Reconfigurable Intelligent Surface-Assisted mmWave multi-UAV Wireless Cellular Networks

TL;DR

This paper proposes a joint optimization framework for RIS-assisted mmWave UAV cellular networks, improving capacity and reliability through deployment, scheduling, beamforming, and phase optimization, addressing mobility and blockage issues.

Contribution

It introduces an iterative optimization method combining deployment, scheduling, beamforming, and RIS phase design for UAV networks, utilizing sphere search and sBnB algorithms.

Findings

Significant sum-rate improvements over non-optimized systems

RIS effectively compensates for throughput loss due to blockage

Joint optimization enhances network performance and reliability

Abstract

Unmanned aerial vehicles (UAVs) have brought a lot of flexibility in the network deployment. However, UAVs suffer from the high mobility and instability. To improve the capacity and reliability of the UAV networks, millimeter-wave (mmWave) and reconfigurable intelligent surfaces (RISs) can be used in the system. In this paper, we consider an RIS-assisted mmWave UAV wireless cellular network, where UAVs serve several users with the help of multiple RISs. We jointly optimize the deployment, user scheduling, beamforming vector, and RIS phases to maximize the sum-rate, with the constraints of the minimum rate, the UAV movement, the analog beamforming, and the RIS phases. To solve this complex problem, we use an iterative method, in which when we optimize one variable, we fix the other three variables. When optimizing the deployment, we find the optimal position for the UAV by a sphere search. Then, we formulate a mixed-integer non-linear problem (MINLP) to find the best scheduling. A spatial branch-and-bound (sBnB) method is used to solve the MINLP. When Optimizing the beamforming vector and the RIS phases, we propose an iterative algorithm that relies on the equivalence between the maximization of the sum-rate and the minimization of the summation of weighted mean-square errors (sum-WMMSE). The majority-minimization method is used to deal with the constant-modulus constraints for the analog beamforming and RIS phases. The proposed joint optimization offers significant advantages over the system without beamforming and RIS phase optimization and the system without deployment optimization. In addition, the RIS can compensate for the loss of throughput due to the blockage, especially in low flight altitudes.