Abstract: The present disclosure provides a method for one user equipment (UE) to perform federated learning with a plurality of UEs in a wireless communication system. More specifically, the method performed by the one UE comprises receiving, from a server, a channel state information reference signal (CSI-RS); transmitting, to the server, channel state information (CSI) calculated based on the CSI-RS; receiving, from the server, compression state information for determining a weight compression method of the one UE based on (i) information on a global parameter for the federated learning and (ii) channel state information of each of channels between the server and the plurality of UEs; determining the weight compression method based on (i) a difference between the global parameter and a global parameter received before a reception of the global parameter and (ii) the compression state information; and transmitting, to the server, a local parameter updated based on the determined weight compression method.

Abstract: A sender may: decompose, on the basis of a plurality of semantic elements, data into semantic element values respectively corresponding to the plurality of semantic elements; perform a first transmission including some semantic element values from among the semantic element values; and on the basis of receiving a retransmission request for the first transmission from a receiver, perform a retransmission including at least one semantic element value other than the semantic element values included in the first transmission, from among the plurality of semantic element values.

Abstract: A transmission device in a wireless communication system: determines an encoder having S outputs from among encoders supported by the transmission device, on the basis of the number B of feedback bits, wherein B=S×Q; encodes CSI through the encoder to output S real number values; determines B-bit encoded CSI including Q bits indicating the S real number values, respectively; and transmits the B-bit encoded CSI. The encoders may have different numbers of outputs from each other, and the different numbers of outputs may be predetermined for different feedback bit number ranges.

Abstract: The present disclosure provides a method for reporting, by a terminal, channel state information (CSI) in a wireless communication system. More specifically, the method includes: receiving, from a base station, a pilot signal related to calculation of a quantization rule, in which the quantization rule is determined based on an empirical distribution of an encoder neural network output of the terminal; transmitting, to the base station, quantization rule information related to the quantization rule calculated based on the pilot signal; and receiving, from the base station, information on a gradient calculated based on the quantization rule information, and the quantization rule information includes information on an empirically calculated variance with respect to the empirical distribution of the encoder neural network output.

Abstract: Proposed are a method for performing federated learning in a wireless communication system, and an apparatus therefor. The method performed by a terminal may comprise the steps of: generating a Q-ary code including i) a restriction-based system part and ii) a parity part by coding restriction-based Q-ary information; determining, on the basis of a preset method, the number T of transmissions of the parity part of the Q-ary code; on the basis of specific channel information from among channel information between a plurality of terminals and a base station participating in the federated learning, allocating power to the system part and T parity parts; and transmitting, on the basis of the allocated power, the system part and the T parity parts to the base station.

Abstract: The present disclosure provides a device and method for performing, based on channel information, a device grouping for federated learning based AirCOMP of a non-IID data environment in a communication system. The present disclosure also provides a device and method for performing effective federated learning in a non-IID environment including multiple devices. The present disclosure also provides a device and method for performing a device grouping in consideration of channel environment factors in order to apply AirComp based federated learning to a real communication environment.

Abstract: The present specification provides a method by which a plurality of terminals perform federated learning in a wireless communication system. More particularly, the method performed by one terminal comprises the steps of: receiving, from a base station, a first downlink signal for requesting information regarding learning data for the federated learning, which is used by the one terminal; transmitting, to the base station, the information regarding the learning data on the basis of a type of learning data, the information regarding the learning data being information related to the distribution of the learning data used by the one terminal; receiving, from the base station, a second downlink signal including parameter information regarding a parameter related to a configuration for performing the federated learning, which has been determined on the basis of the learning data; and performing the federated learning on the basis of the information regarding the parameter.

Abstract: Disclosed herein is a method of operating a terminal in a wireless communication system. As an example of the present disclosure, the method may include: receiving information on at least one of a first split point and a second split point from a base station; performing first split inference based on the first split point and generating first intermediate data; transmitting the first intermediate data to the base station; receiving second intermediate data generated based on the second split point from the base station; performing remaining split inference from the second split point based on the second intermediate data; and adjusting the second split point based on the second intermediate data.

Abstract: In order to provide device and method for adjusting split point in wireless communication system, a method of operating a terminal in a wireless communication system may comprise requesting a base station for uplink channel quality information, receiving, from the base station, the uplink channel quality information, determining a first split point based on the channel quality information, performing first split inference based on the determined first split point and transmitting, to the base station, first split inference result information.

Abstract: An intelligent computing device according to an embodiment of the present specification may receive a request for terminal information related to a terminal from a base station, transmit the terminal information to the base station, receive a resource for data transmission or reception, which is allocated by the base station, and transmit or receive data to or from the base station by using the allocated resource, wherein the data is transmitted or received by using a receiver selected from among a normal receiver of the terminal and an artificial intelligence receiver of the terminal on the basis of wireless communication environment information of the terminal, so that the conventional normal receiver and the artificial intelligence receiver provided at the terminal can be efficiently used in a wireless communication environment.

Abstract: A method for performing reinforcement learning by a V2X communication device in an autonomous driving system, specifically, a method for performing reinforcement learning in consideration of a reward application ratio over time, is proposed. Action information is transmitted to a second V2X communication device, reward information is received from the second V2X communication device, and reinforcement learning is performed on the basis of a reward, wherein a reward corresponding to a ratio determined by a first V2X communication device is applied to the reinforcement learning, the ratio is determined on the basis of a time interval from a time point of transmission of the action information to a time point of reception of the reward information, and the ratio is between 0 and 1, both inclusive.

Abstract: A method for performing reinforcement learning by a V2X communication device in an autonomous driving system, specifically, a method for performing reinforcement learning in consideration of an application rate of a reward according to age in terms of the freshness of a reward for an action, is proposed. An agent transmits an action message and controls a reflection rate of a reward through AoI management for a reward message, so that rewards transmitted from a plurality of devices are suitably reflected in an environment of a reinforcement learning-based autonomous driving system, and an optimal policy can be found accordingly.

Abstract: Disclosed is a method for controlling, by a terminal, an operation of a deep neural network in a wireless communication system. The method according to an embodiment of the present disclosure receives a downlink from abase station in a wireless communication system; and preprocesses the downlink on the basis of the result of an operation of a deep neural network of a terminal, wherein at least one reference signal is applied to the downlink while a statistical feature related to noise of the downlink are maintained. The terminal of the present disclosure may be linked to an artificial intelligence module, a drone (unmanned aerial vehicle (UAV)), a robot, an augmented reality (AR) device, a virtual reality (VR) device, a device related to 6G services, and the like.

Abstract: Disclosed is a method by which a terminal controls the calculations of a deep neural network in a wireless communication system. A method according to one embodiment of the present disclosure receives a downlink from a base station of a wireless communication system by using a multi input multi output (MIMO) operation, pre-processes the downlink on the basis of a result of calculations of a deep neural network in a terminal, acquires the number of a plurality of transmission antennas connected to the terminal, acquires the number of a plurality of reception antennas connected to the terminal, and forms a plurality of overlapping neural networks overlapping in the deep neural network, on the basis of a preset number of reference antennas, the number of transmission antennas, and the number of reception antennas.

Abstract: Disclosed is a method for transmitting and receiving data by a terminal in a wireless communication system according to the present document, the method comprising the steps of: receiving configuration information from a base station; and receiving data, on the basis of the configuration information. The configuration information includes information about a first receiver model and a first zone to which the base station belongs, the first zone includes one or more base stations adjacent to the base station, and the first receiver model is generated on the basis of a combination of receiver models that are trained by the base stations within the first zone, according to a machine learning step.

Abstract: Disclosed is a wireless communication method using an on-device learning-based machine learning network. A method according to one embodiment of the present specification comprises the steps of: generating a learning list on the basis of one or more location samples collected while a terminal communicates through a network; when an event associated with any one of one or more first items included in the learning list is detected, obtaining training data for updating, to a second machine learning network, a first machine learning network to be applied to a receiver of the terminal; and updating the first machine learning network by using the training data, and thereby the method provides a wireless communication service adaptive to an actual communication environment.

Abstract: The present disclosure relates to a method for operating a terminal and a base station in a wireless communication system and an apparatus for supporting the same. In an embodiment of the present disclosure, a method for operating a terminal in a wireless communication system may include: transmitting a first message including information related to learning; receiving a second message including configuration information for learning; transmitting an uplink reference signal; and transmitting channel information related to a downlink channel measured based on a downlink reference signal.

Abstract: Provided is a method for transmitting or receiving data, by a user equipment (UE), to or from a base station (BS). The method includes transmitting, by the UE, capability information of the UE to the BS, wherein the capability information includes information related to artificial intelligence (AI) calculation for the data transmission or reception, receiving, by the UE, at least one of a plurality of AI parameters from the BS, and applying the at least one AI parameter to an encoding process for the data transmission or a decoding process for the data reception, wherein the encoding process or the decoding process is performed by information on a network structure in the at least one AI parameter, and wherein the at least one AI parameter comprises a plurality of information for performing the encoding process or the decoding process by the network structure.

Abstract: Provided is a method for transmitting or receiving data, by a user equipment (UE), to or from a base station (BS). The method includes transmitting, by the UE, capability information of the UE to the BS, wherein the capability information includes information related to artificial intelligence (AI) calculation for the data transmission or reception, receiving, by the UE, at least one of a plurality of AI parameters from the BS, and applying the at least one AI parameter to an encoding process for the data transmission or a decoding process for the data reception, wherein the encoding process or the decoding process is performed by information on a network structure in the at least one AI parameter, and wherein the at least one AI parameter comprises a plurality of information for performing the encoding process or the decoding process by the network structure.