Low complexity semi-blind channel estimation algorithm in two-way relay networks

In this work, we propose a low complexity semiblind channel estimation algorithm, referred to as the variance squared maximum likelihood (VSML) estimator, which employs only one training symbol in each channel estimation, to estimate general non-reciprocal flat-fading channels in amplify-and-forward (AF) two-way relay networks (TWRNs). We formulate a non-convex objective function and obtain closed-form channel estimates by minimizing its approximate expression. Theoretical analysis proves that the derived channel estimation is asymptotically optimal in large sample size scenarios. Monte-Carlo simulation results show that the VSML estimator outperforms the existing relaxed maximum likelihood (RML) estimator in terms of mean squared error (MSE) performance and remarkably reduces the computational complexity by completely avoiding the grid-search algorithm under M-ary phase-shift-keying (MPSK) modulation.