ParEst: joint estimation of the OFDM channel state information in MIMO systems

Abstract

OFDM channel state information (CSI) is needed for determining key parameters in MIMO transmissions. In this paper, a novel CSI estimation method, ParEst, is proposed, which estimates the CSI from multiple transmitting antennas simultaneously. ParEst is based on a recent discovery that the CSI can be approximated very well by the linear combination of complex-based sinusoids on constant frequencies. ParEst finds the CSI of all antennas jointly by solving an optimization problem and achieves higher accuracy than existing heuristic methods. As the base sinusoids are on constant frequencies, ParEst pre-computes most key steps and reduces the run-time computation complexity to just a modest number of vector dot products equivalent to a few FFT calculations. ParEst can be applied to MIMO links in networks such as LTE, 5G, or Wi-Fi.