Joint Distributed Compression and Encryption of Correlated Data in Sensor Networks

In this paper, we propose, formulate, and study a joint distributed data compression and encryption scheme suitable for wireless sensor networks where we adopt the structured encryption system of advanced encryption standard (AES). The distributed compression is achieved as per the Slepian-Wolf coding theorem, using channel codes. Core to achieving optimal compression in the joint compression and encryption is the preservation of correlation among different blocks of data despite applying cryptographic primitives. We establish that the correlation between sources remains unchanged when cryptographic primitives, namely key addition and substitution are applied. However, as a requirement of security in the encryption, any correlation between two inputs to a encryption system is removed with diffusion techniques. Compliance to the requirements of diffusion layer of AES cipher is achieved by designing the compression function so as to maintain branch number property. We establish the necessary and sufficient condition for achieving a compression function with branch number property and show that distributed compression using non-systematic Reed Solomon (RS) code can satisfy this condition