Integrated Malware Detection and Compression Framework for Secure and Efficient Cloud Storage

Effective compression methods and security mechanisms have to be devised in order to fulfill the increasing requirement of cloud storage systems. Malware detection and compression are considered two different processes in the traditional method. As the files are constantly being evaluated, these processes involve a significant processing overhead. The author of this study has proposed a novel system named "Multi-Layer Intelligent Compression Pipeline" (MLICP) that employs an intelligent three-stage structure in order to integrate malware detection and compression in an efficient way. The Random Forest classifier is employed in the first stage in order to determine the hazard levels on the basis of 14 characteristics. The best compression methods are selected in the second stage on the basis of file characteristics and security risk profiles. A set of 1200 files, with an average size of 273 bytes, consisting of 1000 benign samples from news headlines and 200 malicious API request sequences, was employed to test the system. The experimental results indicate that the average compression ratio is 17.71%, with 35.94% compression for high-risk files and 14.09% for benign files. The malware detection program correctly classified all 200 high-risk files with 100% accuracy and without false positives. An algorithm selection experiment revealed that LZMA was used in 16.7% of high-risk files and GZIP in 83.3% of benign text files. The performance analysis indicates that 99% of files are processed in 7.77 milliseconds, with a median processing time of 3.55 milliseconds per file. Data integrity is confirmed in the third step through cryptographic hash verification. The proposed approach can be extremely beneficial to data centers, backup solutions, and organizations dealing with sensitive information in the legal, financial, and healthcare industries.