Ultra-Reliable Low-Latency Communication for Aerial Vehicles via Multi-Connectivity

TL;DR

This paper evaluates multi-connectivity strategies for aerial vehicle communication to meet ultra-reliable low-latency requirements, highlighting the limitations of single connectivity and existing cellular networks in supporting AVs.

Contribution

It investigates the effectiveness of multi-connectivity options, including air-to-ground, air-to-air, and high altitude platforms, for URLLC in AV communication under finite blocklength constraints.

Findings

Single connectivity struggles due to LoS interference.

Existing cellular networks cannot meet URLLC needs for AVs.

Multi-connectivity significantly improves reliability and latency.

Abstract

Aerial vehicles (AVs) such as electric vertical take-off and landing (eVTOL) make aerial passenger transportation a reality in urban environments. However, their communication connectivity is still under research to realize their safe and full-scale operation, which requires stringent end-to-end (E2E) reliability and delay. In this paper, we evaluate reliability and delay for the downlink communication of AVs, i.e., remote piloting, control/telemetry traffic of AVs. We investigate direct air-to-ground (DA2G) and air-to-air (A2A) communication technologies, along with high altitude platforms (HAPs) to explore the conditions of how multi-connectivity (MC) options satisfy the demanding E2E connectivity requirements under backhaul link bottleneck. Our considered use case is ultra-reliable low-latency communication (URLLC) under the finite blocklength (FBL) regime due to the nature of downlink control communication to AVs. In our numerical study, we find that providing requirements by single connectivity to AVs is very challenging due to the line-of-sight (LoS) interference and reduced gains of downtilt ground base station (BS) antenna. We also find that even with very efficient interference mitigation, existing cellular networks designed for terrestrial users are not capable of meeting the URLLC requirements calling for MC solutions.