Resilient to shared spectrum noise scheme for protecting cognitive radio smart grid readings - BCH based steganographic approach

Cognitive Radio smart grids have recently attracted attention because of high efficiency and throughput performance. They transmit (1) periodically collected readings (e.g. monitoring) and (2) highly sensitive data (e.g. geometric location). However, robustness, efficiency and security of the transmitted data compose an unaddressed unique challenge due to CR shared spectrum possible noise. This paper proposes the first novel hybrid model that combines advanced steganographic algorithms with error detection and correction techniques (BCH syndrome codes) in the CR smart meter context. This will allow us to (a) detect and recover any loss from the hidden confidential information without privacy disclosure, and (b) remedy the received normal readings by using the corrected version of the secret hidden data. To randomize hiding and minimize the distortion, 3D wavelet is used to decompose normal readings into a set of coefficients. To strengthen the security, a key is utilized to generate a 3D randomly selected order used in the hiding process. To accurately measure the detection and recovery capabilities, random noise levels are applied to the transmitted readings. The recovered sensitive information and stego readings are extensively measured using BER, PRD and RMS. It is obvious from the experiments that our technique has robust recovery capabilities (i.e. BER = 0, PRD < 1% and RMS < 0.01%).