Estimating Age of Information Using Finite Order Moments (2303.02423v1)

Abstract: Age of information (AoI) has been proposed as a more suitable metric for characterizing the freshness of information than traditional metrics like delay and throughput. However, the calculation of AoI requires complex analysis and strict end-to-end synchronization. Most existential AoI-related works have assumed that the statistical characterizations of the arrival process and the service process are known. In fact, due to the randomness of the sources and the channel noises, these processes are often unavailable in reality. To this end, we propose a method to estimate the average AoI on a point-to-point wireless Rayleigh channel, which uses the available finite order statistical moments of the arrival process. Based on this method, we explicitly present the upper and lower bounds on the average AoI of the system. Our results show that 1) with the increase of the traffic intensity, the absolute error of the estimated average AoI bounds is first increasing and then decreasing, while the average AoI is monotonically increasing; 2) the average AoI can be effectively approximated by using the first two order moment estimation bounds, especially when traffic intensity is small or approaches unity; 3) tighter bounds can be obtained by using more moments.