Joint Distribution of Ages of Information in Networks

TL;DR

This paper develops a novel stochastic hybrid system framework to analyze the joint distribution of Age of Information (AoI) in networks, enabling comprehensive understanding of information freshness across multiple monitors.

Contribution

It introduces a tensor-based approach to derive differential equations for joint AoI moments and MGFs, extending analysis beyond marginal distributions to the joint distribution in network systems.

Findings

Derived differential equations for joint AoI moments and MGFs.

Established conditions for asymptotic stability of the joint distribution.

Provided closed-form stationary joint MGF expressions for multi-source systems.

Abstract

We study a general setting of status updating systems in which a set of source nodes provide status updates about some physical process(es) to a set of monitors. The freshness of information available at each monitor is quantified in terms of the Age of Information (AoI), and the vector of AoI processes at the monitors (or equivalently the age vector) models the continuous state of the system. While the marginal distributional properties of each AoI process have been studied for a variety of settings using the stochastic hybrid system (SHS) approach, we lack a counterpart of this approach to systematically study their joint distributional properties. Developing such a framework is the main contribution of this paper. In particular, we model the discrete state of the system as a finite-state continuous-time Markov chain, and describe the coupled evolution of the continuous and discrete states of the system by a piecewise linear SHS with linear reset maps. Using the notion of tensors, we first derive first-order linear differential equations for the temporal evolution of both the joint moments and the joint moment generating function (MGF) for an arbitrary set of age processes. We then characterize the conditions under which the derived differential equations are asymptotically stable. The generality of our framework is demonstrated by recovering several existing results as special cases. Finally, we apply our framework to derive closed-form expressions of the stationary joint MGF in a multi-source updating system under non-preemptive and source-agnostic/source-aware preemptive in service queueing disciplines.