Performance Evaluation of Non-coherent DPSK Signal Detection Algorithms in Massive MIMO Systems

Abstract

In the 5G technology, massive MIMO is a very important share, where the base station is equipped with huge number of antennas. Massive MIMO can significantly improve signal strength, channel rate, power consumption, and throughput. However, there are major signal detection challenges that face the system such as the complexity of channel estimation. Most of the commonly used detection techniques in massive MIMO are coherent detection, which requires complex channel estimation techniques. In this paper, non-coherent detection is used to avoid the complexity of channel estimation in massive MIMO systems by using differential phase shift keying modulation (DPSK). The performance of Differential Detection (DD), sorted Decision Feedback Differential Detection (DFDD), Multiple Symbol Differential Detection (MSDD) and Multiple Symbol Differential Sphere Detection (MSDSD) algorithms for massive MIMO are studied, implemented and compared together along with coherent Minimum Mean Square Error (MMSE) for the same studied massive MIMO system model. Results showed that, MSDSD outperforms DFDD with trade off complexity and time. Nonetheless for channels with high coherence time, the transmission burst length is increased for better performance. However, the complexity and time needed for MSDSD is higher than DFDD.