Optimized Survival Mode to Guarantee QoS for Time-critical Services

We consider time critical transmissions for wireless industrial IoT services, where information should not be older than a survival time in the millisecond order, and extreme service availability is targeted. With non-zero survival time, a feedback based survival mode is possible, where the network incorporates additional robustness if the previous transmission was not received. We consider the tradeoff between reliability of normal mode and survival mode transmission, and analyze the effect of feedback errors on these. We formulate an optimization to minimize average resource usage, guaranteeing time critical quality of service (QoS), and show numerical results in scenarios with typical system assumptions. For a user with a Rayleigh fading channel, average signal-to-noise ratio 10 dB, and control channel error probability 10
⁻³
, optimization based on judiciously choosing survival and normal mode reliability targets can reduce resource usage needed to guarantee QoS with a factor of 30.