This paper addresses the problem of channel estimation in multi-cellinterference-limited cellular networks. We consider systems employing multipleantennas and are interested in both the finite and large-scale antenna numberregimes (so-called "massive MIMO"). Such systems deal with the multi-cellinterference by way of per-cell beamforming applied at each base station.Channel estimation in such networks, which is known to be hampered by the pilotcontamination effect, constitute a major bottleneck for overall performance. Wepresent a novel approach which tackles this problem by enabling a low-ratecoordination between cells during the channel estimation phase itself. Thecoordination makes use of the additional second-order statistical informationabout the user channels, which are shown to offer a powerful way ofdiscriminating across interfering users with even strongly correlated pilotsequences. Importantly, we demonstrate analytically that in thelarge-number-of-antennas regime, the pilot contamination effect is made tovanish completely under certain conditions on the channel covariance. Gainsover the conventional channel estimation framework are confirmed by oursimulations for even small antenna array sizes.
展开▼
