Survival Time-aware Dynamic Multi-connectivity for Industrial Control Applications

Abstract

Achieving high reliability while maintaining low-latency is one key challenge to make wireless technologies suitable for critical applications in the industrial domain. In 5G, multi-connectivity (MC) in combination with packet duplication (PD) has been identified as a viable solution to meet the stringent reliability requirements. However, this approach can quickly lead to an inefficient usage of scarce radio resources. We present a dynamic PD model tailored for deterministic control applications with known survival times. Based on the risk of exceeding the survival time, our model controls redundancy efficiently on demand. We validate our model via system-level simulation and analyze the effectiveness of MC under correlated slow fading channels. The results reveal that with our proposed dynamic scheme, we are able to achieve outages very close to standard PD while only utilizing approximately 50% the radio resources in comparison.