Real-time Implementation of Chaos Synchronization Using Neural Network based on FPGA

Abstract

As a kind of non-periodic, random-like, and initial sensitive nonlinear dynamic phenomenon, chaos is an important transmission carrier for secure communication. Aiming at the problem of complex structure and difficult realization of the traditional chaos synchronization method, this paper proposes a neural network method based on Field Programmable Gate Array (FPGA) to synchronize chaos. At the transmitter side, FPGA is used to generate a chaotic signal according to the Lorenz state equation, and the transmission signal is encrypted in real-time mode. The neural network is used to model the chaotic transmitter precisely, and the signal is decrypted by the trained neural network at the receiver side, which is verified on FPGA. Finally, the real-time encryption and decryption of a 10 Mbit/s signal are realized successfully. This method greatly simplifies the synchronization structure of a chaotic secure communication system and makes it possible for point to multipoint real-time chaotic communication.

Keywords

• chaos synchronization
• neural network
• real-time
• fpga