Noise considerations and cryptanalysis of chaos-based secure communication systems

Abstract

This paper discusses the topic of using chaotic models for constructing secure communication systems. It investigates two different case studies that use encryption/decryption functions with varying degrees of complexity and performance. The first case study explores the use of an improved cryptography method for scrambling the transmitted signal. The improvement is based on using both the transmitter states and parameters for performing the encryption. The second case study augments a parameter update law to the previous design such that the encryption method is more robust. It uses a decoupling technique for which the synchronization process is completely isolated from the parameter identification algorithm. The effect of using a noisy public channel is investigated. The security of the proposed methods and their advantages and limitation are analyzed using intruder return map attacks to the communication channel. The Lorenz system was used to exemplify the suggested techniques, while transmitting a digital signal. The paper ends with a discussion that investigates various techniques to optimize the performance of the proposed systems.