Communication method and user equipment

Abstract

The present invention pertains to a communication method for a mobile communication system. This communication method has step in which a first user device transmits, to a network device, training data and / or reasoning data. The communication method also has a step in which the network device deletes secure data and / or private data from the training data and / or reasoning data. In addition, the communication method has a step in which the network device transmits, to a second user device, the post-deletion training data and / or post-deletion reasoning data. Moreover, the communication method has a step in which the second user device performs machine learning by using the post-deletion training data and / or post-deletion reasoning data.

Description

Communication Method

[0001] The present disclosure relates to a communication method.

[0002] In recent years, the Third Generation Partnership Project (3GPP) (registered trademark), a standardization project for mobile communication systems, has been studying the application of artificial intelligence (AI) technology, particularly machine learning (ML) technology, to wireless communication (air interface) in mobile communication systems.

[0003] 3GPP contribution: RP-213599, “New SI: Study on Artificial Intelligence (AI) / Machine Learning (ML) for NR Air Interface”

[0004] A communication method according to one aspect is a communication method in a mobile communication system. The communication method includes a step in which a first user device transmits training data and / or inference data to a network device. The communication method also includes a step in which the network device deletes at least one of security-targeted data and privacy-targeted data from the training data and / or inference data. The communication method further includes a step in which the network device transmits the deleted training data and / or the deleted inference data to a second user device. The communication method also includes a step in which the second user device performs machine learning using the deleted training data and / or the deleted inference data.

[0005] A communication method according to one aspect is a communication method in a mobile communication system. The communication method includes a step in which a network device receives training data and / or inference data from a first user device. The communication method also includes a step in which the network device requests a public key from a second user device. The communication method further includes a step in which the second user device, in response to the request, transmits a public key created from a private key in a machine learning model to the network device. The communication method also includes a step in which the network device encrypts the training data and / or inference data using the public key and transmits the encrypted training data and / or encrypted inference data to the second user device. The communication method also includes a step in which the machine learning model of the second user device decrypts the encrypted training data and / or the encrypted inference data using the private key and performs machine learning using the decrypted training data and / or the decrypted inference data.

[0006] Furthermore, according to one aspect, there is provided a communication method in a mobile communication system. The communication method includes a step in which a first machine learning model of a first user device encrypts training data and / or inference data using a common key. The communication method also includes a step in which the first user device transmits the encrypted training data and / or the encrypted inference data to a network device. The communication method also includes a step in which the network device transmits the encrypted training data and / or the encrypted inference data to a second user device. The communication method also includes a step in which a second machine learning model of the second user device decrypts the encrypted training data and / or the encrypted inference data using the common key, and performs machine learning using the decrypted training data and / or the decrypted inference data.

[0007] Furthermore, a communication method according to one aspect is a communication method in a mobile communication system. The communication method includes a step in which a network device transmits a program for performing machine learning to a first user device and a second user device. The communication method also includes a step in which the first user device encrypts training data and / or inference data using a common key created from the program, and transmits the encrypted training data and / or the encrypted inference data to the network device. The communication method also includes a step in which the network device transmits the encrypted training data and / or the encrypted inference data to a second user device. The communication method also includes a step in which the second user device decrypts the encrypted training data and / or the encrypted inference data using the common key created from the program, and performs machine learning using the decrypted training data and / or the decrypted inference data.

[0008] FIG. 1 is a diagram showing an example of the configuration of a mobile communication system according to the first embodiment. FIG. 2 is a diagram showing an example of the configuration of a UE (user equipment) according to the first embodiment. FIG. 3 is a diagram showing an example of the configuration of a gNB (base station) according to the first embodiment. FIG. 4 is a diagram showing an example of the configuration of a protocol stack according to the first embodiment. FIG. 5 is a diagram showing an example of the configuration of a protocol stack according to the first embodiment. FIG. 6 is a diagram showing an example of the configuration of functional blocks of AI / ML technology according to the first embodiment. FIG. 7 is a diagram showing an example of the configuration of a mobile communication system according to the first embodiment. FIGS. 8(A) and 8(B) are diagrams showing an example of deletion of learning inference data according to the first embodiment. FIG. 9 is a diagram showing an example of the configuration of a UE and a gNB according to the first embodiment. FIG. 10 is a diagram showing an example of the configuration of a UE and a gNB according to the first embodiment. FIG. 11 is a diagram showing an example of operation according to the first embodiment. FIG. 12 is a diagram showing an example of the configuration of a UE and a gNB according to the first embodiment. FIG. 13 is a diagram showing an example of the configuration of a UE and a gNB according to the first embodiment. FIG. 14 is a diagram showing an example of the configuration of a UE and a gNB according to the first embodiment. Fig. 15 is a diagram showing a configuration example of a UE and a gNB according to the first embodiment. Fig. 16 is a diagram showing a configuration example of a mobile communication system according to the second embodiment. Fig. 17 is a diagram showing a first operation example according to the second embodiment. Fig. 18 is a diagram showing a second operation example according to the second embodiment. Fig. 19 is a diagram showing an example of program transmission according to the third embodiment. Fig. 20 is a diagram showing an operation example according to the third embodiment.

[0009] When applying machine learning technology to mobile communication systems, it has not yet been established how to utilize machine learning technology, especially when the data used in machine learning is data subject to security or privacy.

[0010] Therefore, an object of the present disclosure is to make it possible to ensure security or privacy in machine learning in mobile communication systems.

[0011] [First embodiment] A mobile communication system according to a first embodiment will be described with reference to the drawings. In the description of the drawings, the same or similar parts are denoted by the same or similar reference numerals.

[0012] (Configuration of mobile communication system) The configuration of a mobile communication system according to the first embodiment will be described. FIG. 1 is a diagram showing an example of the configuration of a mobile communication system 1 according to the first embodiment. The mobile communication system 1 conforms to the 5th Generation System (5GS) of the 3GPP standard. Although 5GS will be described below as an example, the mobile communication system may also be at least partially applied to an LTE (Long Term Evolution) system. The mobile communication system may also be at least partially applied to a sixth generation (6G) system or later system.

[0013] The mobile communication system 1 includes a user equipment (UE) 100, a 5G radio access network (NG-RAN: Next Generation Radio Access Network) 10, and a 5G core network (5GC: 5G Core Network) 20. Hereinafter, the NG-RAN 10 may be simply referred to as the RAN 10. Furthermore, the 5GC 20 may be simply referred to as the core network (CN) 20.

[0014] The UE 100 is a mobile wireless communication device. The UE 100 may be any device that is used by a user. For example, the UE 100 may be a mobile phone terminal (including a smartphone) and / or a tablet terminal, a notebook PC, a communication module (including a communication card or a chipset), a sensor or a device provided in a sensor, a vehicle or a device provided in a vehicle (Vehicle UE), or an aircraft or a device provided in an aircraft (Aerial UE).

[0015] The NG-RAN 10 includes a base station (called a "gNB" in a 5G system) 200. The gNBs 200 are connected to each other via an Xn interface, which is an interface between base stations. The gNB 200 manages one or more cells. The gNB 200 performs wireless communication with a UE 100 that has established a connection with its own cell. The gNB 200 has a radio resource management (RRM) function, a routing function for user data (hereinafter simply referred to as "data"), a measurement control function for mobility control and scheduling, and the like. The term "cell" is used to indicate the smallest unit of a wireless communication area. The term "cell" is also used to indicate a function or resource for wireless communication with the UE 100. One cell belongs to one carrier frequency (hereinafter simply referred to as "frequency").

[0016] In addition, gNBs can also be connected to the Evolved Packet Core (EPC), which is the core network of LTE. LTE base stations can also be connected to 5GC. LTE base stations and gNBs can also be connected via an inter-base station interface.

[0017] The 5GC20 includes an AMF (Access and Mobility Management Function) and a UPF (User Plane Function) 300. The AMF performs various mobility controls for the UE 100. The AMF manages the mobility of the UE 100 by communicating with the UE 100 using NAS (Non-Access Stratum) signaling. The UPF controls data forwarding. The AMF and the UPF 300 are connected to the gNB 200 via an NG interface, which is an interface between a base station and a core network. The AMF and the UPF 300 may be core network devices included in the CN 20.

[0018] 2 is a diagram showing an example of the configuration of a UE 100 (user equipment) according to the first embodiment. The UE 100 includes a receiving unit 110, a transmitting unit 120, and a control unit 130. The receiving unit 110 and the transmitting unit 120 constitute a communication unit that performs wireless communication with the gNB 200. The UE 100 is an example of a communication device.

[0019] The receiving unit 110 performs various types of reception under the control of the control unit 130. The receiving unit 110 includes an antenna and a receiver. The receiver converts a radio signal received by the antenna into a baseband signal (received signal) and outputs the baseband signal to the control unit 130.

[0020] The transmitting unit 120 performs various transmissions under the control of the control unit 130. The transmitting unit 120 includes an antenna and a transmitter. The transmitter converts a baseband signal (transmission signal) output by the control unit 130 into a radio signal and transmits it from the antenna.

[0021] The control unit 130 performs various controls and processes in the UE 100. Such processes include processes of each layer described below. The control unit 130 includes at least one processor and at least one memory. The memory stores programs executed by the processor and information used in the processes by the processor. The processor may include a baseband processor and a CPU (Central Processing Unit). The baseband processor performs modulation / demodulation and encoding / decoding of baseband signals. The CPU executes programs stored in the memory to perform various processes.

[0022] 3 is a diagram showing an example of the configuration of a gNB 200 (base station) according to the first embodiment. The gNB 200 includes a transmitter 210, a receiver 220, a controller 230, and a backhaul communication unit 250. The transmitter 210 and the receiver 220 constitute a communication unit that performs wireless communication with the UE 100. The backhaul communication unit 250 constitutes a network communication unit that communicates with the CN 20. The gNB 200 is another example of a communication device.

[0023] The transmitting unit 210 performs various transmissions under the control of the control unit 230. The transmitting unit 210 includes an antenna and a transmitter. The transmitter converts a baseband signal (transmission signal) output by the control unit 230 into a radio signal and transmits it from the antenna.

[0024] The receiving unit 220 performs various types of reception under the control of the control unit 230. The receiving unit 220 includes an antenna and a receiver. The receiver converts a radio signal received by the antenna into a baseband signal (received signal) and outputs the baseband signal to the control unit 230.

[0025] The control unit 230 performs various controls and processes in the gNB 200. Such processes include processes for each layer, which will be described later. The control unit 230 includes at least one processor and at least one memory. The memory stores programs executed by the processor and information used in the processes by the processor. The processor may include a baseband processor and a CPU. The baseband processor performs modulation / demodulation and encoding / decoding of baseband signals. The CPU executes programs stored in the memory to perform various processes.

[0026] The backhaul communication unit 250 is connected to adjacent base stations via an Xn interface, which is an interface between base stations. The backhaul communication unit 250 is connected to the AMF / UPF 300 via an NG interface, which is an interface between a base station and a core network. The gNB 200 may be configured (i.e., functionally divided) with a central unit (CU) and a distributed unit (DU), and the two units may be connected via an F1 interface, which is a fronthaul interface.

[0027] FIG. 4 is a diagram showing an example of the configuration of a protocol stack of a radio interface of a user plane that handles data.

[0028] The user plane air interface protocol includes a physical (PHY) layer, a medium access control (MAC) layer, a radio link control (RLC) layer, a packet data convergence protocol (PDCP) layer, and a service data adaptation protocol (SDAP) layer.

[0029] The PHY layer performs encoding / decoding, modulation / demodulation, antenna mapping / demapping, and resource mapping / demapping. Data and control information are transmitted between the PHY layer of UE100 and the PHY layer of gNB200 via a physical channel. The PHY layer of UE100 receives downlink control information (DCI) transmitted from gNB200 on a physical downlink control channel (PDCCH). Specifically, UE100 performs blind decoding of the PDCCH using a radio network temporary identifier (RNTI) and acquires successfully decoded DCI as DCI addressed to the UE. The DCI transmitted from gNB200 has a CRC (Cyclic Redundancy Code) parity bit scrambled by the RNTI added.

[0030] In NR, UE100 can use a bandwidth narrower than the system bandwidth (i.e., the cell bandwidth). The gNB200 configures a bandwidth portion (BWP) consisting of contiguous PRBs (Physical Resource Blocks) for UE100. UE100 transmits and receives data and control signals in the active BWP. For example, up to four BWPs may be configurable for UE100. Each BWP may have a different subcarrier spacing. The frequencies of the BWPs may overlap with each other. When multiple BWPs are configured for UE100, the gNB200 can specify which BWP to apply by controlling the downlink. This allows the gNB200 to dynamically adjust the UE bandwidth according to the amount of data traffic of UE100, etc., and reduce UE power consumption.

[0031] The gNB 200 can configure, for example, up to three control resource sets (CORESETs) for each of up to four BWPs on the serving cell. The CORESET is a radio resource for control information to be received by the UE 100. Up to 12 or more CORESETs may be configured on the serving cell for the UE 100. Each CORESET may have an index of 0 to 11 or more. The CORESET may consist of six resource blocks (PRBs) and one, two, or three consecutive Orthogonal Frequency Division Multiplex (OFDM) symbols in the time domain.

[0032] The MAC layer performs data priority control, retransmission processing using Hybrid Automatic Repeat reQuest (HARQ), random access procedures, etc. Data and control information are transmitted between the MAC layer of the UE 100 and the MAC layer of the gNB 200 via a transport channel. The MAC layer of the gNB 200 includes a scheduler. The scheduler determines the uplink and downlink transport format (transport block size, modulation and coding scheme (MCS)) and the resource blocks to be allocated to the UE 100.

[0033] The RLC layer transmits data to the receiving RLC layer using the functions of the MAC layer and PHY layer. Data and control information are transmitted between the RLC layer of the UE 100 and the RLC layer of the gNB 200 via a logical channel.

[0034] The PDCP layer performs header compression / decompression, encryption / decryption, and the like.

[0035] The SDAP layer maps IP flows, which are units for Quality of Service (QoS) control by the core network, to radio bearers, which are units for QoS control by the access stratum (AS). Note that if the RAN is connected to the EPC, SDAP may not be required.

[0036] FIG. 5 is a diagram showing the configuration of a protocol stack of a radio interface of a control plane that handles signaling (control signals).

[0037] The protocol stack of the radio interface of the control plane includes a radio resource control (RRC) layer and a non-access stratum (NAS) instead of the SDAP layer shown in FIG.

[0038] RRC signaling for various settings is transmitted between the RRC layer of UE100 and the RRC layer of gNB200. The RRC layer controls logical channels, transport channels, and physical channels according to the establishment, re-establishment, and release of radio bearers. When there is a connection (RRC connection) between the RRC of UE100 and the RRC of gNB200, UE100 is in an RRC connected state. When there is no connection (RRC connection) between the RRC of UE100 and the RRC of gNB200, UE100 is in an RRC idle state. When the connection between the RRC of UE100 and the RRC of gNB200 is suspended, UE100 is in an RRC inactive state.

[0039] The NAS, which is located above the RRC layer, performs session management, mobility management, etc. NAS signaling is transmitted between the NAS of the UE 100 and the NAS of the AMF 300. Note that the UE 100 has an application layer and the like in addition to the radio interface protocol. Also, the layer below the NAS is called an Access Stratum (AS).

[0040] (AI / ML Technology) Next, the AI / ML technology according to the embodiment will be described. Fig. 6 is a diagram showing an example of the configuration of functional blocks of the AI / ML technology in the mobile communication system 1 according to the first embodiment.

[0041] The functional block configuration example shown in FIG. 6 includes a data collection unit A1, a model learning unit A2, a model inference unit A3, and a data processing unit A4.

[0042] The data collection unit A1 collects input data, specifically, learning data and inference data. The data collection unit A1 outputs the learning data to the model learning unit A2. The data collection unit A1 also outputs the inference data to the model inference unit A3. The data collection unit A1 may acquire data in the device on which the data collection unit A1 is provided as input data. The data collection unit A1 may also acquire data in another device as input data.

[0043] The model learning unit A2 performs model learning. Specifically, the model learning unit A2 optimizes parameters of the learning model through machine learning using the learning data, and derives (or generates, or updates) a learned model. The model learning unit A2 outputs the derived learned model to the model inference unit A3. For example, considering y = ax + b, a (slope) and b (intercept) are parameters, and optimizing these corresponds to machine learning. Generally, machine learning is classified into supervised learning, unsupervised learning, and reinforcement learning. Supervised learning is a method that uses correct answer data as learning data. Unsupervised learning is a method that does not use correct answer data as learning data. For example, in unsupervised learning, feature points are memorized from a large amount of learning data, and the correct answer is determined (range estimation). Reinforcement learning is a method of assigning a score to an output result and learning how to maximize the score. Although supervised learning will be described below, unsupervised learning or reinforcement learning may also be applied as machine learning.

[0044] The model inference unit A3 performs model inference. Specifically, the model inference unit A3 infers an output from inference data using a trained model and outputs the inference result data to the data processing unit A4. For example, in the case of y = ax + b, x corresponds to the inference data and y corresponds to the inference result data. Note that "y = ax + b" is a model. A model in which the slope and intercept are optimized, for example, "y = 5x + 3", is a trained model. Here, there are various model approaches, such as linear regression analysis, neural networks, and decision tree analysis. The above "y = ax + b" can be considered a type of linear regression analysis. The model inference unit A3 may provide model performance feedback to the model learning unit A2.

[0045] The data processing unit A4 receives the inference result data and performs processing that utilizes the inference result data.

[0046] (Communication Method According to First Embodiment) When a learner performs machine learning, the learner may use training data from an environment other than the learner's own environment. This allows the learner to overcome insufficient learning and create a trained model from a large amount of training data.

[0047] 7 is a diagram illustrating an example of the configuration of the mobile communication system 1 according to the first embodiment. The example illustrated in FIG. 7 illustrates an example in which the learning data and / or inference data (hereinafter, "learning data and / or inference data" may be referred to as "learning inference data") used when performing machine learning in the UE 100-1 is transmitted to the UE 100-2 via the gNB 200. In this case, the UE 100-2 creates a learning model using the learning data used in the UE 100-1, and can obtain an inference result using the inference data used in the UE 100-1.

[0048] In such a case, for example, the following case is assumed. That is, UE100-1 transmits "input: its own terminal model name, DL-TDOA (Downlink Time Difference Of Arrival), RSRP" and "output: location information" as learning inference data to UE100-2 via gNB200. Note that DL-TDOA represents a calculation method for the location information of UE100-1 calculated from the arrival time difference at UE100-1 of the signal transmitted from gNB200. Furthermore, RSRP (Reference Signal Received Power) represents the received power of the reference signal transmitted from gNB200.

[0049] In such a case, the UE 100-2 acquires the "terminal model name" and the current "location information" of the UE 100-1. That is, the UE 100-2 can grasp the current location of the terminal model name of the UE 100-1. Therefore, there is a possibility that security or privacy may become an issue for the UE 100-1.

[0050] Therefore, the first embodiment aims to ensure security or privacy in machine learning in a mobile communication system.

[0051] Therefore, in the first embodiment, first, a first user device (e.g., UE100-1) transmits training data and / or inference data to a network device (e.g., gNB200 or CN20). Second, the network device deletes at least one of security-targeted data and privacy-targeted data from the training data and / or inference data. Second, the network device transmits the deleted training data and / or the deleted inference data to a second user device (e.g., UE100-2). Third, the second user device performs machine learning using the deleted training data and / or the deleted inference data.

[0052] As described above, in the first embodiment, gNB200 or CN20 deletes security target data and / or privacy target data (hereinafter, "security target data and / or privacy target data" may be referred to as "target data") from the learning and inference data used by UE100-1. As a result, the learning and inference data of UE100-1 is transmitted to UE100-2 with the target data deleted. Therefore, UE100-2 does not receive target data related to the security or privacy of UE100-1. Therefore, in the first embodiment, it is possible to ensure security or privacy.

[0053] 8(A) and 8(B) are diagrams showing an example of deletion of learning inference data according to the first embodiment. As shown in FIG. 8(A), the gNB 200 may delete the target data. As shown in FIG. 8(B), the CN 20 may delete the target data. In FIGS. 8(A) and 8(B), "input: own terminal model name, DL-TDOA (Downlink Time Difference Of Arrival), RSRP" and "output: location information" are examples of learning inference data. Of these, this shows an example in which "own terminal model name" is deleted as target data.

[0054] (Configuration example of UE100 and gNB200) Next, a configuration example of UE100 and gNB200 will be described.

[0055] Figure 9 shows an example configuration of UE100-1 and gNB200 according to the first embodiment, and Figure 10 shows an example configuration of UE100-2 and gNB200 according to the first embodiment.

[0056] The configuration examples shown in Figures 9 and 10 represent a configuration example in which a "positioning accuracy enhancement" scenario is used as an AIML (Artificial Intelligence Machine Learning) operation scenario. The "positioning accuracy enhancement" scenario is an operation scenario in which the accuracy of the location information measured by the UEs 100-1 and 100-2 is improved using machine learning technology. In the "positioning accuracy enhancement" scenario shown in Figures 9 and 10, for example, "input: positioning reference signal (PRS)" and "output: location data" are used as learning inference data.

[0057] As shown in Figure 9, the UE 100-1 has a receiving unit 110-1, a transmitting unit 120-1, and a control unit 130-1. The control unit 130-1 includes a location information generating unit 133-1, a data collecting unit A1, a model learning unit A2, and a model inference unit A3. Also, as shown in Figure 9, the gNB 200 has a transmitting unit 210, a receiving unit 220, and a control unit 230. The control unit 230 includes a data processing unit A4.

[0058] The data collection unit A1 collects the PRS (the PRS received by the UE 100-1 may be referred to as PRS #1) received by the receiving unit 110 and the location data generated by the location information generation unit 133-1. The model learning unit A2 generates a learned model from the learning data (PRS #1 and location data), and the model inference unit A3 uses the learned model to obtain inference result data (location data) from the inference data (PRS #1 and location data).

[0059] In this way, in UE100-1, "input: PRS" and "output: location data" are used as learning and inference data. The transmitter 120-1 transmits the learning and inference data output from the data collector A1 to the gNB 200. In the gNB 200, the learning and inference data received by the receiver 220 is output to the transmitter 210 via the data processor A4.

[0060] The location information generating unit 133-1 generates location data of the UE 100-1 based on the PRS #1 received by the receiving unit 110-1. The positioning method may be the above-mentioned DL-TDOA method. The positioning method may be a multi-RTT (Roundtrip Time) method or a DL-AoD (Downlink Angle-of-Departure) method. The multi-RTT method is a positioning method in which, in each cell, a round trip time (round trip time) is measured from the time difference between transmission and reception, and distances (at least three distances) are calculated from the round trip times to locate the position of the UE 100-1. In addition, the DL-AoD method is a positioning method in which the angle of emission (AoD) of the PRS #1 is calculated from the received power of the PRS #1, and the location data of the UE 100-1 is obtained from the intersection position of the three directions.

[0061] The position information generation unit 133-1 may also generate position data based on GNSS (Global Navigation Satellite System) reception signals received by the GNSS receiver 150-1. In this case, the model training unit A2 generates a trained model using the training data (GNSS reception signals and position data). Furthermore, the model inference unit A3 uses the trained model to obtain inference result data (position data) from the inference data (GNSS reception signals and position data). In this case, the training inference data has "input: GNSS signal" and "output: position data."

[0062] In the "improving position accuracy" scenario, the "input" that is the subject of learning inference data may be, in addition to "PRS" and "GNSS received signal", at least one of the following, for example.

[0063] (X1) RSRP (Reference Signal Received Power), RSRQ (Reference Signal Received Quality), SINR (Signal-to-interference-plus-noise ratio), AD converter output waveform (These measurements may be PRS. These measurements may also be other received signals received from gNB200.)

[0064] (X2) LOS (Line of Sight) or NLOS (Non Line of Sight)

[0065] (X3) Measurement timing, accuracy, likelihood

[0066] (X4) RF fingerprint

[0067] (X5) Angle of Arrival (AOA) of the received signal, reception level for each antenna, reception phase for each antenna, and observed time difference of arrival (OTDOA) for each antenna

[0068] (X6) Received information of beacons used in wireless LANs (Local Area Networks) such as Wi-Fi (registered trademark) or short-range wireless communications such as Bluetooth (registered trademark)

[0069] (X7) Moving speed of UE 100 (The moving speed may be measured by GNSS receiver 150-1 or 150-2. The moving speed may be measured by a speed sensor in UE 100.)

[0070] In the gNB 200, the learning and inference data transmitted from the UE 100-1 is received by the receiving unit 220 and transmitted to the control unit 230. The control unit 230 (or the data processing unit A4) checks whether the learning and inference data includes target data, and if the target data is included, deletes the target data. The control unit 230 (or the data processing unit A4) outputs the deleted learning and inference data to the transmitting unit 210.

[0071] 10, the UE 100-2 includes a receiving unit 110-2, a transmitting unit 120-2, and a control unit 130-2. The control unit 130-2 includes a location information generating unit 133-2, a data collecting unit A1, a model learning unit A2, and a model inference unit A3.

[0072] The transmitting unit 210 of the gNB 200 transmits the learning inference data after deleting the target data to the UE 100-2.

[0073] The receiver 110-2 of UE100-2 receives the PRS transmitted from gNB200 (the PRS received by UE100-2 may be referred to as PRS #2) and the deleted learning and inference data transmitted from gNB200. The data collection unit A1 outputs the PRS received by the receiver 110-2 and the location data generated by the location information generation unit 133-2 to the model learning unit A2 as learning data and to the model inference unit A3 as inference data. At that time, the data collection unit A1 outputs the learning data of the deleted learning and inference data received by the receiver 110-2 to the model learning unit A2 and outputs the inference data to the model inference unit A3. In other words, the model learning unit A2 generates a trained model using the learning data (PRS #2 and location data) acquired by itself and the learning data (PRS #1 and location data) used by UE100-1 that does not include target data. The model inference unit A3 also obtains an inference result (location data) using inference data acquired by itself (PRS #2 and location data) and inference data used by UE 100-1 that does not include target data (PRS #1 and location data).

[0074] Note that the UE 100-2 may also acquire location data using the GNSS receiver 150-2. When acquiring location data using the GNSS receiver 150-2, the learning inference data may be a "GNSS reception signal" instead of "PRS#2".

[0075] (Example of Operation According to First Embodiment) Next, an example of operation according to the first embodiment will be described.

[0076] Fig. 11 is a diagram illustrating an example of operation according to the first embodiment. Fig. 11 illustrates an example of operation in the case where a gNB 200 is used as an example of a network device.

[0077] As shown in FIG. 11 , in step S10, the UE 100-1 transmits learning inference data to the gNB 200. The UE 100-1 may transmit the learning inference data to the gNB 200 by using an RRC message. Note that the gNB 200 may request the UE 100-1 to transmit the learning inference data. The gNB 200 may make the request by transmitting an RRC message (or a MAC CE (Control Element), or DCI (Downlink Control Information)) including information indicating a request to transmit the learning inference data. The UE 100 may transmit the learning inference data in response to receiving the request.

[0078] In step S11, the gNB 200 checks whether the learning inference data includes data that is problematic in terms of security or privacy (i.e., target data). The target data may be specified by the UE 100-1 or UE 100-2. The target data may be hard-coded in the specifications. When the UE 100-1 or UE 100-2 specifies the target data, the UE 100-1 or UE 100-2 may specify the target data by transmitting an RRC message (or MAC CE, or DCI) including information indicating the target data to the gNB 200. Note that when machine learning is performed in the gNB 200, the target data is not visible to general users, so machine learning may be performed including the target data.

[0079] In step S12, the gNB 200 deletes the target data from the learning inference data. For example, in the example described above, the gNB 200 deletes the "terminal model name" (target data) from "input: its own terminal model name, DL-TDOA, RSRP" and "output: location information" (learning inference data). The UE 100-2 may specify the target data to be deleted. The UE 100-2 may perform this specification by transmitting an RRC message (or MAC CE, or DCI) including information indicating the target data from the gNB 200. The gNB 200 will delete the specified target data from the learning inference data.

[0080] In step S13, the gNB 200 transmits the deleted learning and inference data to the UE 100-2. The gNB 200 may transmit the deleted learning and inference data to the UE 100-2 using an RRC message. The UE 100-2 performs machine learning using the deleted learning and inference data.

[0081] (Another example 1 of the first embodiment) In the first embodiment, the gNB 200 has been described as an example of a network device, but this is not limited to this. The network device may be a core network device (CN) 20. The CN 20 may be an AMF. The CN 20 may be a UPF. The CN 20 may be another core network device.

[0082] Figure 11 is also used for an example of operation when the network device is CN20. In this case, in Figure 11, gNB200 can be replaced with CN20. Furthermore, an NAS message may be used instead of an RRC message between UE100-1 and gNB200. Similarly, an NAS message may be used instead of an RRC message between gNB200 and UE100-2.

[0083] (Another Example 2 According to First Embodiment) In the first embodiment, an example in which the "position accuracy improvement" scenario is used as the operation scenario has been described, but the present invention is not limited to this. For example, "CSI (Channel State Information) feedback enhancement" may be used as the operation scenario.

[0084] The "CSI Feedback Improved" scenario represents, for example, an operation scenario in which machine learning technology is applied to CSI feedback fed back from the UE 100 to the gNB 200. The learning inference data in the "CSI Feedback Improved" scenario is, for example, "input: CSI-RS" and "output: CSI".

[0085] Figure 12 shows an example configuration of UE100-1 and gNB200 when the "Improved CSI Feedback" scenario is applied, and Figure 13 shows an example configuration of UE100-2 and gNB200 when the "Improved CSI Feedback" scenario is applied.

[0086] As shown in FIG. 12, in the "CSI feedback improvement" scenario, the CSI reference signal (CSI-RS) received from the gNB 200 (the CSI-RS received by the UE 100-1 may be referred to as "CSI-RS #1") and the CSI generated by the CSI generation unit 131-1 from the CSI-RS #1 are used as learning data. The model learning unit A2 uses the learning data (CSI-RS #1 and CSI) to generate a learned model. Furthermore, the model inference unit A3 uses the learned model to obtain inference result data (CSI) from the inference data (CSI-RS #1 and CSI). The transmission unit 120-1 transmits the learning inference data to the gNB 200. In this case, the learning inference data is "input: CSI-RS" and "output: CSI". In gNB200, the target data is deleted from the learning inference data, as in the first embodiment.

[0087] The CSI includes at least one of a channel quality indicator (CQI), a precoding matrix indicator (PMI), and a rank indicator (RI). The gNB 200 performs downlink scheduling and the like based on the CSI.

[0088] As shown in FIG. 13, the transmitter 210 of the gNB 200 transmits the learning inference data (CSI-RS #1 and CSI) received from the UE 100-1 to the UE 100-2. The model learning unit A2 of the UE 100-2 generates a trained model using the learning data (CSI-RS #2 and CSI) acquired by itself and the learning data (CSI-RS #1 and CSI) used by the UE 100-1 after the target data has been deleted. In addition, the model inference unit A3 obtains an inference result (CSI) using the inference data (CSI-RS #2 and CSI) acquired by itself and the inference data (CSI-RS #1 and CSI) used by the UE 100-1 after the target data has been deleted.

[0089] The operation example in the "Improved CSI Feedback" scenario also uses Fig. 11. In this case, the same processing may be performed, except that the target of the learning inference data is different from that in the "Improved Location Accuracy" scenario.

[0090] In the "CSI feedback improvement" scenario, the data used for "input" of learning inference data may be any of the following data in addition to CSI-RS.

[0091] (Y1) RSRP, RSRQ, SINR, or AD converter output waveform (these measurements may be CSI-RS. These measurements may also be other received signals received from gNB200.)

[0092] (Y2) Bit Error Rate (BER) or Block Error Rate (BLER) (The total number of transmitted bits (or the total number of transmitted blocks) is known, and the BER (or BLER) may be measured based on the CSI-RS.)

[0093] (Y3) Moving speed of UE 100 (The moving speed may be measured by GNSS receiver 150-1 or 150-2 in UE 100. The moving speed may be measured by a speed sensor in UE 100.)

[0094] (Another example 3 according to the first embodiment) In the first embodiment, a "beam management" scenario may be applied as the operation scenario. The "beam management" scenario is, for example, an operation scenario that uses machine learning technology to manage which beam is the optimal beam among the beams transmitted from gNB200. The learning inference data in the "beam management" scenario is, for example, "input: CSI-RS" and "output: optimal beam (information representing)".

[0095] Figure 14 shows an example configuration of UE100-1 and gNB200 when the "beam management" scenario is applied. Figure 15 shows an example configuration of UE100-2 and gNB200 when the "beam management" scenario is applied.

[0096] As shown in FIG. 14, UE 100-1 uses, as learning data, the CSI-RS received from gNB 200 and the optimal beam determined from the CSI-RS in optimal beam determination unit 132-1. Model learning unit A2 generates a learned model using the learning data (CSI-RS #1 and optimal beam). Furthermore, model inference unit A3 uses the learned model to obtain an inference result (optimal beam) from the inference data (CSI-RS #1 and optimal beam). Transmitting unit 120 transmits the learning inference data to gNB 200. The learning inference data is "input: CSI-RS" and "output: optimal beam." In gNB 200, target data is deleted from the learning inference data, as in the first embodiment.

[0097] As shown in FIG. 15, the transmitter 210 of the gNB 200 transmits the learning inference data (CSI-RS #1 and optimal beam) received from the UE 100-1 to the UE 100-2. The model learning unit A2 of the UE 100-2 generates a trained model using the learning data (CSI-RS #2 and optimal beam) acquired by itself and the learning data (CSI-RS #1 and optimal beam) used by the UE 100-1 after the target data has been deleted. In addition, the model inference unit A3 obtains an inference result (optimal beam) using the inference data (CSI-RS #2 and optimal beam) acquired by itself and the inference data (CSI-RS #1 and optimal beam) used by the UE 100-1 after the target data has been deleted.

[0098] The operation example in the "beam management" scenario also uses Fig. 11. In this case, the only difference is that the target of the learning inference data is different from that in the "position accuracy improvement" scenario, and the same processing may be performed.

[0099] In addition, in the "beam management" scenario, the data used to "input" the learning inference data may be any of the following data in addition to "CSI-RS".

[0100] (Z1) SSB (Synchronization Signal Block) received from gNB200

[0101] (Z2) RSRP, RSRQ, SINR, or AD converter output waveform (these measurements may be CSI-RS. These measurements may also be other received signals received from gNB200.)

[0102] (Z3) BER or BLER (BER (or BLER) may be measured based on CSI-RS with the total number of transmission bits (or the total number of transmission blocks) known.)

[0103] (Z4) Number of beams or beam pattern

[0104] (Z5) Beam measurement value(s)

[0105] (Z6) Moving speed of UE 100 (The moving speed may be measured by GNSS receiver 150-1 or 150-2 in UE 100. The moving speed may be measured by a speed sensor in UE 100.)

[0106] Second Embodiment Next, a second embodiment will be described.

[0107] In the second embodiment, an example in which learning and inference data is encrypted will be described.

[0108] Fig. 16 is a diagram illustrating an example of the configuration of a mobile communication system 1 according to the second embodiment. As shown in Fig. 16, the communication content is kept secret by encrypting the learning inference data. Generally, encryption keeps the data content secret from anyone other than the user who received the data.

[0109] In contrast, in the second embodiment, the UE 100-2 decrypts the encrypted learning and inference data within the machine learning model. In principle, the user of the UE 100-2 cannot (or does not have specific authority to) understand what processing is being performed within the machine learning model. Therefore, even if the encrypted learning and inference data is decrypted within the machine learning model of the UE 100-2, the confidentiality of the learning and inference data content can be maintained. Therefore, in the second embodiment, it is possible to ensure security or privacy in machine learning in the mobile communication system 1.

[0110] The machine learning model refers to, for example, a component (or block) on which machine learning is performed. The machine learning model may be the entire block represented by the functional block diagram shown in FIG. 6. For example, in the second embodiment, the data collection unit A1 of the machine learning model may encrypt and / or decrypt the learning and inference data. In principle, the user of UE 100-2 cannot access the data collection unit A1, and therefore cannot understand the contents of the decrypted learning and inference data (the learning and inference data used by UE 100-1).

[0111] In the second embodiment, two encryption methods, a secret key method and a common key method, will be described.

[0112] (First Operation Example According to Second Embodiment: Secret Key Scheme) Next, a first operation example according to the second embodiment will be described. In the first operation example, an example will be described in which a secret key scheme is used as the encryption scheme.

[0113] Specifically, first, a network device (e.g., gNB200 or CN20) receives training data and / or inference data from a first user device (e.g., UE100-1). Second, the network device requests a public key from a second user device (e.g., UE100-2). Third, in response to the request, the second user device transmits a public key created from a private key in the machine learning model to the network device. Fourth, the network device encrypts the training data and / or inference data using the public key and transmits the encrypted training data and / or encrypted inference data to the second user device. Fifth, the second user device decrypts the encrypted training data and / or encrypted inference data using the private key and performs machine learning in the machine learning model using the decrypted training data and / or decrypted inference data.

[0114] Thus, in the first operation example, the gNB 200 uses a public key to encrypt the learning inference data used by the UE 100-1, thereby maintaining the confidentiality of the learning inference data between the gNB 200 and the UE 100-2. Also, in the first operation example, the encrypted learning inference data is decrypted in the machine learning model of the UE 100-2, so it is not easy for the user using the UE 100-2 to understand the contents of the learning inference data. Therefore, in the first operation example, it is possible to ensure security or privacy in machine learning in the mobile communication system 1.

[0115] FIG. 17 is a diagram illustrating a first operation example according to the second embodiment.

[0116] 17, in step S20, the UE 100-1 transmits learning inference data to the gNB 200. The UE 100-1 may transmit the learning inference data by using an RRC message or the like. The gNB 200 may request the UE 100-1 to transmit the learning inference data. The request may be made by using an RRC message, a MAC CE, a DCI, or the like.

[0117] In step S21, the gNB 200 requests a public key from the UE 100-2. The gNB 200 may make the request to the AS of the UE 100 using an RRC message, a MAC CE, or a DCI. The UE 100-2 may make a request to the gNB 200 to transmit learning inference data. At this time, the UE 100-2 may specify an encryption method (for example, a secret key method). The transmission request and the specification may also use an RRC message, a MAC CE, or a DCI.

[0118] In step S22, the UE 100-2 creates a public key from the private key in the machine learning model in response to the request for the public key. In the UE 100-2, for example, the following processing is performed. That is, the AS of the UE 100-2 requests a public key from the data collection unit A1 of the machine learning model in response to the request. The data collection unit A1 holds the private key in a memory or the like, and creates a public key from the private key in response to the request. The data collection unit A1 outputs the created public key to the AS of the UE 100-2 (or the NAS of the UE 100-2). Note that the AS of the UE 100-2 (or the NAS of the UE 100-2) may have received the public key in advance from the machine learning model (i.e., the data collection unit A1).

[0119] In step S23, the UE 100-2 transmits the public key to the gNB 200. The AS of the UE 100-2 may transmit the public key by using an RRC message or the like.

[0120] In step S24, gNB200 encrypts the learning inference data using the public key.

[0121] In step S25, the gNB 200 transmits the encrypted learning inference data to the UE 100-2. The gNB 200 may transmit the encrypted learning inference data using an RRC message or the like.

[0122] In step S26, the UE 100-2 decrypts the encrypted learning inference data using a private key. In the UE 100-2, for example, the following processing is performed. That is, the AS of the UE 100-2 (or the NAS of the UE 100-2) outputs the encrypted learning inference data received from the gNB to the data collection unit A1 of the machine learning model. The data collection unit A1 decrypts the encrypted learning inference data using a private key stored in a memory or the like.

[0123] In step S27, the UE 100-2 performs machine learning using the decoded learning and inference data. The UE 100-2 performs, for example, the following processing. That is, the data collection unit A1 outputs the decoded learning data to the model learning unit A2, and outputs the decoded inference data to the model inference unit A3. That is, the machine learning model performs machine learning using the decoded learning and inference data.

[0124] In the first operation example, each of the operation scenarios described above (the "position accuracy improvement" scenario, the "CSI feedback improvement" scenario, or the "beam management" scenario) can be applied. When the "position accuracy improvement" scenario is applied, the learning and inference data may be, for example, "input: PRS" and "output: position data." When the "CSI feedback improvement" scenario is applied, the learning and inference data may be, for example, "input: CSI-RS" and "output: CSI." Furthermore, when the "beam management" scenario is applied, the learning and inference data may be, for example, "input: CSI-RS" and "output: optimal beam." The learning and inference data may be data according to each operation scenario.

[0125] (Another example 1 of the first operation example) In the first operation example, an example of encrypting learning inference data has been described, but the object to be encrypted may be all of the learning inference data. The object to be encrypted may be part of the learning inference data. For example, the gNB 200 may encrypt data related to privacy among the learning inference data. The object to be encrypted may be determined by the gNB 200.

[0126] (Another example 2 of the first operation example) In the first operation example, the gNB200 has been described as an example of a network device, but this is not limiting. The network device may be the CN20. When the network device is the CN20, the gNB200 may be read as the CN20 in the operation example shown in FIG. 17 . In this case, a NAS message may be used instead of an RRC message between the UE100-1 and the CN20. Furthermore, the processing performed in the AS of the UE100-1 may be read as the processing performed in the NAS of the UE100-1. Furthermore, a NAS message may be used instead of an RRC message between the CN20 and the UE100-2. In this case, the processing performed in the AS of the UE100-2 may be read as the processing performed in the NAS of the UE100-2.

[0127] (Second Operation Example According to Second Embodiment) The second operation example is an example in which a common key is used as an encryption method.

[0128] Specifically, first, a first machine learning model of a first user device (e.g., UE100-1) encrypts the training data and / or the inference data using a common key. Second, the first user device transmits the encrypted training data and / or the encrypted inference data to a network device (e.g., gNB200 or CN20). Third, the network device transmits the encrypted training data and / or the encrypted inference data to a second user device (e.g., UE100-2). Fourth, a second machine learning model of the second user device decrypts the encrypted training data and / or the encrypted inference data using a common key, and performs machine learning using the decrypted training data and / or the decrypted inference data.

[0129] In this way, in the second operation example, UE 100-1 encrypts the learning inference data using a public key, so that the confidentiality of the learning inference data can be maintained between UE 100-1 and UE 100-2. Also, in the second operation example, decryption is performed in the machine learning model of UE 100-2, so as in the first operation example, it is not easy for a user using UE 100-2 to understand the contents of the learning inference data. Therefore, the second operation example also makes it possible to ensure security or privacy.

[0130] FIG. 18 is a diagram illustrating a second operation example according to the second embodiment.

[0131] As shown in FIG. 18, in step S30, the UE 100-1 encrypts the learning inference data using a common key. For example, the data collection unit A1 in the machine learning model of the UE 100-1 encrypts the learning inference data using a common key stored in a memory or the like. The data collection unit A1 outputs the encrypted learning inference data to the AS of the UE 100-1 (or the NAS of the UE 100-2). Note that the gNB 200 may request the UE 100-1 to transmit the learning inference data, or may specify an encryption method (for example, a common key method). The request and the specification may be made using an RRC message, a MAC CE, a DCI, or the like.

[0132] The common key may be transmitted in advance from the gNB 200 to the data collection unit A1 of the UE 100-1. Conversely, the data collection unit A1 may transmit the common key used by itself to the gNB 200. In the latter case, the gNB 200 may transmit the common key to the UE 100-2 (data collection unit A1). Alternatively, the common key may be hard-coded by being stored in advance in the memory of the data collection unit A1, etc.

[0133] In step S31, the UE 100-1 transmits the encrypted learning inference data to the gNB 200. For example, the AS of the UE 100-1 may transmit the learning inference data to the gNB 200 using an RRC message or the like.

[0134] In step S32, the gNB 200 transmits the encrypted learning inference data to the UE 100-2. The UE 100-2 may request the gNB 200 to transmit the learning inference data in advance. The UE 100-2 may specify an encryption method (for example, a common key method) to the gNB 200. The request and the specification may be made using an RRC message, a MAC CE, a DCI, or the like.

[0135] In step S33, the UE 100-2 decrypts the encrypted learning inference data using a common key. The UE 100-2 performs, for example, the following processing. That is, the AS of the UE 100-2 outputs the received learning inference data to a data collection unit A1 in the machine learning model of the UE 100-2. The data collection unit A1 decrypts the encrypted learning inference data using a common key stored in a memory or the like. The common key may be transmitted from the gNB 200. The common key may be hard-coded, for example, by being stored in advance in the memory.

[0136] In step S34, the UE 100-2 performs machine learning using the decoded learning and inference data. For example, the data collection unit A1 outputs the decoded learning data to the model learning unit A2, and outputs the decoded inference data to the model inference unit A3. In the machine learning model of the UE 100-2, machine learning is performed using the learning and inference data of the UE 100-1.

[0137] In the second operation example, as in the first operation example, each of the above-mentioned operation scenarios (the "position accuracy improvement" scenario, the "CSI feedback improvement" scenario, or the "beam management" scenario) can be applied. The learning inference data may be data corresponding to each operation scenario.

[0138] (Another example 1 of the second operation example) In the second operation example, the target to be encrypted may be all of the learning inference data. The target to be encrypted may be a portion of the learning inference data. For example, the gNB200 encrypts data related to privacy among the learning inference data. The data to be encrypted may be determined by the gNB200. The gNB200 may instruct the UE100-2 on the data to be encrypted.

[0139] (Another example 2 of the second operation example) In the second operation example, an example in which encryption is performed by UE100-1 has been described, but this is not limiting. For example, encryption may be performed by gNB200. In this case, UE100-1 transmits the learning inference data to gNB200 without encryption (step S31), and the learning inference data is encrypted in gNB200 using a common key. The common key may be transmitted in advance from UE100-1 or UE100-2 to gNB200. The common key may be stored in advance in a memory or the like in gNB200 (or may be hard-coded).

[0140] (Another example 3 of the second operation example) In the second operation example, the network device is not limited to gNB200, and may be CN20. In this case, in the operation example shown in FIG. 18, gNB200 may be read as CN20. Furthermore, between UE100-1 and gNB200, and between gNB200 and UE100-2, NAS messages may be used instead of RRC messages. In this case, the processing performed in the AS of UE100-1 may be read as the processing performed in the NAS of UE100-1. Furthermore, the processing performed in the AS of UE100-2 may be read as the processing performed in the NAS of UE100-2.

[0141] Third Embodiment Next, a third embodiment will be described.

[0142] In the first operation example of the second embodiment, an example in which a public key is transmitted has been described. In the second operation example of the second embodiment, an example in which a common key is transmitted has been described. In the third embodiment, an example in which such key information is not transmitted or received will be described.

[0143] When machine learning is performed in the mobile communication system 1, the data can be divided into parts that change and parts that do not change as the machine learning is performed. For example, the model part of a trained model or the data part such as training data are parts that change. On the other hand, for example, the program that executes machine learning is a part that does not change. In the mobile communication system 1 according to the third embodiment, a common key is created using the part that does not change, i.e., the program that executes machine learning, and the common key is used to encrypt and decrypt the learning inference data.

[0144] Specifically, first, a network device (e.g., gNB200 or CN20) transmits a program for performing machine learning to a first user device (e.g., UE100-1) and a second user device (e.g., UE100-2). Second, the first user device encrypts the training data and / or inference data using a common key created from the program, and transmits the encrypted training data and / or encrypted inference data to the network device. Third, the network device transmits the encrypted training data and / or encrypted inference data to the second user device. Fourth, the second user device decrypts the encrypted training data and / or encrypted inference data using a common key created from the program, and performs machine learning using the decrypted training data and / or decrypted inference data.

[0145] As described above, in the third embodiment, the UE 100-1 encrypts the learning and inference data using a common key created from a program for executing machine learning, and the UE 100-2 decrypts the encrypted learning and inference data using a common key created from the program. Therefore, since the learning and inference data is encrypted between the UE 100-1 and the UE 100-2, the learning and inference data used by the UE 100-1 can be kept secret.

[0146] In addition, in the UE 100-2, since the common key is created from the program used when executing machine learning, the common key is created in the machine learning model, and as in the second embodiment, the user who uses the UE 100-2 cannot, in principle, grasp the decrypted learning inference data. Therefore, as in the second embodiment, the third embodiment can also ensure security or privacy.

[0147] Furthermore, in the third embodiment, since there is no need to send or receive a common key, processing is easier between UE100-1 and 100-2 and gNB200 compared to when a common key is sent or received, and it is also possible to make effective use of communication resources.

[0148] FIG. 19 is a diagram illustrating an example of program transmission according to the third embodiment.

[0149] As shown in FIG. 19, gNB200 transmits a program for executing a machine learning model (e.g., AIML_α) to UE100-1 and UE100-2. Also, gNB200 transmits a program for executing a machine learning model (e.g., AIML_β) to UE100-3. Then, assume that gNB200 creates a common key from the program for executing AIML_α, encrypts the learning inference data using the common key, and transmits the encrypted learning inference data (e.g., broadcasts). In such a case, UE100-1 and 100-2 can create a common key using the same AIML_α as gNB200, and therefore can decrypt the encrypted learning inference data. However, since UE100-3 holds a program for executing AIML_β, even if a common key is created using AIML_β, the learning inference data encrypted in gNB200 cannot be decrypted.

[0150] As described above, in the third embodiment, the gNB 200 can also broadcast learning and inference data without considering which machine learning model the UEs 100-1 to 100-3 are using. That is, when the gNB 200 transmits learning and inference data, if the learning and inference data is consistent with the UE 100's own machine learning model, the UE 100 can decode the learning and inference data transmitted from the gNB 200 and successfully generate a machine learning model. On the other hand, if the learning and inference data is not consistent with the UE 100's own machine learning model, the UE 100 cannot decode the learning and inference data transmitted from the gNB 200 and cannot use the learning and inference data. Therefore, the gNB 200 can broadcast the learning and inference data without considering the consistency with the machine learning model used by the UEs 100-1 to 100-3. Alternatively, the gNB 200 can broadcast the learning and inference data without receiving notification by signaling from the UEs 100-1 to 100-3 as to which machine learning model they are using.

[0151] For example, as in the first embodiment, assume that UE 100-1 transmits the learning inference data that it has used to UE 100-2 via gNB 200. Even in such a case, for the machine learning performed by UE 100-1 and UE 100-2, gNB 200 transmits a program for performing the machine learning to UE 100-1 and 100-2, so that UE 100-1 and 100-2 can perform encryption and decryption using the same common key created from the program.

[0152] In addition, the program for performing machine learning may be transmitted from each of the UEs 100-1 to 100-3 to the gNB 200. In this case, the machine learning may be performed in the gNB 200.

[0153] (Example of Operation According to Third Embodiment) Next, an example of operation according to the third embodiment will be described.

[0154] FIG. 20 is a diagram showing an example of operation according to the third embodiment. The example of operation shown in FIG. 20 will be described using an example (FIG. 19) in which the gNB 200 transmits a program for performing machine learning. Therefore, before the example of operation shown in FIG. 20 is started, the gNB 200 transmits the same program for performing machine learning (for example, a program for executing AIML_α) to the UE 100-1 and the UE 100-2. For example, the gNB 200 may transmit the program using an RRC message or the like.

[0155] In step S40, the gNB 200 creates a common key using a program for performing machine learning that was transmitted to the UEs 100-1 and 100-2. Any method may be used to create the common key. For example, the gNB 200 may use the program as input to obtain a hash value using a hash function, and use the hash value as the common key. The hash function may also be any function. The hash function may be, for example, a SHA-2 series such as SHA (Secure Hash Algorithm)-256. The hash function may be a SHA-3 series such as SHA3-256. The hash function may be MD5 (Message Digest 5). For example, the gNB 200 creates a common key from a program for executing AIML_α.

[0156] In step S41, the gNB 200 decides to broadcast the learning inference data and encrypts the learning inference data using a common key. For example, the gNB 200 encrypts the learning inference data using a common key created from a program for executing AIML_α.

[0157] In step S42, the gNB 200 broadcasts the encrypted learning inference data. The gNB 200 may broadcast using an RRC message such as system information (SIB), or may broadcast using a paging message. Alternatively, the gNB 200 may broadcast using an MBS (Multicast and Broadcast Services) message. The gNB 200 may broadcast using a dedicated message of an AIML-dedicated layer. The UE 100-1 and / or the UE 100-2 may request the gNB 200 to transmit the learning inference data. The request may be made using an RRC message, a MAC CE, a DCI, or the like. The gNB 200 may broadcast the encrypted learning inference data in response to the request. The UE 100-1 receives the encrypted learning inference data, and the UE 100-2 also receives the learning inference data.

[0158] In step S43, UE100-1 creates a common key using a program for executing machine learning received from gNB200, and uses the common key to decrypt the encrypted learning inference data. For example, the data collection unit A1 in the machine learning model of UE100-1 creates a common key using a program that executes AIML_α, and uses the created common key to decrypt the encrypted learning inference data. Since UE100-1 uses the same common key as the common key used by gNB200, it can decrypt the learning inference data encrypted by gNB200.

[0159] In step S44, the UE 100-1 performs machine learning using the decoded learning inference data.

[0160] In step S45, UE100-2 also creates a common key using a program that executes machine learning received from gNB200, and uses the common key to decrypt the encrypted learning inference data. For example, the data collection unit A1 in the machine learning model of UE100-2 creates a common key using a program that executes AIML_α, and uses the created common key to decrypt the encrypted learning inference data. Since UE100-2 also uses the same common key as the common key used by gNB200, it can decrypt the learning inference data encrypted by gNB200.

[0161] In step S46, the UE 100-2 also performs machine learning using the decoded learning inference data.

[0162] Step S43 and step S45 may be performed simultaneously. Step S44 and step S46 may also be performed simultaneously. With respect to steps S43 to S46, if machine learning is performed after decoding, the UE 100-1 may perform the process before the UE 100-2. The UE 100-2 may perform the process before the UE 100-1. The UE 100-1 and the UE 100-2 may perform the process simultaneously.

[0163] In the third embodiment, the above-described operation scenarios ("position accuracy improvement" scenario, "CSI feedback improvement" scenario, or "beam management" scenario) can also be applied. The learning inference data is data corresponding to each operation scenario.

[0164] (Another example 1 according to the third embodiment) In the third embodiment, an example has been described in which the gNB 200 transmits a program for performing machine learning to the UEs 100-1 and 100-2, but this is not limiting. For example, the CN 20 may transmit the program to the UEs 100-1 and 100-2. In this case, the CN 20 may transmit the program using a NAS message. Furthermore, the CN 20 may generate a common key from the program, encrypt the learning inference data using the common key, and transmit (or notify) the encrypted learning inference data.

[0165] (Another example 2 according to the third embodiment) In the third embodiment, an example in which learning inference data is transmitted from the gNB 200 to the UE 100 has been described, but this is not limited to this. For example, a trained model on which machine learning has been performed in the gNB 200 may be transmitted to the UE 100-1 and the UE 100-2. The UE 100-1 and the UE 100-2 may further perform machine learning using the trained model.

[0166] [Other Embodiments] In the first to third embodiments, examples in which gNB200 or CN20 is used have been described, but the present invention is not limited to this. Instead of gNB200 or CN20, for example, a dedicated center may be used. For example, in the first embodiment, the dedicated center deletes target data from the learning and inference data received from UE100-1 and transmits the deleted learning and inference data to UE100-2. Also, for example, in the second embodiment, the dedicated center encrypts the learning and inference data received from UE100-1 with the public key received from UE100-2 and transmits the encrypted learning and inference data to UE100-2. Furthermore, for example, in the second embodiment, the dedicated center transmits the learning and inference data encrypted by UE100-1 to UE100-2. Furthermore, for example, in the third embodiment, the dedicated center may send a program that performs machine learning to UE 100-1 and 100-2, encrypt the learning inference data, and send (or notify) the encrypted learning inference data to UE 100-1 and 100-2.

[0167] Although the first to third embodiments have been described above mainly in terms of supervised learning, the present invention is not limited to this. For example, the first to third embodiments may be applied to unsupervised learning or reinforcement learning.

[0168] Furthermore, a program (information processing program) that causes a computer to execute each process or function according to the above-described embodiment may be provided. Alternatively, a program (e.g., a mobile communication program) that causes the mobile communication system 1 to execute each process or function according to the above-described embodiment may be provided. The program may be recorded on a computer-readable medium. Using a computer-readable medium, it is possible to install the program on a computer. Here, the computer-readable medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, and may be, for example, a recording medium such as a CD-ROM or a DVD-ROM. Such a recording medium may be memory included in the UE 100 and the gNB 200.

[0169] The above-described operational flows are not limited to being implemented independently, but can also be implemented by combining two or more operational flows. For example, some steps of one operational flow may be added to another operational flow, or some steps of one operational flow may be replaced with some steps of another operational flow. In each flow, it is not necessary to execute all steps, and only some steps may be executed.

[0170] In the above-described embodiments and examples, an example in which the base station is an NR base station (gNB) has been described, but the base station may be an LTE base station (eNB) or a 6G base station. The base station may also be a relay node such as an IAB (Integrated Access and Backhaul) node. The base station may also be a DU of the IAB node. The UE 100 may also be an MT (Mobile Termination) of the IAB node.

[0171] Also, the term "network node" primarily refers to a base station, but may also refer to a device in the core network or part of a base station (CU, DU, or RU).

[0172] A program may be provided that causes a computer to execute each process performed by the UE 100 or the gNB 200. The program may be recorded on a computer-readable medium. Using a computer-readable medium, the program can be installed on a computer. Here, the computer-readable medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, and may be, for example, a recording medium such as a CD-ROM or a DVD-ROM. Furthermore, circuits that execute each process performed by the UE 100 or the gNB 200 may be integrated, and at least a portion of the UE 100 or the gNB 200 may be configured as a semiconductor integrated circuit (chip set, SoC).

[0173] As used in this disclosure, the terms "based on" and "depending on" do not mean "based only on" or "depending only on," unless expressly stated otherwise. The term "based on" means both "based only on" and "based at least in part on." Similarly, the term "depending on" means both "depending only on" and "depending at least in part on." Furthermore, the terms "include" and "comprise" do not mean including only the listed items, but may mean including only the listed items or may include additional items in addition to the listed items. Furthermore, the term "or" as used in this disclosure is not intended to mean an exclusive or. Furthermore, any reference to elements using designations such as "first," "second," etc., as used in this disclosure does not generally limit the quantity or order of those elements. These designations may be used herein as a convenient method of distinguishing between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed therein or that the first element must precede the second element in some way. In this disclosure, where articles are added by translation, such as a, an, and the in English, these articles shall include the plural unless the context clearly indicates otherwise.

[0174] Although the embodiments have been described in detail above with reference to the drawings, the specific configuration is not limited to the above, and various design changes can be made without departing from the spirit of the invention. Furthermore, it is also possible to combine the embodiments, operation examples, or processes within a consistent range.

[0175] This application claims priority from Japanese Patent Application No. 2022-170700 (filed October 25, 2022), the entire contents of which are incorporated herein by reference.

[0176] (Supplementary Note) (Supplementary Note 1) A communication method in a mobile communication system, comprising: a step in which a first user device transmits training data and / or inference data to a network device; a step in which the network device deletes at least one of security-targeted data and privacy-targeted data from the training data and / or the inference data; a step in which the network device transmits the training data after deletion and / or the inference data after deletion to a second user device; and a step in which the second user device performs machine learning using the training data after deletion and / or the inference data after deletion.

[0177] (Supplementary Note 2) The communication method according to Supplementary Note 1, further comprising a step in which the first user device or the second user device designates at least one of the security-targeted data to be deleted and the privacy-targeted data to be deleted.

[0178] (Supplementary Note 3) A communication method in a mobile communication system, comprising: a step in which a network device receives training data and / or inference data from a first user device; a step in which the network device requests a public key from a second user device; a step in which the second user device, in response to the request, transmits the public key created from a private key in a machine learning model to the network device; a step in which the network device encrypts the training data and / or the inference data using the public key and transmits the encrypted training data and / or the encrypted inference data to the second user device; and a step in which the machine learning model of the second user device decrypts the encrypted training data and / or the encrypted inference data using the private key and performs machine learning using the decrypted training data and / or the decrypted inference data.

[0179] (Supplementary Note 4) The communication method according to Supplementary Note 3, wherein the step of transmitting the public key includes a step in which a data collection unit of the machine learning model creates the public key from the private key.

[0180] (Supplementary Note 5) A communication method in a mobile communication system, comprising: a step in which a first machine learning model of a first user device encrypts training data and / or inference data using a common key; a step in which the first user device transmits the encrypted training data and / or the encrypted inference data to a network device; a step in which the network device transmits the encrypted training data and / or the encrypted inference data to a second user device; and a step in which a second machine learning model of the second user device decrypts the encrypted training data and / or the encrypted inference data using the common key, and performs machine learning using the decrypted training data and / or the decrypted inference data.

[0181] (Supplementary Note 6) The communication method described in Supplementary Note 5, wherein the encrypting step includes a step in which a data collection unit of the first machine learning model encrypts the training data and the inference data using the common key, and the decrypting step includes a step in which a data collection unit of the second machine learning model decrypts the encrypted training data and / or the encrypted inference data using the common key.

[0182] (Supplementary Note 7) A communication method in a mobile communication system, comprising: a step in which a network device transmits a program for performing machine learning to a first user device and a second user device; a step in which the first user device encrypts training data and / or inference data using a common key created from the program, and transmits the encrypted training data and / or the encrypted inference data to the network device; a step in which the network device transmits the encrypted training data and / or the encrypted inference data to the second user device; and a step in which the second user device decrypts the encrypted training data and / or the encrypted inference data using the common key created from the program, and performs machine learning using the decrypted training data and / or the decrypted inference data.

[0183] (Supplementary Note 8) The communication method according to any one of Supplementary Notes 1 to 7, wherein the network device is a base station or a core network device.

[0184] 1: Mobile communication system 20: CN 100 (100-1, 100-2, 100-3): UE 110 (110-1, 110-2): Receiving unit 120 (120-1, 120-2): Transmitting unit 130 (130-1, 130-2): Control unit 131-1, 131-2: CSI generating unit 132-1, 132-2: Optimal beam determining unit 133-1, 133-2: Position information generating unit 150-1, 150-2: GNSS receiver 200: gNB 210: Transmitting unit 220: Receiving unit 230: Control unit A1: Data collecting unit A2: Model learning unit A3: Model inference unit A4: Data processing unit

Claims

1. A communication method in a mobile communication system, comprising: a first user device transmitting learning data and / or inference data to a network node; the network node deleting at least one of security target data and privacy target data from the learning data and / or the inference data; the network node transmitting the learning data after deletion and / or the inference data after deletion to a second user device; the second user device performing machine learning using the learning data after deletion and / or the inference data after deletion. A communication method.

2. The communication method according to claim 1, further comprising: either the first user device or the second user device designating at least one of the security target data to be deleted and the privacy target data to be deleted.

3. A communication method in a mobile communication system, comprising: a network node receiving learning data and / or inference data from a first user device; the network node requesting a public key from a second user device; the second user device transmitting, in response to the request, the public key created from a secret key in a machine learning model to the network node; the network node encrypting the learning data and / or the inference data using the public key, and transmitting the encrypted learning data and / or the encrypted inference data to the second user device; the machine learning model of the second user device decrypting the encrypted learning data and / or the encrypted inference data using the secret key, and performing machine learning using the decrypted learning data and / or the decrypted inference data. A communication method.

4. The communication method according to claim 3, wherein transmitting the public key includes: a data collection unit of the machine learning model creating the public key from the secret key.

5. A communication method in a mobile communication system, comprising: a first machine learning model of a first user device encrypting learning data and / or inference data using a common key; the first user device transmitting the encrypted learning data and / or the encrypted inference data to a network node. The network node transmits the encrypted training data and / or the encrypted inference data to a second user device; A second machine learning model of the second user device decrypts the encrypted training data and / or the encrypted inference data using the common key, and performs machine learning using the decrypted training data and / or the decrypted inference data; A communication method.

6. The encrypting includes encrypting the training data and the inference data using the common key by a data collection unit of the first machine learning model; The decrypting includes decrypting the encrypted training data and / or the encrypted inference data using the common key by a data collection unit of the second machine learning model; The communication method according to claim 5.

7. A communication method in a mobile communication system, A network node transmits a program for executing machine learning to a first user device and a second user device; The first user device encrypts training data and / or inference data using a common key created from the program, and transmits the encrypted training data and / or the encrypted inference data to the network node; The network node transmits the encrypted training data and / or the encrypted inference data to the second user device; The second user device decrypts the encrypted training data and / or the encrypted inference data using the common key created from the program, and performs machine learning using the decrypted training data and / or the decrypted inference data; A communication method.

8. The network node is a base station or a core network node The communication method according to any one of claims 1 to 7.

9. A user device, A receiving unit that receives, from the network node, the training data after deletion and / or the inference data after deletion, after at least one of the security target data and the privacy target data has been deleted from the training data and / or the inference data transmitted by another user device to the network node; A control unit that performs machine learning using the training data after deletion and / or the inference data after deletion; A user device.

10. A user device, comprising: a transmission unit configured to transmit, in response to a request for a public key from a network node, the public key created from a private key in a machine learning model to the network node; a reception unit configured to encrypt learning data and / or inference data using the public key, and receive the encrypted learning data and / or the encrypted inference data from the network node; a control unit configured to cause the machine learning model to decrypt the encrypted learning data and / or the encrypted inference data using the private key, and perform machine learning using the decrypted learning data and / or the decrypted inference data;[[ / END]] the user device.

11. A user device, comprising: a reception unit configured to receive encrypted learning data and / or encrypted inference data from a network node; a control unit configured to cause the machine learning model to decrypt the encrypted learning data and / or the encrypted inference data using a common key, and perform machine learning using the decrypted learning data and / or the decrypted inference data; the user device.

12. A user device, comprising: a reception unit configured to receive, from a network node, a program for executing machine learning and encrypted learning data and / or encrypted inference data; a control unit configured to decrypt the encrypted learning data and / or the encrypted inference data using a common key created from the program, and perform machine learning using the decrypted learning data and / or the decrypted inference data; the user device.