Distributed Pilot Assignment for UAV-Enabled Cell-Free and Small Cell Systems

This paper addresses the critical pilot contamination problem in the uplink of UAV-enabled wireless networks, covering both cell-free (CF) and small cell (SC) architectures. In these systems, UAV-enabled communication relays (UCRs) collaboratively serve users without fixed terrestrial infrastructure; CF has all UCRs jointly serving all users, while SC has users served by the strongest gain UCR. The reuse of limited orthogonal pilot sequences severely degrades channel estimation accuracy, especially in dynamic UCR environments. To overcome this, we propose a pilot assignment for master UCRs (PAMU), a novel decentralized strategy where the UCR with the strongest connection to the user locally assigns the pilot. This effectively minimizes pilot interference on dominant links and significantly reduces the central unit computational load. We benchmark the performance of PAMU against random pilot assignment and the spatial pilot reassignment (SPR) algorithm, evaluating it with and without power control optimization through extensive simulations across various network settings, including different numbers of UCRs, users, pilot sequence lengths, and UCR altitudes. Results show PAMU dramatically improves average user throughput and substantially reduces pilot contamination in both CF and SC systems. While SPR is more effective for enhancing the minimum user rate, and random assignment consistently yields the lowest performance, PAMU delivers strong overall performance, proving to be a practical and efficient decentralized pilot control solution for user-centric wireless networks, offering valuable insights for future aerial communication system design.