Reliability Analysis and Optimization for Sensing Data Collection and Processing in New Energy Internet Systems

Abstract

In this paper, we consider a sensing network in the future energy internet systems. The network has a processing server collecting sensing data from multiple sensors and processing them together. The whole framework of data collection and processing is required to satisfy a low overall latency and high- reliability. Therefore, the data collection phase is short and the transmission in such phase is carried by finite blocklength codes, while the data processing should be finished within the remaining time duration (taking out the spent time for data collection from the overall latency constraint/budget). We characterize the overall reliability of data collecting and processing. Following the characterization, we provide an optimal framework design of time-allocation aiming to minimize the instantaneous overall error probability. In particular, we prove that the proposed framework is jointly convex in the lengths of data collection and data processing phases. Via simulations, we validate the convexity proven in our analytical and show the performance advantage of proposed analytical solution.