second aspect.

[0053] It can be understood that the above model determination apparatus, semantic quantization apparatus, communication device, communication system, storage medium, program product, and computer program are all used to execute the method proposed in the embodiments of the present disclosure. Therefore, the beneficial effects they can achieve can refer to the beneficial effects in the corresponding method, which will not be described here.

[0054] The model determination and semantic quantification method and device, the communication device and the storage medium are provided in the embodiments of the present disclosure. In some embodiments, the model determination and semantic quantification method can be replaced by the information processing method, the communication method and other terms, the model determination and semantic quantification device can be replaced by the information processing device, the communication device and other terms, and the information processing system and the communication system can be replaced by other terms.

[0055] 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, some or all steps of different embodiments can be combined arbitrarily, an embodiment can be combined with the optional implementation manners of other embodiments.

[0056] In the embodiments of the present disclosure, the terms and / or descriptions between the embodiments are consistent and can be referred to each other if there is no special description and logical conflict, and the technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0057] 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.

[0058] In the embodiments of the present disclosure, unless otherwise specified, the elements expressed in singular form, such as "one", "a", "the", "above", "said", "preceding", "this" and the like, can represent "one and only one", or "one or more", "at least one" and the like.

[0059] For example, in the case of using articles such as "a", "an", "the" and the like in translation, the noun after the article can be understood as singular expression, or can be understood as plural expression.

[0060] In the embodiments of the present disclosure, "a plurality of" means two or more.

[0061] In some embodiments, the terms "at least one of", "one or more", "a plurality of", "multiple" and the like can be replaced by each other.

[0062] In some embodiments, the description of "at least one of A, B", "A and / or B", "one of A or B", "at least one of A or B", "one of A or B" and the like, can include the following technical solutions according to the situation: in some embodiments, A is executed (A is executed regardless of B); in some embodiments, B is executed (B is executed regardless of A); in some embodiments, A and B are selectively executed (A and B are selected from A and B); in some embodiments, A and B are executed (A and B are executed). When there are more branches such as A, B, C, and the like, the above description is similar.

[0063] In some embodiments, the description of "A or B" and the like can include the following technical solutions according to the situation: in some embodiments, A is executed (A is executed regardless of B); in some embodiments, B is executed (B is executed regardless of A); in some embodiments, A and B are selectively executed (A and B are selected from A and B). When there are more branches such as A, B, C, and the like, the above description is similar.

[0064] The prefix words "first", "second" and the like in the embodiments of the present disclosure are only used to distinguish different description objects, and do not constitute a limitation on the position, order, priority, quantity or content of the description objects. The description of the description objects should refer to the description in the context of the claims or embodiments, and should not constitute an unnecessary limitation because of the use of the prefix words.

[0065] 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". "First" and "second" do not limit whether the "fields" they modify are in the same message, nor do they 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 number of description objects is not limited by ordinal words, and can be one or more. For example, "first device", where the number of "devices" can be one or more. In addition, 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 their types 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 their contents can be the same or different.

[0066] In some embodiments, "including A", "containing A", "for indicating A", "carrying A" can be interpreted as directly carrying A, or indirectly indicating A.

[0067] In some embodiments, the terms "in response to", "in response to determining", "in the case of", "when", "if", "if", etc. can be replaced with each other.

[0068] In some embodiments, the terms "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", etc. can be replaced with each other, and the terms "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", etc. can be replaced with each other.

[0069] In some embodiments, the device, etc. can be interpreted as an entity, and can also be interpreted as virtual, and the name is not limited to the name described in the embodiments. The terms "device", "equipment", "device", "circuit", "network element", "node", "function", "unit", "section", "system", "network", "chip", "chip system", "entity", "subject", etc. can be replaced with each other.

[0070] In some embodiments, "network" can be interpreted as a device (e.g., access network device, core network device, etc.) contained in the network.

[0071] In some embodiments, the terms “access network device (AN device),” “radio access network device (RAN device),” “base station (BS),” “radio base station,” “fixed station,” “node,” “access point,” “transmission point (TP),” “reception point (RP),” “transmission / reception point (TRP),” “panel,” “antenna panel,” “antenna array,” “cell,” “macro cell,” “small cell,” “femto cell,” “pico cell,” “sector,” “cell group,” “serving cell,” “carrier,” “component carrier,” “bandwidth part (BWP),” and the like can be used interchangeably.

[0072] In some embodiments, the terms "terminal," "terminal device," "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 so on can be replaced with each other.

[0073] In some embodiments, the access network device, the core network device, or the network device can be replaced with a terminal. For example, the embodiments of the present disclosure can also be applied to a structure in which communication between the access network device, the core network device, or the network device and the terminal is replaced with communication between a plurality of terminals (e.g., device-to-device (D2D), vehicle-to-everything (V2X), etc.). In this case, the terminal can also be configured to have all or part of the functions of the access network device. In addition, the terms "uplink," "downlink," and the like can also be replaced with terms corresponding to the inter-terminal communication (e.g., "side"). For example, the uplink channel, the downlink channel, and the like can be replaced with the side channel, and the uplink, the downlink, and the like can be replaced with the sidelink.

[0074] In some embodiments, the terminal can be replaced with the access network device, the core network device, or the network device. In this case, the access network device, the core network device, or the network device can also be configured to have all or part of the functions of the terminal.

[0075] In some embodiments, obtaining data, information, etc. can comply with laws and regulations of the country where the location is.

[0076] In some embodiments, data, information, etc. can be obtained after obtaining user consent.

[0077] 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.

[0078] FIG. 1A is a schematic diagram of an architecture of a communication system according to an embodiment of the present disclosure.

[0079] As shown in FIG. 1A, the communication system 100 includes a first device 101 and a second device 102, for example, the first device can be a reference signal receiving device, and the second device can be a model inference device.

[0080] For example, the types of the first device and the second device include, but are not limited to, a terminal, a network device.

[0081] In some embodiments, the network device includes at least one of the following: an access network device, a core network device.

[0082] In some embodiments, the terminal 101 includes at least one of the following, for example: a mobile phone, a wearable device, an Internet of Things device, a communication-enabled car, a smart car, a tablet computer (Pad), a computer with wireless transceiver function, 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 smart grid, a wireless terminal device in transportation safety, a wireless terminal device in smart city, a wireless terminal device in smart home, but is not limited thereto.

[0083] In some embodiments, the access network device is, for example, a node or device that accesses a terminal to a wireless network, and the access network device can include at least one of an evolved NodeB (eNB) in a 5G communication system, a next generation eNB (ng-eNB), a next generation NodeB (gNB), a node B (NB), a home node B (HNB), a home evolved node B (HeNB), a wireless backhaul device, a radio network controller (RNC), a base station controller (BSC), a base transceiver station (BTS), a base band unit (BBU), a mobile switching center, a base station in a 6G communication system, an Open RAN, a Cloud RAN, a base station in other communication systems, an access node in a Wi-Fi system, but is not limited thereto.

[0084] In some embodiments, the core network device can be one device including one or more network elements, or can be multiple devices or device groups including all or part of the one or more network elements described above. The network element can be virtual or physical. The core network includes, for example, at least one of an evolved packet core (EPC), a 5G core network (5GCN), and a next generation core (NGC).

[0085] In some embodiments, the technical solutions of the present disclosure can be applied to an Open RAN architecture, at which time the interfaces between or within the access network devices involved in the embodiments of the present disclosure can become internal interfaces of the Open RAN, and the processes and information interactions between these internal interfaces can be realized through software or programs.

[0086] In some embodiments, the access network device can be composed of a central unit (CU) and a distributed unit (DU), where the CU can also be referred to as a control unit. The CU-DU structure can split the protocol layers of the access network device, and some of the protocol layers are controlled by the CU, and the remaining or all of the protocol layers are distributed in the DU and controlled by the CU, but is not limited thereto.

[0087] It can be understood that the communication system described in the embodiments of the present disclosure is for more clearly illustrating the technical solutions of the embodiments of the present disclosure, and does not constitute a limitation on the technical solutions proposed by the embodiments of the present disclosure. Those skilled in the art can know that, with the evolution of system architecture and the appearance of new business scenarios, the technical solutions proposed by the embodiments of the present disclosure are also applicable to similar technical problems.

[0088] 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 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.

[0089] 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), 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 thereon, 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).

[0090] In some embodiments, in order to meet the demand of various commercial service scenarios and industrial Internet of Things scenarios for location services, indoor navigation, augmented and virtual reality, automatic driving, and other service services can be implemented based on high-precision positioning technology.

[0091] In some embodiments, high-precision positioning in indoor and outdoor scenarios can be achieved based on time measurement and various positioning methods based on angles, etc. With the increasing demand for related business services, positioning accuracy enhancement for commercial scenarios and Industrial Internet of Things (IIOT) scenarios needs to be considered, aiming to achieve decimeter-level high-precision positioning to meet the high-precision location service needs of consumer and enterprise markets.

[0092] However, the positioning algorithms in the related art, such as TDOA (Time Difference of Arrival), Multi-RTT (Round Trip Time), etc., are difficult to meet the very strict positioning accuracy requirements in IIoT scenarios, etc., thereby affecting related business services.

[0093] In some embodiments, to address the above problems, an AI solution can be introduced. Deep neural network models can better complete complex data processing and feature modeling processes and have been applied to solve many key problems in wireless communication. High-precision positioning based on AI is an important application case of artificial intelligence technology in the communication field, and the related technology can model the mapping relationship between channel measurement data and terminal position coordinates using a deep neural network model, which can achieve higher positioning accuracy compared to traditional positioning methods.

[0094] In some embodiments, an AI model can be used to implement a semantic encoder at the sending end and a semantic decoder at the receiving end in a semantic communication system. The neural network model can autonomously learn key semantic feature information in the data based on large-scale training data, thereby completing data compression and data reconstruction and achieving low-overhead and effective data transmission.

[0095] In some embodiments, AI-based high-precision positioning technology research mainly considers two specific implementation methods, including direct positioning based on an AI model and indirect positioning assisted by an AI model.

[0096] FIG. 1B is a schematic diagram of two positioning methods according to an embodiment of the present disclosure.

[0097] As shown in FIG. 1B, positioning method (1) is direct positioning based on an AI model: the AI positioning model input is channel measurement data used for positioning, and the output is the positioning result, i.e., the target position coordinates;

[0098] The positioning method (2) is AI model assisted indirect positioning: the input of the AI model is channel measurement data for positioning, and the output is intermediate positioning parameters, which can include ToA (Time-of-Arrival), AoA (Angle-of-Arrival), etc., and the target position coordinates, i.e., the positioning result, can be calculated based on these intermediate parameters using traditional positioning methods such as TDOA.

[0099] In the above direct positioning method and indirect positioning method, the input of the AI model is channel measurement data for positioning, which is usually the channel impulse response (CIR) calculated based on the positioning reference signal. For example, the terminal can calculate the downlink CIR based on the positioning reference signal (PRS) sent by the network device (e.g., base station); for example, the network device can calculate the uplink CIR based on the channel sounding reference signal for positioning (SRS-Pos) sent by the terminal.

[0100] In addition, based on the specific device implementing positioning, AI-based high-precision positioning can be implemented at the terminal, base station, or location management function (LMF). For example, the AI-based high-precision positioning scheme includes five specific application modes.

[0101] FIG. 1C is a schematic diagram of the five application modes according to an embodiment of the present disclosure. As shown in FIG. 1C:

[0102] Mode 1: Terminal-side direct or indirect positioning: the terminal calculates channel measurement data such as CIR for positioning based on the PRS sent by the base station, inputs the AI model to directly obtain the terminal position coordinates, or the terminal inputs the measurement data for positioning into the AI model to obtain intermediate positioning parameters such as ToA and AoA, and then uses traditional positioning methods such as TDOA to obtain the terminal position coordinates. After obtaining the positioning result, i.e., the position coordinates, the terminal reports the result to the LMF.

[0103] Mode 2: Terminal-assisted LMF-side indirect positioning: the terminal calculates channel measurement data such as CIR for positioning based on the PRS sent by the base station, inputs the AI model to obtain intermediate positioning parameters such as ToA and AoA, and reports the intermediate positioning parameters to the LMF, and the LMF uses traditional positioning methods such as TDOA to obtain the terminal position coordinates.

[0104] Mode 3: Terminal-assisted LMF-side direct positioning: The terminal calculates channel measurement data such as CIR for positioning according to PRS sent by the base station and reports the measurement data to the LMF, and the LMF inputs the received measurement data into an AI model to obtain target position coordinates.

[0105] Mode 4: Base station-assisted LMF-side indirect positioning: The base station calculates channel measurement data such as CIR for positioning according to SRS-Pos sent by the terminal, inputs the measurement data into an AI model to obtain intermediate positioning parameters such as ToA and AoA, and reports the intermediate positioning parameters to the LMF, and the LMF uses a traditional positioning method such as TDOA to obtain terminal position coordinates.

[0106] Mode 5: Base station-assisted LMF-side direct positioning: The base station calculates channel measurement data such as CIR for positioning according to SRS-Pos sent by the terminal, and reports the measurement data for positioning to the LMF, and the LMF inputs the measurement data into an AI model to directly obtain terminal position coordinates.

[0107] In the above five modes, data acquisition and model inference can be implemented in different devices. For example, for modes 1, 2, and 4, the channel measurement data required by the AI positioning model is measured by the terminal or the base station according to the positioning reference signal, and the AI model is also deployed on the terminal or the base station, so the terminal or the base station can directly input the channel measurement data obtained by it into the AI positioning model to obtain the positioning result, and there is no need to report and transmit the channel measurement data for positioning. For modes 3 and 5, the channel measurement data for AI positioning needs to be first calculated by the terminal or the base station according to the positioning reference signal, and then reported to the LMF, input into the AI positioning model, and the positioning result, i.e., the target position coordinates, is obtained.

[0108] In some embodiments, the main application process of the deep learning-based high-precision positioning scheme is as follows: after the AI network model for target positioning is trained using training data, the trained AI positioning model is deployed on the terminal, the base station, or the LMF, and then the target positioning work in the actual system is completed, i.e., model inference is performed.

[0109] For modes 3 and 5 of AI-based high-precision positioning, the channel measurement data for positioning needs to be first reported by the terminal or the base station to the LMF, and then input into the AI model to complete model inference to obtain positioning accuracy. Although the AI positioning model can achieve high positioning accuracy, if the data reporting overhead for the AI model input is too large and occupies too many communication transmission resources, it will have a certain impact on the system work, which is still not conducive to the application of AI positioning technology. Therefore, the trade-off between positioning data reporting overhead and positioning accuracy needs to be considered.

[0110] For example, in the AI-based high-precision positioning standard research and simulation evaluation, the channel impulse response (CIR) calculated and obtained by the receiving end based on the positioning reference signal can be used as the input of the AI model.

[0111] The CIR can represent the influence of the signal during channel propagation and reflect the changes of the signal after channel propagation, including the energy attenuation of the impulse signal caused by path loss and shadow fading, and the superposition of multiple different impulse signals received at the receiving end due to multiple propagation paths in the channel.

[0112] The CIR data is a complex matrix, the dimensions of which are mainly related to the number of transmission nodes (such as transmission reception points (TRPs)) of the reference signal and the number of time domain sampling points. The CIR data of each TRP is a complex number at multiple time domain sampling points, containing power, phase and time delay information.

[0113] Although directly using complete CIR data as the input data of the AI model can achieve high positioning accuracy, the data reporting overhead is large. In addition, in the standard simulation evaluation, the CIR is processed to retain only part of the information to obtain power delay profile (PDP) and delay profile (DP) data, which can also be used as the input of the AI positioning model. The PDP retains the power and delay information of the CIR data, and the DP only retains the delay information of the CIR data.

[0114] In some embodiments, considering the problem of excessive data reporting overhead, in the application process, the channel measurement data for AI positioning can be transmitted after quantization, that is, the accurate original data is mapped to a limited set of discrete numerical values to simplify the data representation, and then a certain length of bit data stream is used to represent the original data to complete the data transmission.

[0115] The use of data quantization in a wireless communication system can achieve one of the following technical effects:

[0116] (1) Reduce the size and processing complexity of the data, and greatly improve the transmission efficiency;

[0117] (2) The storage space required by the quantized data is less, which is helpful to save more data on the storage device;

[0118] (3) In a low-bandwidth or high-latency network environment, since quantization can reduce the amount of data, it is beneficial to reduce the transmission delay;

[0119] (4) The quantized data representation is relatively simple, which can reduce the computational complexity;

[0120] (5) Quantization can increase the ambiguity of data, thereby improving the privacy and security of data.

[0121] In some embodiments, quantization processing of CIR and other channel measurement data types mainly includes uniform quantization, non-uniform quantization and other methods, which convert high-precision floating-point numbers and other types of data in the original data into low-precision data represented by bit stream data according to relatively fixed rules. However, in some cases, these fixed rule quantization methods do not conform to the characteristics of the actual data to be quantized, resulting in the loss of some key data feature information, which in turn adversely affects the positioning accuracy of the AI positioning model.

[0122] For example, in the channel measurement data to be quantized, the data at a certain position is 0.501, and the data type is floating-point number, which needs to be accurately represented by 32 bits (bits). Quantizing the data 0.501 can convert the data into a value represented by relatively fewer bits (e.g., 4 bits, 8 bits, etc.), such as 0.5.

[0123] Since uniform quantization, non-uniform quantization and other methods are used to quantize data according to fixed rules, but the characteristics of different data are different.

[0124] Taking uniform quantization as an example, for example, for data A, the importance of each bit value is the same, such as quantization to a value represented by n1 bits, so that each bit value before quantization corresponds to the same number of bits in n1 bits. Each bit value before quantization can be represented with the same precision, and the quantization result can basically meet the characteristics of data A.

[0125] However, for data B, the importance of a bit (or multiple bits) is different from that of other bits, so for the value of the bit with relatively high importance, it needs to be represented with relatively high precision. Therefore, when quantized to a value represented by n2 (which can be equal to or different from n1) bits, the value of the bit with relatively high importance needs to correspond to relatively more bits relative to other bits.

[0126] Since uniform quantization quantizes data according to fixed rules, quantizing data B based on this quantization method will result in each bit value before quantization corresponding to the same number of bits in n2 bits, thereby failing to meet the need for the value of the bit with relatively high importance to be represented by relatively more bits. The quantization result obtained on this basis will adversely affect the positioning accuracy of the subsequent AI positioning model.

[0127] For example, for data C, the quantized value is represented by n3 bits, and the value of the i th bit in data C corresponds to j bits in n3 bits. For data C, the value of each bit is the same, and it is assumed that the quantization result can basically meet the characteristics of data C.

[0128] However, for data D, the value of the i th bit is less important than the values of other bits, and the value of the i th bit with relatively low importance needs to be represented with relatively low precision. Therefore, when quantized to a value represented by n4 (which can be equal to or different from n3) bits, the value of the i th bit with relatively low importance needs to correspond to relatively few bits compared to other bits.

[0129] Since the non-uniform quantization quantizes data with a fixed rule, quantizing data D based on this quantization method will result in the value of the i th bit before quantization still corresponding to j bits in n4 bits, thereby failing to meet the need to represent the value of the bit with relatively low importance with relatively few bits. The quantization result obtained on this basis will adversely affect the positioning accuracy of the subsequent AI positioning model.

[0130] In a first aspect, embodiments of the present disclosure provide a model determination method. FIG. 2 is a schematic flowchart of a model determination method according to an embodiment of the present disclosure.

[0131] The model determination method shown in this embodiment can be executed by a training device. For example, the type of the training device can be a communication device, for example, the training device can be a first device, or a second device, or a third device other than the first device and the second device.

[0132] For example, the first device, the second device, and the third device can be a terminal, a network device, a server, and the like, and the present disclosure does not limit the same.

[0133] As shown in FIG. 2, the model determination method can include the following steps:

[0134] In step S201, a training sample set is determined.

[0135] In some embodiments, the samples in the training sample set include channel measurement data corresponding to a positioning reference signal, and the labels of the samples include semantic quantization data.

[0136] In step S202, the first initial model and the second initial model are trained based on the training sample set. The first initial model at the time of stopping training is used as a semantic quantization model, and the second initial model at the time of stopping training is used as a positioning model.

[0137] In some embodiments, the input of the semantic quantization model comprises channel measurement data, the output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, the input of the positioning model comprises the predicted semantic quantization data, and the output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

[0138] It should be noted that the embodiment shown in FIG. 2 can be independently implemented, or can be implemented in combination with at least one other embodiment of the present disclosure. The specific implementation can be selected as needed, and the present disclosure does not limit.

[0139] In some embodiments, a training sample set can be constructed first. The samples in the training sample set can comprise channel measurement data, and the labels of the samples can comprise semantic quantization results corresponding to the channel measurement data, which can be referred to as semantic quantization data.

[0140] In some embodiments, an initial model for training the semantic quantization model, which can be referred to as a first initial model, and an initial model for training the positioning model, which can be referred to as a second initial model, can be determined. Then, the first initial model and the second initial model can be trained based on the training sample set. For example, the first initial model and / or the second initial model can be a neural network model. For example, the training process can be implemented based on supervised learning, or can be implemented based on unsupervised learning, and the present disclosure does not limit this.

[0141] The first initial model at the time when the training is stopped can be used as the semantic quantization model, and the second initial model at the time when the training is stopped can be used as the positioning model. The semantic quantization model thus obtained can perform feature extraction, data compression, etc. on the input channel measurement data, and output predicted semantic quantization data obtained by predicting the channel measurement data. For example, the predicted semantic quantization data can contain key semantic features in the channel measurement data, and the predicted semantic quantization data can be represented by a bit stream, for example.

[0142] In some embodiments, the predicted semantic quantization data can be used to predict a positioning result corresponding to the channel measurement data. For example, the predicted semantic quantization data can be input into the positioning model, and the positioning model can predict a positioning result corresponding to the channel measurement data based on the predicted semantic quantization data, which can be referred to as a predicted positioning result.

[0143] According to an embodiment of the present disclosure, the first initial model can be trained to obtain the semantic quantization model based on a training sample set by using an algorithm such as machine learning or deep learning. Since the semantic quantization model is an AI model trained based on an algorithm such as machine learning or deep learning, when performing semantic quantization on the channel measurement data, the semantic quantization model will not quantize the channel measurement data according to a fixed rule, but will fully consider the characteristics of the channel measurement data and quantize the channel measurement data based on the characteristics of the channel measurement data, so that appropriate semantic quantization results can be predicted for channel measurement data with different characteristics.

[0144] Therefore, the predicted semantic quantization data output by the semantic quantization model can accurately reflect the characteristics of the channel measurement data input into the semantic quantization model. On this basis, the accuracy of the predicted semantic quantization data used by the positioning model to predict the positioning result of the channel measurement data can be further ensured.

[0145] For example, the semantic quantization model determined based on an embodiment of the present disclosure quantizes the data B, and in the case that the output predicted semantic quantization data is n2 bits, the value of the bit with relatively high importance in the channel measurement data can be ensured to correspond to relatively more bits than the values of other bits, thereby meeting the need for the value of the bit with relatively high importance to be represented by relatively more bits.

[0146] It should be noted that the semantic quantization model and the positioning model obtained by training can also be verified by constructing a verification set and tested by constructing a test set. The specific selection can be made according to the need, and the present disclosure will not be repeated here.

[0147] In some embodiments, during the training of the first initial model and the second initial model based on the training sample set, the difference between the predicted positioning result and the actual positioning result is used as a loss function.

[0148] In some embodiments, the training stopping condition includes at least one of the following:

[0149] The result of the loss function is within a threshold range;

[0150] The number of training cycles is greater than or equal to a number threshold.

[0151] In some embodiments, a loss function can be constructed, for example, the loss function is the difference between the predicted positioning result and the actual positioning result. The actual positioning result may, for example, be the label value of the sample in the sample set, or a value set artificially.

[0152] For example, the condition for stopping training can include that the result of the loss function is within a threshold range. For example, the loss function can be processed by a stochastic gradient descent algorithm, which can calculate the gradient of the loss function and update the model parameters in the direction of the negative gradient to minimize the loss function. As the training of each cycle, the loss function as a whole tends to gradually decrease, and it can be determined whether the loss function in each training cycle is within the threshold range. If it is not within the threshold range, the next training cycle is continued until the loss function is within the threshold range (for example, the loss function is within the threshold range once or multiple times). In this case, the model prediction result is close to the label value of the sample, which can meet the requirement of prediction accuracy, and therefore the training can be stopped.

[0153] For example, the condition for stopping training can include that the number of training cycles is greater than or equal to a number threshold. For example, one training cycle can be from the sample input to the first initial model, to the adjustment of the parameter model position of the first initial model and the second initial model. The number of training cycles can be recorded, and when the number of training cycles is greater than or equal to the number threshold, in this case, the model training process has been relatively long in time, and therefore the training can be stopped to avoid excessive overhead.

[0154] It should be noted that the condition for stopping training is not limited to the two conditions shown in the above embodiments, and other conditions can also be included, which are not limited by the present disclosure. For example, a positioning accuracy condition can be included, where the positioning accuracy can be represented as the difference between the predicted positioning result and the target label result, where the target label result can not be associated with the sample, for example, a value set by a human.

[0155] In some embodiments, the channel measurement data includes at least one of:

[0156] related information of the positioning reference signal;

[0157] a channel impulse response corresponding to the positioning reference signal;

[0158] a complex corresponding to the channel impulse response;

[0159] power delay profile data corresponding to the channel impulse response;

[0160] delay profile data corresponding to the channel impulse response;

[0161] spectrum information of a channel corresponding to the positioning reference signal;

[0162] point cloud information of a channel corresponding to the positioning reference signal.

[0163] For example, the point cloud information can be obtained by statistically analyzing the propagation paths and other information contained in the positioning reference signal.

[0164] In some embodiments, the spectrum information comprises at least one of: time delay spread spectrum information; Doppler spectrum information; micro-Doppler spectrum information; angle spectrum information; signal strength spectrum information.

[0165] For example, the spectrum information can be estimated based on data in each dimension of the channel state information matrix by a perception algorithm (which can be selected as needed, and the present disclosure does not limit it). For example, the spectrum information contains information of multiple paths or multiple motion modes, and each path or each motion mode can be reflected by an independent spectrum line or parameter.

[0166] In some embodiments, the related information of the positioning reference signal comprises at least one of:

[0167] an amplitude of the positioning reference signal;

[0168] a phase of the positioning reference signal;

[0169] I path information of the positioning reference signal;

[0170] Q path information of the positioning reference signal.

[0171] In some embodiments, the positioning reference signal comprises at least one of:

[0172] a positioning reference signal sent by a network device;

[0173] a sounding reference signal sent by a terminal.

[0174] For example, when positioning is based on the aforementioned mode 3, the positioning reference signal can be a positioning reference signal sent by a network device, such as a PSR. For example, when positioning is based on the aforementioned mode 5, the positioning reference signal can be a sounding reference signal sent by a terminal, such as a SRS-Pos.

[0175] It should be noted that the mode to which the embodiments of the present disclosure are applicable is not limited to mode 3 and mode 5, and can also be applicable to other modes.

[0176] In some embodiments, the semantic quantization model is deployed on a first device, and / or the positioning model is deployed on a second device; the first device is configured to determine channel measurement data by receiving a positioning reference signal, and send predicted semantic quantization data to the second device.

[0177] It should be noted that the first device and the second device can be different devices, or can be the same device, and in the case where the first device and the second device are the same device, the first device can not necessarily perform the operation of sending predicted semantic quantization data to the second device.

[0178] For example, in the mode 1 based positioning, the first device and the second device can be the same device, i.e., a terminal, and the predicted positioning result comprises a terminal position.

[0179] For example, in the mode 2 based positioning, the first device and the second device can be the same device, i.e., a terminal, and the predicted positioning result comprises a positioning intermediate parameter.

[0180] For example, in the mode 3 based positioning, the first device can be a terminal, and the second device can be an LMF, and the predicted positioning result comprises a terminal position.

[0181] For example, in the mode 4 based positioning, the first device and the second device can be the same device, i.e., a base station, and the predicted positioning result comprises a positioning intermediate parameter.

[0182] For example, in the mode 5 based positioning, the first device can be a base station, and the second device can be an LMF, and the predicted positioning result comprises a terminal position.

[0183] It should be noted that the semantic quantization model and the positioning model can be trained by a communication device, or can be trained by different devices, and the present disclosure does not limit this. The following describes the case where the semantic quantization model and the positioning model are trained by different devices through several embodiments.

[0184] In a second aspect, embodiments of the present disclosure propose a model training method, executed by a first training device, comprising: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; training a first initial model based on the training sample set, and the first initial model at the time of training stop is taken as a semantic quantization model; wherein an input of the semantic quantization model comprises channel measurement data, and an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data.

[0185] In some embodiments, a training sample set can be constructed first, a sample in the training sample set can comprise channel measurement data, and a label of the sample comprises a semantic quantization result corresponding to the channel measurement data, which can be referred to as semantic quantization data.

[0186] In some embodiments, an initial model for training the semantic quantization model can be determined, which can be referred to as a first initial model. Then, the first initial model can be trained based on the training sample set. For example, the first initial model can be a neural network model. For example, the training process can be implemented based on supervised learning, or can be implemented based on unsupervised learning, and the present disclosure does not limit this.

[0187] The first initial model is trained based on the training sample set, and the first initial model at the time of stopping training can be used as a semantic quantization model. The semantic quantization model obtained in this way can perform feature extraction, data compression, etc. on the input channel measurement data, and output semantic quantization data predicted from the channel measurement data, which can be referred to as predicted semantic quantization data. For example, the predicted semantic quantization data can contain key semantic features in the channel measurement data, and the predicted semantic quantization data can be represented by a bit stream.

[0188] In some embodiments, the predicted semantic quantization data can be used to predict the positioning result corresponding to the channel measurement data. For example, the predicted semantic quantization data can be input into a positioning model trained by a second training device, and the positioning model can predict the positioning result corresponding to the channel measurement data based on the predicted semantic quantization data, which can be referred to as a predicted positioning result.

[0189] In some embodiments, the model training method further comprises: sending the predicted semantic quantization data output by the first initial model to a second training device, for the second device to train a second initial model; wherein the second initial model at the time of stopping training is used as a positioning model, the input of the positioning model includes the predicted semantic quantization data, and the output of the positioning model includes the predicted positioning result corresponding to the channel measurement data.

[0190] In some embodiments, the first training device can interact with the second training device during the training of the first initial model, so as to collaboratively train the model with the second training device.

[0191] For example, the first training device can send the predicted semantic quantization data output by the first initial model to the second training device during the training process, and the second training device can train the second initial model based on the predicted semantic quantization data, for example, input the predicted semantic quantization data into the second initial model, and the second initial model can output the predicted positioning result of the channel measurement data corresponding to the predicted semantic quantization data.

[0192] For example, the first training device can send the label value corresponding to the sample input into the first initial model to the second training device during the training process, and the second training device can determine the loss function according to the predicted positioning result output by the second initial model and the label value, and then determine whether to stop the training process based on the loss function.

[0193] For example, the second training device can feed back the predicted positioning result output by the second initial model to the first training device during the training process, and the first training device can determine the loss function according to the predicted positioning result fed back by the second initial model and the label value, and then determine whether to stop the training process based on the loss function.

[0194] In a third aspect, embodiments of the present disclosure provide a model training method, executed by a second training device, the method comprising: receiving predicted semantic quantization data output by a first initial model trained by a first training device; training a second initial model based on the predicted semantic quantization data; wherein an input of the second initial model comprises the predicted semantic quantization data.

[0195] In some embodiments, the training of the second initial model, the second initial model at the time when the training stops can be used as a positioning model, and an output of the positioning model can comprise a predicted positioning result corresponding to the channel measurement data.

[0196] In some embodiments, the second training device can interact with the first training device during the training of the second initial model, so as to collaboratively train the model with the first training device.

[0197] For example, the second training device can receive, during the training, predicted semantic quantization data output by the first initial model in the first training device, the second training device can train the second initial model based on the predicted semantic quantization data, for example, input the predicted semantic quantization data into the second initial model, and the second initial model can output a predicted positioning result of channel measurement data corresponding to the predicted semantic quantization data.

[0198] For example, the second training device can receive, during the training, a label value corresponding to a sample input into the first initial model sent by the first training device, the second training device can determine a loss function according to the predicted positioning result output by the second initial model and the label value, and then determine whether to stop the training process based on the loss function.

[0199] For example, the second training device can feed back the predicted positioning result output by the second initial model to the first training device during the training, and the first training device can determine a loss function according to the predicted positioning result fed back by the second initial model and the label value, and then determine whether to stop the training process based on the loss function.

[0200] It should be noted that the concepts (such as training stop, channel measurement data, etc.) involved in the embodiments of the third aspect can refer to the related content in the embodiments of the second aspect described above, which will not be repeated here.

[0201] In a fourth aspect, embodiments of the present disclosure provide a semantic quantization method. For example, the semantic quantization method is executed by a first device, and a semantic quantization model is deployed in the first device. The method comprises: determining first channel measurement data by measuring a positioning reference signal; and inputting the first channel measurement data into the semantic quantization model based on the first aspect and any one of the optional embodiments of the first aspect, to determine first predicted semantic quantization data corresponding to the first channel measurement data.

[0202] In some embodiments, the semantic quantization model is determined based on the following manner: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; training a first initial model and a second initial model based on the training sample set, wherein the first initial model at a training stop time is the semantic quantization model, and the second initial model at the training stop time is the positioning model; wherein an input of the semantic quantization model comprises channel measurement data, an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

[0203] In some embodiments, the semantic quantization model is determined based on the following manner: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; training a first initial model based on the training sample set, wherein the first initial model at a training stop time is the semantic quantization model; wherein an input of the semantic quantization model comprises channel measurement data, and an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data.

[0204] In some embodiments, the semantic quantization method further comprises: sending the first predicted semantic quantization data to a second device, wherein the second device is deployed with a positioning model, and the positioning model determines a first predicted positioning result corresponding to the first channel measurement data based on the first predicted semantic quantization data.

[0205] It should be noted that the concepts (e.g., training stop, channel measurement data, etc.) involved in the embodiments of the fourth aspect can refer to the related content in the embodiments of the second aspect described above, which will not be repeated here.

[0206] FIG. 3 is a schematic diagram of predicting a positioning result, according to an embodiment of the present disclosure.

[0207] As shown in FIG. 3, based on the training of the semantic quantization model and the positioning model, the semantic quantization model can be deployed in the first device, and the semantic quantization model can be deployed in the second device.

[0208] For example, the first device can be referred to as a channel data measurement device, which is used to obtain channel measurement data.

[0209] For example, the second device can be referred to as a model inference device, which is used to predict the channel measurement data to obtain a predicted positioning result.

[0210] In some embodiments, the semantic quantization model deployed in the channel data measurement device can perform feature extraction and compression on the channel measurement data input into the semantic quantization model, and predict the semantic quantization result corresponding to the channel measurement data, which can be referred to as predicted semantic quantization data.

[0211] In some embodiments, the positioning model is deployed in the model inference device, and the channel data measurement device can send the predicted semantic quantization data to the model inference device. The model inference device can input the predicted semantic quantization data into the positioning model, and perform positioning based on the input to predict the positioning result corresponding to the channel measurement data, which can be referred to as predicted positioning result, including but not limited to terminal position (e.g., including coordinate information), ToA, AoA, etc.

[0212] FIG. 4 is an interaction diagram of semantic quantization, according to an embodiment of the present disclosure.

[0213] As shown in FIG. 4, based on the training of the semantic quantization model and the positioning model, the semantic quantization model can be deployed in the first device, and the semantic quantization model can be deployed in the second device.

[0214] For example, the first device can be referred to as a reference signal receiving device, and the second device can be referred to as a model inference device.

[0215] The reference signal sending device can send a reference signal for positioning, which can include at least one of PRS and SRS-Pos.

[0216] The reference signal receiving device can receive the reference signal for positioning, determine the channel measurement data based on the received reference signal, input the channel measurement data into the semantic quantization model, and predict the semantic quantization result corresponding to the channel measurement data, which can be referred to as predicted semantic quantization data. Further, the reference signal receiving device can send the predicted semantic quantization data to the model inference device.

[0217] The model inference device can input the predicted semantic quantization data into the positioning model, and predict the positioning result corresponding to the channel measurement data, including but not limited to direct positioning result (e.g., terminal position) and indirect positioning result (e.g., ToA, AoA).

[0218] In some embodiments, the names of information and the like are not limited to the names described in the embodiments, and terms such as "information", "message", "signal", "signaling", "report", "configuration", "indication", "instruction", "command", "channel", "parameter", "domain", "field", "symbol", "symbol", "codebook", "codeword", "codepoint", "bit", "data", "program", "chip", and the like can be replaced with each other.

[0219] In some embodiments, "acquire", "obtain", "get", "receive", "transmit", "bidirectional transmission", "send and / or receive", and the like can be replaced with each other, and can be interpreted as receiving from other subjects, acquiring from protocols, acquiring from higher layers, obtaining by self-processing, autonomously implementing, and the like.

[0220] In some embodiments, the terms "send", "transmit", "report", "issue", "transmit", "bidirectional transmission", "send and / or receive", and the like can be replaced with each other.

[0221] In some embodiments, the terms "certain", "preset", "pre-set", "set", "indicated", "certain", "arbitrary", "first", and the like can be replaced with each other, and "certain A", "preset A", "pre-set A", "set A", "indicated A", "certain A", "arbitrary A", "first A" can be interpreted as A specified in advance in protocols and the like, can be interpreted as A obtained by setting, configuring, or indicating, and the like, and can be interpreted as certain A, certain A, arbitrary A, or first A, but are not limited thereto.

[0222] Corresponding to the foregoing embodiments of the model determination method and the semantic quantization method, the disclosure also provides embodiments of a model determination device and a semantic quantization device.

[0223] FIG. 5 is a schematic block diagram of a model determination device according to an embodiment of the disclosure. As shown in FIG. 5, the model determination device includes a processing module 501.

[0224] In some embodiments, the processing module is configured to determine a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; train a first initial model and a second initial model based on the training sample set, wherein the first initial model at a training stop is a semantic quantization model, and the second initial model at the training stop is a positioning model; wherein an input of the semantic quantization model comprises channel measurement data, an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

[0225] In some embodiments, in the process of training the first initial model and the second initial model based on the training sample set, a difference between the predicted positioning result and an actual positioning result is used as a loss function.

[0226] In some embodiments, the condition for stopping training comprises at least one of the following: a result of the loss function is within a threshold range; a number of training cycles is greater than or equal to a number threshold.

[0227] In some embodiments, the positioning reference signal comprises at least one of the following: a positioning reference signal sent by a network device; a sounding reference signal sent by a terminal.

[0228] In some embodiments, the semantic quantization model is deployed on a first device, and / or the positioning model is deployed on a second device; the first device is configured to determine the channel measurement data by receiving a positioning reference signal, and send the predicted semantic quantization data to the second device.

[0229] In some embodiments, the channel measurement data comprises at least one of the following: related information of a positioning reference signal; a channel impulse response corresponding to the positioning reference signal; a complex number corresponding to the channel impulse response; power delay profile data corresponding to the channel impulse response; delay profile data corresponding to the channel impulse response; spectrum information of a channel corresponding to the positioning reference signal; point cloud information of the channel corresponding to the positioning reference signal.

[0230] In some embodiments, the related information of the positioning reference signal comprises at least one of the following: an amplitude of the positioning reference signal; a phase of the positioning reference signal; I path information of the positioning reference signal; Q path information of the positioning reference signal.

[0231] In some embodiments, the spectrum information comprises at least one of the following: time delay spread spectrum information; Doppler spectrum information; micro-Doppler spectrum information; angle spectrum information; signal strength spectrum information.

[0232] FIG. 6 is a schematic block diagram of a semantic quantization apparatus, according to an embodiment of the present disclosure. For example, the semantic quantization apparatus can be disposed in a first device. As shown in FIG. 6, the model determination apparatus includes a processing module 601 and a sending module 602.

[0233] In some embodiments, the processing module is configured to determine first channel measurement data by measuring a positioning reference signal; input the first channel measurement data into a semantic quantization model based on any one of the first aspect or the optional embodiments of the first aspect to determine first predicted semantic quantization data corresponding to the first channel measurement data.

[0234] In some embodiments, the sending module is configured to send the first predicted semantic quantization data to a second device, wherein the second device is deployed with a positioning model, and the positioning model determines a first predicted positioning result corresponding to the first channel measurement data based on the first predicted semantic quantization data.

[0235] Embodiments of the present disclosure also propose a model training apparatus. For example, the model training apparatus can be disposed in a first training device. For example, the first training device can be a communication device. For example, the first training device can be the first device, the second device, or the third device in the above embodiments, or can be another device.

[0236] In some embodiments, the model training apparatus includes a processing module configured to determine a training sample set, wherein a sample in the training sample set includes channel measurement data corresponding to a positioning reference signal, and a label of the sample includes a positioning result corresponding to the channel measurement data; train a first initial model based on the training sample set, and the first initial model at the time of stopping training is used as a semantic quantization model; wherein an input of the semantic quantization model includes channel measurement data, and an output of the semantic quantization model includes predicted semantic quantization data corresponding to the channel measurement data.

[0237] In some embodiments, the model training apparatus further includes a sending module configured to send predicted semantic quantization data output by the first initial model to a second training device, for the second device to train a second initial model; wherein the second initial model at the time of stopping training is used as a positioning model, an input of the positioning model includes the predicted semantic quantization data, and an output of the positioning model includes a predicted positioning result corresponding to the channel measurement data.

[0238] Embodiments of the present disclosure also provide a model training apparatus. For example, the model training apparatus can be arranged in a second training device. For example, the second training device can be a communication device. For example, the second training device can be the first device, the second device, or the third device in the above embodiments, or can be another device.

[0239] In some embodiments, the model training apparatus comprises a receiving module configured to receive predicted semantic quantization data output by the first initial model trained by the first training device; and a processing module configured to train a second initial model based on the predicted semantic quantization data. An input of the second initial model comprises the predicted semantic quantization data.

[0240] For the apparatus embodiments, since they basically correspond to the method embodiments, the relevant parts are described in the part of the method embodiments. The apparatus embodiments described above are merely illustrative. The modules described as separate components can or can not be physically separate, and the components displayed as modules can or can not be physical modules, i.e., they can be located in one place or distributed on multiple network modules. According to actual needs, some or all of the modules can be selected to achieve the purpose of the present embodiment. Those skilled in the art can understand and implement it without creative labor.

[0241] Embodiments of the present disclosure also provide an apparatus for implementing any of the above methods. For example, an apparatus is provided, which comprises units or modules for implementing the steps performed by a terminal in any of the above methods. For another example, another apparatus is provided, which comprises units or modules for implementing the steps performed by a network device (such as an access network device, a core network function node, a core network device, etc.) in any of the above methods.

[0242] 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 any of the above methods or realize the functions of each unit or module 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 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 of 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 above 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.

[0243] 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), and 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 reconfigurable. 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), and the like.

[0244] FIG. 7A is a structural schematic diagram of a communication device 7100 according to an embodiment of the present disclosure. The communication device 7100 can be a network device (for example, an access network device, a core network device, and the like), or a terminal (for example, a user equipment, and the like), or a chip, a chip system, or a processor supporting the network device to implement any of the above methods, or a chip, a chip system, or a processor supporting the terminal to implement any of the above methods. The communication device 7100 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.

[0245] As shown in FIG. 7A, the communication device 7100 includes one or more processors 7101. The processor 7101 can be a general processor or a special-purpose processor, etc., such as a baseband processor or a central processing unit. The baseband processor can be configured to process communication protocols and communication data, and the central processing unit can be configured to control a communication apparatus (e.g., a base station, a baseband chip, a terminal device, a terminal device chip, a DU or a CU, etc.), execute programs, and process data of the programs. Optionally, the communication device 7100 is configured to perform any of the above methods. Optionally, the one or more processors 7101 are configured to invoke instructions to cause the communication device 7100 to perform any of the above methods.

[0246] In some embodiments, the communication device 7100 further includes one or more transceivers 7102. When the communication device 7100 includes one or more transceivers 7102, the transceiver 7102 performs at least one of the communication steps (e.g., steps S201 and S202, but not limited to) in the above methods, and the processor 7101 performs at least one of the other steps (e.g., steps S201 and S202, but not limited to). In optional 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, interface circuit, interface, etc. can be replaced by each other, and the terms transmitter, transmitting unit, transmitter, transmitting circuit, etc. can be replaced by each other, and the terms receiver, receiving unit, receiver, receiving circuit, etc. can be replaced by each other.

[0247] In some embodiments, the communication device 7100 further includes one or more memories 7103 for storing data. Optionally, all or part of the memory 7103 can also be outside the communication device 7100. In optional embodiments, the communication device 7100 can include one or more interface circuits 7104. Optionally, the interface circuit 7104 is connected to the memory 7102, and the interface circuit 7104 can be configured to receive data from the memory 7102 or other devices, and can be configured to send data to the memory 7102 or other devices. For example, the interface circuit 7104 can read data stored in the memory 7102 and send the data to the processor 7101.

[0248] The communication device 7100 described in the above embodiments can be a network device or a terminal, but the scope of the communication device 7100 described in the present disclosure is not limited thereto, and the structure of the communication device 7100 can not be limited by FIG. 7A. The communication device can be a standalone device or can be part of a larger device. For example, the communication 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 vehicle-mounted device, a network device, a cloud device, an artificial intelligence device, and the like; (6) other devices, and the like.

[0249] FIG. 7B is a structural schematic diagram of a chip 7200 according to an embodiment of the present disclosure. For the case where the communication device 7100 is a chip or a chip system, the structural schematic diagram of the chip 7200 shown in FIG. 7B can be referred to, but is not limited thereto.

[0250] The chip 7200 includes one or more processors 7201. The chip 7200 is configured to perform any of the above methods.

[0251] In some embodiments, the chip 7200 further includes one or more interface circuits 7202. Optionally, the terms interface circuit, interface, transceiver pin, and the like can be replaced with each other. In some embodiments, the chip 7200 further includes one or more memories 7203 for storing data. Optionally, all or part of the memory 7203 can be outside the chip 7200. Optionally, the interface circuit 7202 is connected to the memory 7203, and the interface circuit 7202 can be configured to receive data from the memory 7203 or other devices, and the interface circuit 7202 can be configured to send data to the memory 7203 or other devices. For example, the interface circuit 7202 can read data stored in the memory 7203 and send the data to the processor 7201.

[0252] In some embodiments, the interface circuit 7202 performs at least one of the communication steps (such as steps S201, S202, but not limited thereto) of transmitting and / or receiving in the above methods. The interface circuit 7202 performing the communication steps such as transmitting and / or receiving in the above methods means that the interface circuit 7202 performs data interaction between the processor 7201, the chip 7200, the memory 7203, or a transceiver device. In some embodiments, the processor 7201 performs at least one of the other steps (such as steps S201, S202, but not limited thereto).

[0253] The modules and / or devices described in various embodiments of the virtual device, the physical device, the chip, etc. can be combined or separated according to circumstances. Alternatively, part or all of the steps can also be performed by multiple modules and / or devices in cooperation, which is not limited here.

[0254] The disclosure further provides a storage medium having instructions stored thereon, which, when executed on the communication device 7100, causes the communication device 7100 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.

[0255] The disclosure further provides a program product, which, when executed by the communication device 7100, causes the communication device 7100 to perform any of the above methods. Alternatively, the program product is a computer program product.

[0256] The disclosure further provides a computer program, which, when executed on a computer, causes the computer to perform any of the above methods.

Claims

1. A model determination method characterized by, The method is executed by a training device, and the method comprises: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises semantic quantization data; training a first initial model and a second initial model based on the training sample set, wherein the first initial model at a training stop time is a semantic quantization model, and the second initial model at the training stop time is a positioning model; wherein an input of the semantic quantization model comprises channel measurement data, an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

2. The method of claim 1, wherein, In the process of training the first initial model and the second initial model based on the training sample set, a difference between the predicted positioning result and an actual positioning result is used as a loss function.

3. The method of claim 1, wherein, The training stop condition comprises at least one of the following: a result of the loss function is within a threshold range; and a number of training cycles is greater than or equal to a number threshold.

4. The method according to any one of claims 1 to 3, characterized in that, The channel measurement data comprises at least one of the following: related information of the positioning reference signal; a channel impulse response corresponding to the positioning reference signal; a complex number corresponding to the channel impulse response; power-time delay spectrum data corresponding to the channel impulse response; time delay spectrum data corresponding to the channel impulse response; spectrum information of a channel corresponding to the positioning reference signal; and point cloud information of the channel corresponding to the positioning reference signal.

5. The method of claim 4, wherein, The related information of the positioning reference signal comprises at least one of the following: an amplitude of the positioning reference signal; a phase of the positioning reference signal; I-path information of the positioning reference signal; and Q-path information of the positioning reference signal.

6. The method of claim 4, wherein, The spectrum information comprises at least one of the following: time delay spread spectrum information; Doppler spectrum information; micro-Doppler spectrum information; angle spectrum information; and signal strength spectrum information.

7. The method according to any one of claims 1 to 6, characterized in that, The positioning reference signal comprises at least one of the following: a positioning reference signal sent by a network device; and a sounding reference signal sent by a terminal.

8. A model training method, comprising: The method is executed by a first training device, and the method comprises: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; training a first initial model based on the training sample set, wherein the first initial model at a training stop time is a semantic quantization model; and wherein an input of the semantic quantization model comprises channel measurement data, and an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data.

9. The method of claim 8, wherein, The method further comprises: sending, by the first training device, the predicted semantic quantization data output by the first initial model to a second training device, so as to train a second initial model by the second device; wherein the second initial model at a training stop time is a positioning model, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

10. A model training method, comprising: The method is executed by a second training device, and the method comprises: receive predicted semantic quantization data of a first initial model trained by a first training device; train a second initial model based on the predicted semantic quantization data; wherein an input of the second initial model comprises the predicted semantic quantization data.

11. A method of semantic quantification, characterized in that, performed by a first device, wherein the semantic quantization model is deployed in the first device, the method comprising: determining first channel measurement data by measuring a positioning reference signal; inputting the first channel measurement data into the semantic quantization model to determine first predicted semantic quantization data corresponding to the first channel measurement data.

12. The method of claim 11, wherein, The semantic quantization model is determined based on the following manner: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; training a first initial model and a second initial model based on the training sample set, wherein the first initial model at a training stop time is taken as the semantic quantization model, and the second initial model at the training stop time is taken as a positioning model; wherein an input of the semantic quantization model comprises channel measurement data, an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

13. The method of claim 11, wherein, The semantic quantization model is determined based on the following manner: determining a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; training a first initial model based on the training sample set, wherein the first initial model at a training stop time is taken as the semantic quantization model; wherein an input of the semantic quantization model comprises channel measurement data, and an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data.

14. The method according to any one of claims 11 to 13, characterized in that, The method further comprises: sending the first predicted semantic quantization data to a second device, wherein the positioning model is deployed in the second device, and the positioning model determines a first predicted positioning result corresponding to the first channel measurement data based on the first predicted semantic quantization data.

15. A model determination apparatus characterized by comprising: The apparatus comprises: a processing module configured to determine a training sample set, wherein a sample in the training sample set comprises channel measurement data corresponding to a positioning reference signal, and a label of the sample comprises a positioning result corresponding to the channel measurement data; train a first initial model and a second initial model based on the training sample set, wherein the first initial model at a training stop time is taken as the semantic quantization model, and the second initial model at the training stop time is taken as a positioning model; wherein an input of the semantic quantization model comprises channel measurement data, an output of the semantic quantization model comprises predicted semantic quantization data corresponding to the channel measurement data, an input of the positioning model comprises the predicted semantic quantization data, and an output of the positioning model comprises a predicted positioning result corresponding to the channel measurement data.

16. A model training apparatus, comprising: The apparatus comprises: The processing module is configured to determine a training sample set, wherein a sample in the training sample set includes channel measurement data corresponding to a positioning reference signal, and a label of the sample includes a positioning result corresponding to the channel measurement data; and train a first initial model based on the training sample set, and the first initial model at a training stop time is taken as a semantic quantization model. The input of the semantic quantization model includes channel measurement data, and the output of the semantic quantization model includes predicted semantic quantization data corresponding to the channel measurement data.

17. A model training apparatus, comprising: The apparatus comprises: The receiving module is configured to receive predicted semantic quantization data output by a first initial model trained by a first training device; The processing module is configured to train a second initial model based on the predicted semantic quantization data; The input of the second initial model includes the predicted semantic quantization data.

18. A semantic quantization apparatus, characterized by comprising: The apparatus comprises: The processing module is configured to determine first channel measurement data by measuring a positioning reference signal; and input the first channel measurement data into a semantic quantization model to determine first predicted semantic quantization data corresponding to the first channel measurement data.

19. A communications device, characterized by One or more processors; The communication device is configured to perform the model training method in any one of claims 1 to 10, and / or the semantic quantization method in any one of claims 11 to 14. The instructions, when executed on the communication device, cause the communication device to perform the model training method in any one of claims 1 to 10, and / or the semantic quantization method in any one of claims 11 to 14.

20. A storage medium, the storage medium storing instructions, wherein, The program product, when executed by the communication device, causes the communication device to perform the model training method in any one of claims 1 to 10, and / or the semantic quantization method in any one of claims 11 to 14.

21. A program product, characterized by ​

Images