Pilot Assignment Based on Graph Coloring with Sum-Rate Maximization in Cell-Free Massive MIMO

Rongheng Chen,Hong Wang,Rongfang Song
DOI: https://doi.org/10.1109/iccc54389.2021.9674337
2021-12-10
Abstract:To mitigate the severe effect of pilot contamination, a novel graph coloring (NGC) based algorithm is applied to pilot assignment in cell-free massive multiple-input multiple-output (MIMO) systems. Specifically, firstly, the potential pilot contamination intensity among users is calculated by utilizing the large-scale fading between users and access points (APs). Then, the contamination threshold is set to perform the sparsification and an interference graph describing the user interference relationship is constructed. Finally, the proposed algorithm is used to implement pilot assignment in the interference graph. In addition, an iterative grid search (IGS) approach is adopted to update the threshold to ensure that the system downlink sum-rate is maximized. The numerical results reveal that the proposed NGC pilot assignment algorithm with optimized thresholds can remarkably improve the system performance compared with previous pilot assignment schemes.
What problem does this paper attempt to address?