Secret Matrix Lifecycle Management for Compressive Sensing based Cryptographic Mechanisms

In Compressive Sensing (CS) -based cryptographic mechanisms, the measurement matrix is treated as a secret key, so that the encryption and/or the generation of authentication codes are performed during the compressed sampling process. Investigations on the complexity of the signal reconstruction without knowledge of the measurement matrix show that only key management is required to perform the above-mentioned cryptographic operations if the secret CS-matrix is used once for each signal. In standard cryptographic key management, only secret binary information is considered. In contrast, different secret measurement matrices may have different sizes, properties, and structures, depending on the characteristics of the input vector, the desired performance, the compressive sampling rate, and the expected reconstruction results. Therefore, an appropriate key management mechanism is required for secret CS-matrices. In this paper, the protection requirements and properties of the secret CS-matrix and its associated metadata are defined. Then, a high-level secret CS-matrix management architecture with the corresponding lifecycle are designed. The association between the secret measurement matrix, the associated metadata, and their management functions are discussed in our analysis.