Joint Beamforming and Dynamic Relay Positioning for mmWave Urban Communications

TL;DR

This paper proposes a joint beamforming and relay positioning strategy for mmWave urban communications, using stochastic programming to optimize relay locations and beamforming weights for enhanced signal quality.

Contribution

It introduces a novel joint optimization framework for beamforming and predictive relay positioning in mmWave systems, leveraging stochastic programming for urban environments.

Findings

The proposed method outperforms randomized relay positioning policies.

Efficient distributed solution for relay placement and beamforming.

Significant improvement in signal-to-interference-plus-noise ratio.

Abstract

The use of millimeter wave (mmWave) spectrum for commercial wireless communications is expected to offer data rates in the order of Gigabits-per-second, thus able to support future applications such as Vehicle-to-Vehicle or Vehicle-to-Infrastructure communication. However, especially in urban settings, mmWave signal propagation is sensitive to blockage by surrounding objects, resulting in significant signal attenuation. One approach to mitigate the effect of attenuation is through multi-hop communication with the help of relays. Leveraging the unique characteristics of the mmWave medium, we consider a single-source/destination $2$-hop system, where a cluster of spatially distributed and reconfigurable relays is used to cooperatively amplify-and-forward the source signal to the destination. Our system evolves in time slots, during which not only are optimal beamforming weights centrally determined, but also future relay positions for the subsequent time slot are optimally selected, jointly maximizing the expected signal-to-interference+noise ratio at the destination. Optimal predictive relay positioning is achieved by formulating a 2-stage stochastic programming problem, which is efficiently approximated via a conditional sample-average-approximation surrogate, and solved in a purely distributed fashion across relays. The efficacy of the proposed near-optimal positioning policy is corroborated by comparison against a randomized relay positioning policy, clearly confirming its superiority.