Dynamic Server Allocation Under Stochastic Switchover on Time-Varying Links

2026-01-22

AI Summary

This work addresses the performance degradation in dynamic server allocation caused by stochastic heterogeneous switching delays over time-varying links. To tackle this challenge, the authors propose ACI, a non-myopic frame-based scheduling framework that explicitly incorporates switching delays into scheduling decisions for the first time. By amortizing delay overhead and integrating queue backlog states into a tunable urgency metric, ACI flexibly balances delay performance while guaranteeing throughput optimality. Leveraging Lyapunov drift theory and backlog-driven control, the framework achieves throughput optimality within a scaled capacity region. Extensive simulations in a multi-UAV free-space optical (FSO) backhaul network demonstrate that ACI significantly outperforms the conventional Max-Weight policy.

Technology Category

• Planning, Routing, and Scheduling: Scheduling under Uncertainty
• Reasoning under Uncertainty: Stochastic Optimization
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

Dynamic resource allocation to parallel queues is a cornerstone of network scheduling, yet classical solutions often fail when accounting for the overhead of switching delays to queues with superior link conditions. In particular, system performance is further degraded when switching delays are stochastic and inhomogeneous. In this domain, the myopic, Max-Weight policy struggles, as it is agnostic to switching delays. This paper introduces ACI, a non-myopic, frame-based scheduling framework that directly amortizes these switching delays. We first use a Lyapunov drift analysis to prove that backlog-driven ACI is throughput-optimal with respect to a scaled capacity region; then validate ACI's effectiveness on multi-UAV networks with an FSO backhaul. Finally, we demonstrate how adapting its core urgency metric provides the flexibility to navigate the throughput-latency trade-off.

Problem

• Dynamic Server Allocation
• Stochastic Switchover
• Time-Varying Links
• Switching Delays
• Network Scheduling

Innovation

• dynamic server allocation
• stochastic switching delay
• non-myopic scheduling
• Lyapunov drift analysis
• throughput-latency trade-off