Method and apparatus for monitoring and reporting AI model in wireless communication system
The disclosure relates to a 6th generation (6G) communication system for optimizing a network while being applied in various industrial fields through connection between 5th generation (5G) or beyond 5G things and networks for supporting a higher data rate. A method performed by user equipment (UE) in a wireless communication system is provided, including receiving configuration information related to an artificial intelligence (AI) model from a base station; performing monitoring of a first AI model of the UE for encoding and decoding channel state information (CSI); and reporting a monitoring result to the base station. The first AI model includes a first encoder and a first decoder of the UE, and the first decoder may be related to a second AI model of the base station including a second decoder.
1 . A method performed by a user equipment (UE) in a wireless communication system, the method comprising:
receiving, from a base station, configuration information related to a first artificial intelligence (AI) model of the UE and a second AI model of the base station, wherein the first AI model includes a first encoder of the UE and a first decoder of the UE;
generating a plurality of channel state information (CSI) for monitoring the first AI model:
compressing the plurality of CSI based on the first encoder:
identifying whether at least one compressed CSI among the plurality of compressed CSI is a monitoring target, based on the configuration information:
if the at least one compressed CSI is identified as the monitoring target, restoring the at least one compressed CSI based on the first decoder; and
reporting, to the base station, a monitoring result of the first AI model based on comparing the at least one restored CSI and at least one corresponding CSI among the plurality of CSI,
wherein the first decoder is related to the second AI model of the base station including a second decoder of the base station.
2 . The method of claim 1 , wherein the first AI model includes a first autoencoder and the second AI model includes a second autoencoder.
3 . The method of claim 1 , wherein the configuration information comprises at least one of an identifier of the first AI model, an interval for monitoring the first AI model, a method for monitoring the first AI model, or information on a type of reporting for the monitoring result of the first AI model.
4 . The method of claim 1 , wherein comparing the measured at least one CSI and the restored at least one CSI comprises:
identifying whether a difference between the measured at least one CSI and the restored at least one CSI is less than a first threshold value in the configuration information.
5 . The method of claim 4 , further comprising:
receiving, from the base station, a message requesting a change into a non-AI-based CSI feedback mode, in response to transmitting, during a third time period,
wherein the monitoring result is transmitted when accuracy of the first AI model of the monitoring result is less than the first threshold value.
6 . The method of claim 1 , wherein reporting the monitoring result of the first AI model to the base station comprises at least one of:
reporting the monitoring result of the first AI model together with a report of CSI compressed by the first encoder;
reporting the monitoring result of the first AI model when a specific event occurs; or
periodically reporting the monitoring result of the first AI model, and
wherein the monitoring result of the first AI model comprises accuracy of the first AI model and an identifier of the first AI model.
7 . The method of claim 1 , further comprising:
transmitting, to the base station, a message indicating a failure of the first AI model when accuracy of the first AI model is less than a second threshold value during a first time period.
8 . The method of claim 7 , further comprising:
transmitting, to the base station, a message indicating a change into a non-AI-based CSI feedback mode when the accuracy of the first AI model is less than a third threshold value during a second time period.
9 . The method of claim 8 , further comprising:
transmitting, to the base station, a message indicating a change into an AI-based CSI feedback mode when the UE operates in the non-AI-based CSI feedback mode, and when accuracy of the first AI model is greater than the third threshold value during the second time period.
10 . The method of claim 1 ,
wherein the configuration information comprises information on a weight value and a structure of the second AI model.
11 . A user equipment (UE) in a wireless communication system, the UE comprising:
at least one transceiver;
at least one processor communicatively coupled to the at least one transceiver; and
at least one memory, communicatively coupled to the at least one processor, storing instructions executable by the at least one processor individually or in any combination to cause the UE to:
receive, from a base station, configuration information related to a first artificial intelligence (AI) model of the UE and a second AI model of the base station, wherein the first AI model includes a first encoder of the UE and a first decoder of the UE,
generate a plurality of channel state information (CSI) for monitoring the first AI model,
compress the plurality of CSI based on the first encoder,
identify whether at least one compressed CSI among the plurality of compressed CSI is a monitoring target, based on the configuration information,
if the at least one compressed CSI is identified as the monitoring target, restore the at least one compressed CSI based on the first decoder, and
report, to the base station, a monitoring result of the first AI model based on comparing the at least one restored CSI and at least one corresponding CSI among the plurality of CSI,
wherein the first decoder is related to the second AI model of the base station including a second decoder of the base station.
12 . The UE of claim 11 , wherein the first AI model includes a first autoencoder and the second AI model includes a second autoencoder.
13 . The UE of claim 11 , wherein the configuration information comprises at least one of an identifier of the first AI model, an interval for monitoring the first AI model, a method for monitoring the first AI model, or information on a type of reporting for the monitoring result of the first AI model.
14 . A method performed by a base station in a wireless communication system, the method comprising:
transmitting, to a user equipment (UE), configuration information related to a first artificial intelligence (AI) model of the UE and a second AI model of the base station, wherein the first AI model includes a first encoder of the UE and a first decoder of the UE; and
receiving, from the UE, a monitoring result,
wherein the monitoring result is based on a comparison between at least one restored channel state information (CSI) and at least one corresponding CSI, the at least one restored CSI being restored from at least one compressed CSI which corresponds to a monitoring target,
wherein the at least one compressed CSI is compressed from the at least one corresponding CSI among a plurality of CSI of the UE,
wherein the first decoder is related to a second decoder included in the second AI model of the base station, and
wherein the monitoring result of the first AI model is based on a comparison between a first output from the first encoder and a second output from the first decoder.
15 . The method of claim 14 , wherein the first AI model includes a first autoencoder and the second AI model includes a second autoencoder.
16 . The method of claim 14 , wherein the configuration information comprises at least one of an identifier of the first AI model, an interval for monitoring the first AI model, a method for monitoring the first AI model, or information on a type of reporting for the monitoring result of the first AI model.
17 . A base station in a wireless communication system, the base station comprising:
at least one transceiver;
at least one processor communicatively coupled to the at least one transceiver; and
at least one memory, communicatively coupled to the at least one processor, storing instructions executable by the at least one processor individually or in any combination to cause the base station to:
transmit, to a user equipment (UE), configuration information related to a first artificial intelligence (AI) model of the UE and a second AI model of the base station, wherein the first AI model includes a first encoder of the UE and a first decoder of the UE; and
receive, from the UE, a monitoring result,
wherein the monitoring result is based on a comparison between at least one restored channel state information (CSI) and at least one corresponding CSI, the at least one restored CSI being restored from at least one compressed CSI which corresponds to a monitoring target,
wherein the at least one compressed CSI is compressed from the at least one corresponding CSI among a plurality of CSI of the UE,
wherein the first decoder is related to a second decoder included in the second AI model of the base station, and
wherein the monitoring result of the first AI model is based on a comparison between a first output from the first encoder and a second output from the first decoder.
18 . The base station of claim 17 , wherein the first AI model includes a first autoencoder and the second AI model includes a second autoencoder.
19 . The base station of claim 17 , wherein the configuration information comprises at least one of an identifier of the first AI model, an interval for monitoring the first AI model, a method for monitoring the first AI model, or information on a type of reporting for the monitoring result of the first AI model.