Anti-virus software is a key security technology on today's end user
systems. Current anti-virus engines use two complementary techniques to
detect malware. One is to statically scan potential malware sample files
for certain patterns which are known ("malware signatures"). The other
is to dynamically detect typical malicious behavior (e.g., modifications
of registry keys, DLL injections etc.) upon execution of a sample. No
anti-virus product can reliably detect malware. Rather, all products are
plagued by false positives and false negatives. An interesting approach
to improve the reliability of detection is to run several anti-virus
products on a given malware sample. There are several online scanning
services, that implement this approach. However, for performance reasons
these services only use the static signature detection functionality of
the anti- virus products, and thus do not take advantage of the full
functionality of current anti-virus engines. This book explains how to
overcome this limitation and to build an efficient online malware
scanning service that fully utilizes the capabilities of current
anti-virus engines.