Differentially Encoded Blind Multicell Multiuser Massive MIMO

In this paper, we consider the uplink of the narrowband multicell multiuser massive multiple-input multiple-output (MIMO) systems. In traditional approaches, orthogonal pilots among users of all (interfering) cells are utilized to estimate their channels assumed to be constant during channel coherence time. However, in multicell scenarios, such schemes suffer from pilot contamination. In order to address this issue, based on asymptotic pairwise orthogonality of the users' channels in massive regime, blind channel estimation schemes have been proposed. On the other hand, such schemes rely on low interference or high signal to noise ratio (SNR) levels, to achieve reliable performance. Inspired by the statistically noise-free 1-lag autocovariance matrix, we let the users' signal utilize a differential structure, by which, the effect of additive noise in such blind schemes is significantly reduced while preserving system robustness to out of cell interference. Numerical simulations are also used to verify the outperformance of the proposed scheme over the current pilot-based and blind schemes.