EWM-TOPSIS Based Weighted Graph Pilot Assignment for Cell-Free Massive MIMO Networks

In this paper, a novel weighted graph pilot assignment scheme based on entropy weight method (EWM) and technique for order preference by similarity to ideal solution (TOPSIS) is proposed for cell-free massive multiple-input multipleoutput (CF m-MIMO) networks with user heterogeneity. This is the first pilot assignment work in CF m-MIMO networks subject to multi-dimensional heterogeneity of user equipments (UEs), e.g., mobility speed, activation probability and serving priority. By employing EWM-TOPSIS to assess each UE's contribution to the total spectral efficiency (SE) based on its multi-dimensional heterogeneity, the proposed EWM-TOPSIS based weighted graph pilot assignment (ETWGPA) scheme enables pilot reuse among neighboring UEs with good channel conditions only, while assigning saved orthogonal pilots to neighboring UEs that contribute greatly to the total SE. By characterizing pilot contamination among UEs using multi-criteria of EWM-TOPSIS, i.e., multidimensional heterogeneity of UEs and number of common serving access points between UEs, the proposed ETWGPA scheme constructs a weighted interference graph and assigns the pilot with the least interference to each UE sequentially. Simulation results show that the proposed ETWGPA scheme outperforms the state-of-the-art schemes in terms of SE, energy efficiency, while ensuring higher UE fairness and lower complexity.