age-of-information scheduling

This work addresses the challenges of unreliable status updates and heterogeneous operational costs in integrated sensing and communication systems by jointly optimizing the Age of Information (AoI) and long-term aggregate cost. In the single-source setting, a Markov decision process is formulated, revealing an optimal policy with a monotone threshold structure, and a state-space truncation method with rigorous error bounds is developed. For the multi-source scenario, the problem is cast as a restless multi-armed bandit, leading to a broadly applicable approximate Whittle index scheduling policy. Theoretical analysis provides guarantees on both policy structure and truncation error, while numerical experiments demonstrate that the proposed approximate Whittle index significantly outperforms baseline methods in both indexable and non-indexable regimes.

This paper addresses the information freshness optimization problem in multi-source single-server mobile edge computing systems, using Peak Age of Information (Peak AoI) as the performance metric. We jointly optimize the scheduling policy for stochastically arriving sources and the status sampling mechanism. Our key theoretical contribution is the rigorous proof that randomized schedulers are optimal under both preemptive and non-preemptive settings. Leveraging this insight, we decouple the joint optimization into an alternating optimization over scheduling frequency and sampling threshold, and propose a transmission-aware threshold-based sampling policy along with a general alternating optimization framework. The proposed algorithm effectively handles the inherent non-convexity of the problem. Numerical experiments demonstrate that it closely approaches the theoretical optimum across diverse system configurations, significantly reducing Peak AoI while achieving a balanced co-design of communication and computation resources.

This study addresses the optimization of status updates for remote navigation agents in integrated sensing and communication (ISAC) systems, aiming to balance information freshness against system overhead. By formulating a long-term cost model based on the Age of Information (AoI), the work incorporates the stochastic success probabilities and associated costs of both sensing and communication into a sequential decision-making framework defined over a two-dimensional AoI state space. The problem is modeled as a discounted infinite-horizon Markov decision process, and theoretical analysis reveals that the optimal stationary policy exhibits a monotone threshold structure, characterized by a non-decreasing switching curve—offering both interpretability and implementability. Numerical experiments validate the structural properties of the value function and optimal policy, demonstrating that an AoI-driven optimization objective can effectively guide ISAC system design.

Conventional Age of Information (AoI) analysis assumes zero initial age for update packets, neglecting non-negligible prior delays—introducing systematic bias in freshness evaluation. Method: We model AoI in a single-server queue with updates possessing non-zero initial age, leveraging queuing theory and stochastic processes to accommodate forwarding, tandem, and retransmission configurations. Contribution/Results: We derive the first general closed-form expression for average AoI incorporating an initial-age correction term; rigorously establish its upper and lower bounds; and prove that, under independent arrivals, the correction term scales linearly with initial age. Experimental validation on multi-stage tandem networks confirms tight bound convergence and demonstrates substantial improvement in AoI estimation accuracy—particularly critical for multi-hop communication and distributed sensing—thereby rectifying the inherent inaccuracy of the zero-initial-age assumption.

This study addresses the inadequacy of traditional age-of-information (AoI) metrics—based solely on average AoI—in capturing closed-loop LQR tracking performance, as they lack grounding in control theory. Focusing on scalar linear time-invariant systems with delayed intermittent updates, the authors reformulate the infinite-horizon LQR problem as an optimization over the distribution of update intervals. They establish, for the first time from a control-theoretic perspective, that control performance depends critically on higher-order and even exponential moments of the inter-update interval distribution, not merely its mean. Leveraging stochastic control, LQR optimization, and moment analysis, the work demonstrates that distinct scheduling policies yielding identical average AoI can result in markedly different control performance. This phenomenon is validated using real-world NGSIM vehicle trajectory data, underscoring the necessity of distribution-aware AoI metrics and full distributional modeling in control-oriented network design.

This work addresses the challenge of information freshness in wireless monitoring systems, where stochastic data arrivals and unreliable channels cause asynchronous evolution of Age of Information (AoI) at the local sensor and the central monitor. Recognizing that conventional scheduling strategies relying solely on the central AoI are suboptimal, the paper proposes a dual-AoI model to capture this asynchrony and formulates the scheduling problem as one of minimizing the long-term average AoI. By modeling the system as a Markov decision process (MDP), an optimal transmission policy is derived. The study establishes, for the first time, the monotonicity and threshold structure of the optimal policy under the dual-AoI framework and provides necessary and sufficient conditions for AoI stability. Simulation results demonstrate that the proposed low-complexity policy significantly outperforms existing approaches in enhancing overall information freshness.

This work addresses the problem of minimizing the average Age of Information (AoI) for status updates from multiple devices in bandwidth-constrained Internet-of-Things systems where channel states are unobservable. The scheduling problem is formulated as a partially observable restless multi-armed bandit. By applying Lagrangian relaxation, the problem is decoupled into tractable subproblems, and leveraging the threshold structure of the optimal policy, the authors establish indexability for the first time under Markovian channel dynamics. A closed-form Whittle-like index policy is then proposed, which substantially reduces computational complexity. The resulting algorithm achieves near-optimal performance—approaching the theoretical lower bound—in large-scale or resource-limited settings, significantly outperforming existing baseline methods.

This work addresses the problem of minimizing the Age of Information (AoI) in scenarios with randomly arriving updates by proposing a multi-threshold preemption policy that jointly considers packet age and system age in its decision-making. By formulating a stochastic process model that integrates both age metrics into a unified multi-threshold framework, the proposed approach overcomes the limitations of conventional single-threshold or probabilistic preemption strategies. Theoretical analysis reveals structural properties of the optimal policy, and experimental results demonstrate that the proposed method significantly outperforms existing approaches, effectively reducing AoI in stochastic update environments.

This study addresses the minimization of the age of information (AoI) and peak AoI (PAoI) in a single-source, single-server continuous-time status update system by optimizing preemption policies. The work proposes the first general analytical framework capable of exactly computing the average AoI and PAoI under arbitrary age-dependent preemption strategies. Leveraging this framework, two practical policies—probabilistic preemption (PP) and threshold-based preemption (TP)—are efficiently optimized via one-dimensional line search. Numerical results under log-normal service times demonstrate that the optimized PP and TP policies significantly reduce both AoI and PAoI, thereby validating the effectiveness and superiority of the proposed approach.