A Novel Way for Pilot Decontamination in Massive MIMO Multi-Cell Systems

Abstract

The performance of massive Multi-input Multi-output (MIMO) system is limited by pilot contamination, due to the pilots with similar energy can be received by users at the same time, which brings the difficulty of determining the domain pilot. An effective way to solve this problem in Time Division Duplex (TDD) wireless system is using weighted graph for pilot allocation. However, this method is badly influenced by the number of pilots. In this paper, weighted graph is combined with Soft Pilot Reuse Scheme (SPRS) and cell classification. In order to deal with the influence of too much pilots, SPRS and cell classification will be presented. Program of cell classification will divide cells into two groups and prepare orthogonal pilots to make two groups of cells be orthogonal to each other. For whole users in one cell, they will be divided into edge users and center users by large-scale-fading factor. For center users, edge-weighted interference graph (EWIG) will measure their degrees of affection by pilot contamination and form the weighted graph. After that, the graph will be used as reference to assign pilot. As for edge users in the same category, they will use orthogonal pilots to reduce the pilot contamination. The simulation gives powerful results of the interference rejection and the achievable rate enhancement.