Joint Trajectory and Pilot Assignment Optimization for UAV Enabled Cell-Free Massive MIMO

Pilot assignment is a major issue in unmanned aerial vehicle (UAV) enabled user centric (UC) cell- free massive multiple input and multiple output (UC-CF-mMIMO) wireless networks where, UAVs function as distributed mobile aerial access points (MAAPs). In this paper, an efficient Tabu-search (TS) based pilot assignment scheme is proposed to address this issue where, a penalty based objective function maintains a minimum distance between mobile units (MUs). This scheme utilizes a local neighborhood search, where we first define the neighborhood structure and set the objective function. The proposed algorithm iteratively identifies a near-optimal solution with low computational overhead and achieves high average channel estimate power. We implemented a gradient based (GB) algorithm to optimize the 3D locations of MAAPs and find the signal-interference-plus-noise ratio (SINR) expression. Numerical simulations show that proposed TS-based strategy and MAAP trajectory optimization significantly enhance the per-user throughput with 95%-likelihood. The results achieved with proposed method are better than the conventional random pilot assignment schemes such as, gradient based and gibbs sampling (GB-GS), user centric perfect channel state information(UC-PSCI) and mMIMO based PCSI method while maintaining similar complexity.