Detecting and identifying malware: A new signal processing goal


Malware is a range of harmful software that ends up on users' computers allowing unauthorized remote control. Signal processing methods are powerful tools for defense against maltraffic on the Internet. Applying signal processing to maltraffic detection involves mapping observed network behavior to a signal representation and then extracting features from this signal which can be used to determine the presence of maltraffic. Transform-domain techniques have been proposed recently as effective tools to extract underlying information from aggregate traffic. These techniques take off-the-shelf signal representation tools and combine them with equally well-known detection/classification tools.

Publication Title

IEEE Signal Processing Magazine