Cooperative Beamforming With Predictive Relay Selection for Urban mmWave Communications

While millimeter wave (mmWave) communications promise high data rates, their sensitivity to blockage and severe signal attenuation present challenges in their deployment in urban settings. One effective way to mitigate those effects and increase communication range is relaying in combination with beamforming. However, relay-assisted beamforming also requires relay selection, a process that not only is resource demanding due to the need for explicit Channel State Information (CSI), but also introduces network latency. In this paper, we introduce a relay beamforming approach for mmWave communications in an urban scenario, which exploits the shadowing-induced correlation structure of the channel to reduce both latency and CSI estimation overhead. Our system model consists of static relays deployed in clusters across streets. Under that setting, we propose a novel, resource efficient scheme for joint optimal relay selection and distributed cooperative beamforming that maximizes the expected Signal-to-Interference plus Noise Ratio (SINR) at the destination, under power constraints. In particular, assuming a time-slotted system operation, one relay from each cluster is optimally selected at each time slot to participate in optimal beamforming at the next time slot. The key novelty of the proposed scheme is that relay selection is implemented in a predictive and distributed manner, by exploiting channel correlations and by using past and present measurements of magnitude CSI. As a result, at each time slot, optimal beamforming based on relays selected in the previous slot and optimal predictive relay selection for the next time slot are implemented completely in parallel. This parallelization eliminates delays induced by sequential execution of relay selection and beamforming, and substantially reduces CSI estimation overhead. Simulations confirm that the proposed relay selection scheme outperforms any randomized selection policy, while, at the same time, achieves comparable performance to an ideal selection scheme that relies on perfect CSI estimates for all candidate relays.