Joint Identification for Number of Transmit Antennas and Channel Order

Identifying the number of transmit antennas and the channel order is an important step in recovering the original information for a non-cooperative communication system. Traditional methods overestimate the number of transmit antennas and the channel order due to the coupling effect. This paper proposes a novel joint identification scheme to determine the number of transmit antennas and the channel order, which consists of two stages. Specifically, in the first stage, an improved weighted Gerschgorin disk identifier is developed to obtain the number of transmit antennas using frequency domain signals. In the second stage, we first derive an equivalent time-domain model to improve the utilization of the receiver antenna, and then a novel test statistic following
$F$
distribution is constructed to perform the sequential binary hypothesis testing. Finally, the channel order is obtained using the estimated number of transmit antennas in the first stage. The simulation results show that the proposed joint identification scheme significantly outperforms the existing approaches.