Pilot Reuse For Multi-Cell Multi-User Massive MIMO Transmission

To meet the constantly increasing demands for higher data rates, massive multi-input multi-output (MI-MO) technology has drawn considerable attention. By employing a large number of antennas at base stations to serve multiple user equipments (UEs) simultaneously, massive MIMO can improve spectral efficiency and link reliability substantially and all the effects of uncorrected noise and small scale fading vanish. In this thesis, pilot reuse for multi-cell multi-user massive MIMO transmission is investigated.Firstly, coordinated pilot reuse (CPR) for massive MIMO transmission is investigated. When discussing with multi-cell scenario, previous works often assume the pilot allocation scheme that, UEs in the same cell use a set of mutually orthogonal pilot sequences, while the same set of pilot sequences is reused among different cells. This scheme, however, is considered to be precious-resource-consuming. Since MIMO channels are not spatially isotropic, we propose CPR for multi-cell pilot allocation. The CPR allows pilots to be reused among intra-/inter-cell UEs and thus can significantly reduce pilot overhead and improve net spectral efficiency. We first obtain mean square error with uplink pilot reuse (MSE-UPR) of channel estimation. Pilot allocation scheme is decided on the principle of minimizing MSE-UPR. Since the complexity of exhaustive search (ES) is exponentially high, we give the condition under which MSE-UPR can be minimized. Basing on this condition, we propose simplified CPR algorithm to reduce complexity of allocating pilots. We also calculate uplink and downlink net spectral efficiency with the proposed CPR approach as well as the minimum mean square error (MMSE) uplink receiver and downlink precoder applied. Monte-Carlo simulations illustrate the significant performance improvement in pilot overhead and net spectral efficiency offered by the proposed CPR approach comparing with the traditional orthogonal pilot reuse (OPR) approach.Subsequently, we investigate MIMO transmission with Zadoff-Chu (ZC) based pilot reuse approach applied. We propose a noval pilot reuse approach taking advantage of the ideal autocorrelation and well cross-correlation of ZC. With this approach, we allocate pilots for each cell according to the main idea of CPR without coordination among cells. Pilot allocation results are only determined by UEs in the same cells. Simulation results are presented for the proposed ZC-based pilot reuse approach in comparison with CPR and OPR from aspects of complexity and net spectral efficiency based on Monte-Carlo method, which demonstrate that the ZC-based pilot reuse approach is an effective tradeoff of complexity and net spectral efficiency. Since pilot reuse might lead to additional pilot interference, we then develop statistically robust MMSE uplink receiver and downlink precoder with channel estimation errors taken into account. We also present simulation result to evaluate the proposed robust MMSE uplink receiver and downlink precoder. The result shows that, compared with traditional non-robust MMSE uplink receiver and downlink precoder, the robust ones can offer better gains especially in high signal-to-noise ratio region.