Efficient resource allocation in 5G URLLC with packet duplication based macro-diversity

Abstract

Packet duplication based macro-diversity is a potential solution to achieve high reliability and low latency, for ultra-reliable low-latency communication (URLLC) applications, like mission critical internet of things (IoT), in 5G cellular systems. Increased radio resource usage is the price paid for the benefits accrued. This paper proposes a framework to find the relaxed per link reliability targets and to perform radio resource and power allocation over the links, which can guarantee the user equipment (UE) reliability target and also provide spectrum efficient resource usage, when multiple base stations (BSs) serve each UE. The resource allocation problem that involves BS selection, packet length adaptation, and power allocation, is formulated to minimize the overall resource usage within the constraints of reliability, latency, resource availability, and transmit power. Inherent system dynamics that exist among packet length, decoding error probability, and transmit power are used in devising a lightweight heuristic solution. The two important findings are that: (i) with each UE, a stringent reliability target for the best link and relaxed reliability targets for the other links can provide a spectrum efficient solution; (ii) additional resource savings can be achieved by allocating more power to the UEs with poor channel conditions, at each BS. The proposed method can achieve up to 20% resource saving when compared to the blind duplication with equal power allocation.