Reconstruction of Counting Process in Real-Time: the Freshness of Information Through Queues

Meng Wang,Wei Chen,Anthony Ephremides
DOI: https://doi.org/10.1109/icc.2019.8762021
2019-01-01
Abstract:For the emerging Internet of Things (IoT), one of the most important basic problems is how to reconstruct signals in real-time from a set of under-sampled and delayed samples. The sampling omits the details of the signals of interest and the delayed samples against the requirement of real-time. As a result, distortion occurs between the interested signal and the reconstructed signal. In this paper, we focus on minimizing the average distortion defined as the 1-norm of the difference of the two signals under the scenario that a Poisson counting process is reconstructed in real-time on a remote monitor. We derive the average distortion-sampling rate function, with which the optimal sampling rate can be obtained as well as the minimum average distortion. To further decrease the average distortion, an algorithm is proposed by sacrificing the real-time requirement in a small degree.
What problem does this paper attempt to address?