Survey on Semantics-based Online Malware Detection

Recently, malware has more and more become an essential threat to embedded systems, whereas standard software package solutions like antivirus and patches haven't been so triple-crown in defensive the ever-evolving and advanced malicious programs. During this work, we tend to propose a hardware-enhanced design, GuardOL, to perform on-line malware detection. GuardOL could be a combined approach exploitation processor and FPGA. Our approach aims to capture the malicious behavior (i.e., high-level semantics) of malware. to the present finish, we tend to �rst propose the frequency-centric model for feature construction using system call patterns of renowned malware and benign samples. We then develop a machine learning approach (using multilayer perceptron) in FPGA to train classi�er exploitation these options. At runtime, the trained classi�er is employed to classify the unknown samples as malware or benign, with early prediction. The experimental results show that our answer can do high classi�cation accuracy, quick detection, low power consumption and flexibility for simple practicality upgrade to adapt to new malware samples. One in every of the most benefits of our style is that the support of early prediction detecting 46th of malware at intervals �rst 30 minutes of their execution, while 97 of the samples at 100% of their execution, with < 3% false positives.