Technical Program

Paper Detail

Paper IDN.7.4
Paper Title Age-of-Information Revisited: Two-way Delay and Distribution-oblivious Online Algorithm
Authors Cho-Hsin Tsai, Chih-Chun Wang, Purdue University, United States
Session N.7: Age of Information II
Presentation Lecture
Track Networking and Network Coding
Manuscript  Click here to download the manuscript
Virtual Presentation  Click here to watch in the Virtual Symposium
Abstract The ever-increasing needs of supporting real-time applications have spurred a considerable number of studies on minimizing Age-of-Information (AoI), a new metric characterizing the data freshness of the system. This work revisits and significantly strengthens the seminal results of Sun et al. on the following fronts: (i) The optimal waiting policy is generalized from the 1-way delay to the 2-way delay setting; (ii) A new way of computing the optimal policy with quadratic convergence rate, an order-of-magnitude improvement over the state-of-the-art bisection methods; and (iii) A new low-complexity adaptive online algorithm that provably converges to the optimal policy without knowing the exact delay distribution, a sharp departure from the existing AoI algorithms. Contribution (iii) is especially important in practice since the delay distribution can sometimes be hard to know in advance and may change over time. Simulation results in various settings are consistent with the theoretic findings.

Plan Ahead


2021 IEEE International Symposium on Information Theory

11-16 July 2021 | Melbourne, Victoria, Australia

Visit Website!