Recursive Optimization of Finite Blocklength Allocation to Mitigate Age-of-Information Outage

TL;DR

This paper investigates how to allocate finite blocklengths in TDMA systems to reduce the occurrence of high Age-of-Information outages, using a Markov Decision Process model and risk-sensitive policies.

Contribution

It introduces a Markov Decision Process framework for finite blocklength allocation to mitigate AoI outages and proposes a policy iteration approach for improved robustness.

Findings

Risk-sensitive policies outperform average AoI optimization in outage mitigation.

Finite blocklength allocation significantly affects AoI outage probability.

Burstiness analysis provides insights into outage behavior in FBL regimes.

Abstract

As an emerging metric for the timeliness of information delivery, Age-of-Information (AoI) raises a special interest in the research area of tolerance-critical communications, wherein sufficiently short blocklength is usually adopted as an essential requirement. However, the interplay between AoI and finite blocklength is scantly treated. This paper studies the occurrence of high AoI, i.e., AoI outage, in TDMA systems with respect to the blocklength allocation among users. A Markov Decision Process model is set up for the problem, which enables a static state analysis, and therewith a policy iteration approach to improve the AoI robustness is proposed. The burstiness of outages is also analyzed to provide additional insights into this problem in the finite blocklength (FBL) regime. It is shown that, different from average AoI optimizations, a risk-sensitive approach is significantly beneficial for AoI outage optimizations, on account of the FBL regime.