Deception attack detection based on bandwidth allocation for switched systems with quantization

Abstract

This article discusses a more practical deception attack on the continuous-time switched systems, i.e. a malicious attacker tampers with the binary 01 code transmitted through the network. Assume that the switching is slow enough and the attacker’s energy is limited, we propose a coding transmission strategy based on bandwidth allocation, which selects a part of the network bandwidth for reverse data transmission. Under such strategy, partial detection of attacks and synchronous update of quantization rules are realized. Combined the increasing/decreasing rate of Lyapunov function with the success rate of attack detection, the sufficient conditions ensuring the mean square exponential stability of the system are given. Taking the maximum convergence rate as a performance indicator, the optimal bandwidth allocation scheme is provided for the specific values, which realizes a trade-off between attack detection rate and quantization error. Two-tank system is adopted to illustrate the effectiveness of the theoretical analysis.