Scalable Probabilistic Models for 802.11 Protocol Verification
Title: Scalable Probabilistic Models for 802.11 Protocol Verification
Abstract: This research aims to create more scalable and accurate models for the 802.11 protocol, enabling efficient verification of real-time properties. The authors identify ways to simplify probabilistic models, optimize them using probabilistic timed automata, and develop a formal framework to reduce state space. Their optimizations, including abstracting away the acknowledgment protocol and deterministic waits, and considering only a subset of allowed packet sizes, do not change the probability measures. The reduced models are immediately verifiable in PRISM and can be used with other tools like RAPTURE. The results show a significant reduction in state space, making verification more feasible for networks of moderate size.
Link to Article: https://arxiv.org/abs/0403044v1 Authors: arXiv ID: 0403044v1