Lightweight Random Key Generation Based on Speech Signal for Cryptosystems

Speech signal samples have a random nature. In this paper, random cryptographic keys are generated based on this feature of speech signals. In this method, we use values of the speech signal as the most important feature for generating the random keys. In brute force attacks with different combinations of keys, try to hack it. However, this feature leads to confusion and diffusion. The proposed structure of random key generation is constructed based on the human voice and building blocks such as quantization, permutation, and LFSRs. This structure can generate different key sizes. The generated keys' length equals ۱۲۸, ۱۹۶, and ۲۵۶ bits. For generating the key, we require ۳ seconds speech signal from the user, which is used as an input signal. In addition, to increase the randomness values, we apply the permutation operations to intermediate values, and the intermediate values are added to the values of LFSRs by the modular adder. The statistical analysis for the proposed structure is performed, which ascertains that the random key sequence has acceptable properties. Therefore, the proposed structure of random key generation can be used in the encryption of digital signals such as images and voices with a large amount of data

کلیدواژه ها:

Block cipher ، Random key generation ، Lightweight ، Speech signal