Improving reliability of replicated message delivery in cellular machine-type communications

Abstract

In this paper, we introduce a novel approach for cellular machine-type communications (MTC) based on random access transmissions. It targets a substantial increase in the probability of message delivery at the first transmission attempt. This may be achieved by sending multiple message replicas across a set of transmission opportunities (in time, code, or frequency), which are shared between all the potential MTC devices. For the considered class of access algorithms, we propose the optimal scheme with centralized control that delivers the maximum success probability. This developed approach together with the conventional baseline method (with a single replica) is analyzed in the meta-stable state, and a performance estimate is obtained. Our numerical results - supported by both analysis and simulations - confirm a significant improvement in terms of the probability of immediate message delivery comparing to the conventional transmission, as well as indicate a potential gap with respect to the derived upper bound.