The maximum matching degree sifting algorithm for steganography pretreatment applied to IoT

Abstract

Similar to the Internet, the IoT also faces numerous information security issues. Because the multimedia sensors that form Wireless Multimedia Sensor Networks (WMSNs) inherently operate with large amounts of data with high redundancy, steganography appears to be a better way to ensure the security of information in this medium than does cryptography. Considering that computing power and energy resources are often limited in the IoT, it is more effective and feasible to use steganography to obtain the outcomes that people expect, including better concealment and security. Currently, some advanced steganalysis techniques can reliably detect embedded secret messages, To defend against such steganalysis techniques, most scholars in this field have focused on developing or improving advanced embedding algorithms. However, in this article, choosing a suitable carrier is also considered to be a good approach to improve resistance. Thus, this paper proposes the Maximum Matching Degree(MMD) sifting algorithm, which is based on the principle of " minimizing the effect of embedding" (here, we measure the effect of embedding by the number of modified bits) and can be applied to choose the best carrier by minimizing the number of bits that will be modified during embedding. This approach can be regarded as a steganography pretreatment. Moreover, it is easy to implement, which is also important in IoT situations. Using greyscale-images as carriers, we conducted experiments. The results demonstrated that the pretreatment method not only improves embedding efficiency and steganalysis resistance (to some common steganalysis techniques) but is also extremely versatile. This result has significant implications for steganography and broad implementation prospects.