Connectivity jamming game for physical layer attack in peer to peer networks

TL;DR

This paper models wireless network connectivity under jamming attacks using a stochastic game framework, proposing a unified metric based on attack duration to develop anti-jamming strategies that enhance network resilience.

Contribution

It introduces a novel approach to compare diverse attack metrics via expected attack duration and proposes a maxmin strategy for anti-jamming defense in peer-to-peer networks.

Findings

Unified attack metric improves comparison of different attack types.

Maxmin strategy enhances network resilience against interference.

Framework applicable to various network topologies.

Abstract

Because of the open access nature of wireless communications, wireless networks can suffer from malicious activity, such as jamming attacks, aimed at undermining the network's ability to sustain communication links and acceptable throughput. One important consideration when designing networks is to appropriately tune the network topology and its connectivity so as to support the communication needs of those participating in the network. This paper examines the problem of interference attacks that are intended to harm connectivity and throughput, and illustrates the method of mapping network performance parameters into the metric of topographic connectivity. Specifically, this paper arrives at anti-jamming strategies aimed at coping with interference attacks through a unified stochastic game. In such a framework, an entity trying to protect a network faces a dilemma: (i) the underlying motivations for the adversary can be quite varied, which depends largely on the network's characteristics such as power and distance; (ii) the metrics for such an attack can be incomparable (e.g., network connectivity and total throughput). To deal with the problem of such incomparable metrics, this paper proposes using the attack's expected duration as a unifying metric to compare distinct attack metrics because a longer-duration of unsuccessful attack assumes a higher cost. Based on this common metric, a mechanism of maxmin selection for an attack prevention strategy is suggested.