Embedding and Extraction of Data in Speech for Covert Communication

This paper proposes an implementation of speech steganography in real-time for data concealment. Speech steganography is a technique that allows for the covert communication of information by embedding it within an audio signal. This paper discusses a methodology that is efficient in embedding and extracting data in real-time, without introducing any significant delay or distortion to the original speech signal. The technique embeds the data into the speech signal using a combined least significant bit (LSB) and discrete wavelet transform (DWT) methodology. The paper also presents an approach to optimize the embedding process by selecting appropriate segments of the speech signal that is less perceptually important. Experimental results show that the algorithm is capable of achieving high embedding capacity while maintaining good speech quality. Many platforms, including those for secure communication, digital watermarking, and copyright protection, can use the suggested technology.