Online Experimental Design for Network Tomography

2025-04-30

AI Summary

To address the low efficiency of real-time topology and state inference in network tomography, this paper proposes an online sequential decision-making framework for dynamic probe allocation. It is the first to integrate optimal experimental design (OED) with online learning, enabling adaptive tomography for both packet-loss networks and quantum-bit-flip networks. The method models unicast and multicast path probing, jointly leveraging maximum likelihood estimation (MLE) and concentration inequalities; under Lipschitz continuity and parameter concentration assumptions, it establishes a rigorous regret bound framework. Theoretical convergence guarantees are derived for two provably identifiable network classes. Simulations demonstrate that the proposed approach improves estimation accuracy by 23%–41% and accelerates convergence by 1.8–3.2× over state-of-the-art methods, while exhibiting strong generalization capability.

Technology Category

• Machine Learning: Online Learning & Bandits
• Search and Optimization: Learning to Search
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization

Abstract

How to efficiently perform network tomography is a fundamental problem in network management and monitoring. A network tomography task usually consists of applying multiple probing experiments, e.g., across different paths or via different casts (including unicast and multicast). We study how to optimize the network tomography process through online sequential decision-making. From the methodology perspective, we introduce an online probe allocation algorithm that dynamically performs network tomography based on the principles of optimal experimental design and the maximum likelihood estimation. We rigorously analyze the regret of the algorithm under the conditions that i) the optimal allocation is Lipschitz continuous in the parameters being estimated and ii) the parameter estimators satisfy a concentration property. From the application perspective, we present two case studies: a) the classical lossy packet-switched network and b) the quantum bit-flip network. We show that both cases fulfill the two theoretical conditions and provide their corresponding regrets when deploying our proposed online probe allocation algorithm. Besides these two case studies with theoretical guarantees, we also conduct simulations to compare our proposed algorithm with existing methods and demonstrate our algorithm's effectiveness in a broader range of scenarios.

Problem

• Optimize network tomography via online sequential decision-making
• Develop online probe allocation algorithm for dynamic tomography
• Validate algorithm in lossy packet-switched and quantum networks

Innovation

• Online probe allocation algorithm for tomography
• Optimal experimental design principles applied
• Maximum likelihood estimation for parameter analysis