Blind estimation of multi-CFO for distributed MIMO-OFDM system in frequency-selective fading channels

A novel blind multiple carrier frequency offset (multi-CFO) estimation algorithm is developed for distributed multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) systems over frequency-selective fading channels. In the algorithm, different multiple CFOs are estimated by cyclostationarity (CS) of the received signals and least square (LS) criterion. Theory analysis shows that the blind algorithm adapts to synchronization under any distributed stationary noise. Meanwhile, it has no special limitation on the number of transmit and receive antennas. Simulation results verify that the proposed algorithm maintains stability at low signal-to-noise ratio (SNR). It can also achieve excellent performance in single-input singleoutput (SISO) OFDM systems.

I. Introduction

Recently, multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) has gained increased interest because of its high data rate and its ability to combat fading^[1]. However, the performance of the MIMO-OFDM system critically depends on the accurate estimation of the carrier frequency offset (CFO). MIMO systems may be categorized into the centralized and the distributed system. For the centralized system, the antennas of transmitter and receiver are placed near to each other which make the same CFO exists between different transmit and receive antenna. Compared with the centralized MIMO system, the distributed system can achieve a higher diversity gain, which is a promising solution for future MIMO system. However, the transmitters and the receivers in a distributed system are sparsely placed. The oscillator mismatch and Doppler shift are often different between each pair of receive and transmit antenna, which make different CFOs exist between each pair of transmit and receive antenna and the carrier synchronization of distributed system much more complex than that of centralized system and traditional single-input single-output (SISO) system. Hence, effective carrier synchronization algorithms are more important to distributed MIMO-OFDM system.