Minimizing Real-time Estimation Error in IoT Networks via Noisy Past Observations

Age of information (AoI) has been widely used as a performance metric to characterize information freshness. In the Internet of Things (IoT) network, the mean-square error (MSE) is usually used to measure the estimation error and it can be expressed as a nonlinear function of AoI. However, in a practical scenario, status can be negatively affected by noise, which ultimately causes performance degradation on estimation. In this paper, we aim to address the issue of minimizing the maximum average noise-aware MSE in IoT monitoring systems under weak communication conditions, while also satisfying the minimum throughput requirements for each device. Then, an efficient algorithm is proposed to solve the optimization problem. Simulation results indicate that the proposed algorithm meets each device’s minimum throughput requirement, and by examining the effects of process structural characteristics and noise, it can achieve performance improvements in different communication environments, particularly in situations where channel resources are scarce and noise heterogeneity is high.