Joint Channel Parameter Estimation Using Evolutionary Algorithm

This paper proposes to utilise Evolutionary Algorithm (EA) to jointly estimate the Time of Arrival, Direction of Arrival, and amplitude of impinging waves in a mobile radio environment. The problem is presented as the joint Maximum Likelihood (ML) estimation of the channel parameters where typically, the high dimensional non-linear cost function is deemed to be too computationally expensive to be solved directly. Simulation results show that the proposed method is extremely robust to initialisation errors and low SNR environments, while at the same time it is also computationally more efficient than popular iterative ML methods i.e. the Space-Alternating Generalised Expectation-maximisation (SAGE) algorithm.