Enhancement of multimedia security using random permutation with wavelet function

Abstract

Digital data include financial documents and medical records, where security is not guaranteed. Unfortunately, cybercrime has developed for several multimedia applications. To overcome cybercrime, forensic analysis is applied. However, an increasing amount of malware is embedded in video payloads, and to minimize this amount, pseudo arbitrary permutation of movable Haar wavelet coefficients (PAP-MHWC) was developed. Pseudo arbitrary permutation using a secret key is permutated to decrease the probability of malware and reduce the distortion rate. Forensic security collects log file information as verification. Haar wavelet coefficients are a sequence of square-shaped functions used in a Fourier analysis to identify cyber forensic regions in video files. Such coefficients enable working with increasingly complex files by decomposing them into various positions and scales. Video frame overlapping is removed using a translation-invariance wavelet transform, thereby improving the forensic security rate. Experiment results show that the proposed PAP-MHWC method achieves a better performance in terms of malware detection accuracy, false-positive ratio, malware detection time, and malware crime probability rate than previous methods.