Runtime Verification of Interactions Using Automata

2025-11-01

AI Summary

Distributed systems lack a global clock, rendering local execution traces non-reorderable and impeding verification of interaction consistency across multiple components. Method: This paper proposes two automata-theoretic runtime verification techniques for multi-trace analysis. The first models local subsystem traces as a synchronous product automaton and checks conformance against an interaction specification. The second enhances practicality via a preprocessing mechanism and a fault classification strategy to improve result reusability and diagnostic precision. Contribution/Results: Experiments demonstrate that both methods maintain high detection accuracy while exhibiting strong scalability. They enable real-time compliance monitoring of distributed system interactions and support precise fault localization—addressing key challenges in verifying cross-component behavior under partial observability and asynchronous execution.

Technology Category

• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Multiagent Systems: Distributed Problem Solving
• Search and Optimization: Distributed Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• Security and Privacy: Tracking, profiling, and countermeasures against them
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Runtime verification consists in observing and collecting the execution traces of a system and checking them against a specification, with the objective of raising an error when a trace does not satisfy the specification. We consider distributed systems consisting of subsystems which communicate by message-passing. Local execution traces consisting of send and receive events are collected on each subsystem. We do not assume that the subsystems have a shared global clock, which would allow a reordering of the local traces. Instead, we manipulate multitraces, which are collections of local traces. We use interaction models as specifications: they describe communication scenarios between multiple components, and thus specify a desired global behaviour. We propose two procedures to decide whether a multitrace satisfies an interaction, based on automata-theoretic techniques. The first procedure is straightforward, while the second provides more information on the type of error and integrates the idea of reusability: because many multitraces are compared against one interaction, some preprocessing can be done once at the beginning. We implement both procedures and compare them.

Problem

• Runtime verification of distributed message-passing systems
• Checking multitraces against interaction models without global clocks
• Automata-based procedures for error detection and reusable analysis

Innovation

• Runtime verification using automata-based interaction models
• Multitrace analysis without shared global clock assumption
• Two procedures with error diagnosis and preprocessing reuse