Pilot Contamination Mitigation for Wideband Massive MIMO Systems

This paper proposes a novel joint channel estimation and beamforming approach for multicell wideband massive multiple input multiple output (MIMO) systems. With the proposed channel estimation and beamforming approach, we determine the number of cells <tex-math notation="LaTeX">Nc </tex-math> that can utilize the same time and frequency resource while mitigating the effect of pilot contamination. The proposed approach exploits the multipath characteristics of wideband channels. Specifically, when the channel has a maximum of <tex-math notation="LaTeX">L </tex-math> uncorrelated multipath taps (or correlated multipath taps satisfying modest criteria which is valid in most practical scenarios), it is shown that <tex-math notation="LaTeX">Nc=L </tex-math> cells can estimate the channels of their user equipments (UEs) and perform beamforming while mitigating the effect of pilot contamination. The proposed approach can also be applied for general correlated multipath taps, and achieves good performance for this scenario as well. In a typical long term evolution (LTE) channel environment having delay spread <tex-math notation="LaTeX">Td=4.69\,\,μ s </tex-math> and channel bandwidth <tex-math notation="LaTeX">$B=5$ </tex-math> MHz, we have found that <tex-math notation="LaTeX">$L=36$ </tex-math> cells can use this band. In practice, <tex-math notation="LaTeX">Td </tex-math> is constant for a particular environment and carrier frequency, and hence <tex-math notation="LaTeX">L </tex-math> increases as the bandwidth increases. All the analytical expressions have been validated, and the superiority of the proposed design over the existing ones is demonstrated using extensive numerical simulations both for correlated and uncorrelated channels. The proposed channel estimation and beamforming design is linear and simple to implement.