A Novel Receiver Design and Maximum-Likelihood Detection for Distributed MIMO Systems in Presence of Distributed Frequency Offsets and Timing Offsets

Abstract

In distributed multi-input multi-output (MIMO) systems, antennas may not be able to utilize common oscillators at the transmitter side or the receiver side. Consequently, there are multiple carrier frequency offsets (CFOs) and multiple timing offsets (TOs) due to distributed antennas that may significantly degrade the system performance if not properly addressed. In this paper, we analyze the effect of multiple CFOs and multiple TOs on receiver signal energy at the front end of each receive antenna pulse matched filter output in a distributed MIMO system. We then propose a novel receiver design for distributed MIMO systems that can accommodate multiple CFOs and multiple TOs. The proposed receiver structure utilizes a bank of pulse matched filters at each receive antenna where each filter accommodates one CFO specifically. Each filter output is then sampled at the symbol rate, with sampling timing selected according to the corresponding TO, followed by an information symbol detector. For the proposed receiver configuration, we derive the maximum-likelihood (ML) symbol detector for both space-time block coded and uncoded distributed MIMO systems. In addition, we design sub-optimal detectors and show that under certain conditions, the sub-optimal detectors perform close to the ML-optimal detector with significantly lower complexities. Extensive simulation studies illustrate our theoretical development for distributed MIMO systems with various transceiver configurations and channel conditions. The simulation results indicate that the proposed receiver structure together with the ML detection offers significant performance gains compared to the current state of the art.