FPGA Realization of a Reversible Data Hiding Scheme for 5G MIMO-OFDM System by Chaotic Key Generation-Based Paillier Cryptography Along with LDPC and Its Side Channel Estimation Using Machine Learning Technique

Multiple-Input and Multiple-Output (MIMO) technology is a significant and timely subject, which is highly motivated by the needs of 5G wireless communications. Data transmission performs MIMO, which is highly sensitive. There are several security issues while transmitting the data such as loss of data and code injection. Two efficient methods are Encryption and Data Hiding protection of data in wireless communication. This dissertation suggests FPGA Implementation of RDHS by Chaotic Key Generation-Based Paillier Cryptography with LDPC using machine learning technique. RDHS stands for Reversible Data Hiding Scheme. In a reversible method, the initial stage of preprocessing is to shrink the histogram of image before the process of encryption. Hence, the plaintext domain changing the encrypted images to data embedding cannot result from any pixel repletion. A little distortion data embedding may be taken as the original image may recover the directly decrypted image. Here, the performance metrics of throughput, area consumed, latency, delay, packet delivery, network life and overhead are calculated. The proposed Paillier homomorphic cryptosystem proposes higher network throughput as 99%, higher network life 98%, lower delay rate as 60%, packet delivery as 74%, overhead as 66%, latency as 55% and area consumed as 61% with the existing method such as McEliece, Elgamal and Elliptic curve cryptosystem in the security analysis of the proposed method providing decryption time 94% and encryption time 98% better than the existing method.

Automated summary

Multiple-Input and Multiple-Output (MIMO) technology, motivated by the needs of 5G wireless communications, is an important subject. This dissertation suggests the implementation of RDHS using FPGA, Chaotic Key Generation-Based Paillier Cryptography, and LDPC with machine learning technique. The proposed method shows higher network throughput, network life, and lower delay rate compared to existing methods such as McEliece, Elgamal, and Elliptic curve cryptosystem.