A Sum-Rate Prediction Strategy Based On RIS-Aided IoT Networks Power Optimization Algorithm

Reconfigurable intelligent surfaces (RISs), as ar-tificial electromagnetic structures with programmable electro-magnetic characteristics, have the potential to enhance the signals collected by service base stations in multi-cell Internet of things (IoT) networks. This capability improves the information extraction and utilization in communication systems, presenting numerous promising applications in the sixth-generation (6G) communication systems-IoT network environment. This paper proposes an optimization framework for the transmission sum-rate in a multicell, multi-user wireless communication system assisted by RIS. Firstly, establish transmission power allocation constraints for training and data symbols based on the distances of users from the transmitter, along with corresponding sum-rate objective functions. Utilize the multidimensional feature parameters of the RIS and long short-term memory (LSTM) to establish the relationship between user mobility distance and transmission power. Then, based on the obtained power prediction data, calculate the corresponding sum-rate using the previously defined objective functions. The obtained predictions contribute to more accurately assessing future communication environments, thereby aiding in optimizing the performance of communication systems. Simulation results validate the significant potential of LSTM in RIS-assisted multi-user wireless network systems.