An Analytical Framework for Latency-Reliability Optimization in Real-Time Wireless Systems

End-to-end latency-reliability (LR) analysis in real-time wireless systems should model multiple operations, such as compression and communication. Each operation contributes as a separate random variable to the composite latency, leading to a computationally-intensive optimization. We introduce a computationally-efficient framework to analyze LR optimization for uplink (UL) data transmission. The non-convex problem is transformed into convex optimization using Chebyshev’s and Cantelli’s Inequalities wherein UL latency is the sum of mean and scaled version of standard deviation. An algorithm is proposed to find the optimal system parameters. The proposed framework offers an excellent error performance for a reliability level of 0.9999 while reducing time complexity by 95%.