A Genetic Algorithm based Pilot Assignment strategy for Cell-Free massive MIMO system

In this work, two pilot assignment schemes are proposed to mitigate pilot contamination and improve the performance of the downlink cell-free massive MIMO system. A genetic algorithm is applied to determine the pilot sequences of all users, thereby maximizing their average signal-to-interference-plus-noise ratio (SINR). K-medoids clustering algorithm exploiting the spreading characteristics of Access Points (AP) in the system is combined with the considered genetic algorithm for faster convergence. Both algorithms show superior performance about the average downlink rate of all users compared to the traditional greedy algorithms.