Improving Speed of Virus Scanning - Applying TRIZ to Improve Anti-Virus Programs

Abstract

As the number of viruses increases the amount of time to scan for those viruses also increase. This situation is worsened because of the increasing number of files typically stored in a computer system. Besides more complex tests are required to detect the modern day’s intelligent viruses. Thus a combination of all these factors makes a full virus scan very lingering and resource consuming. With thousands of types of viruses and gigabytes of storages, a typical virus scanning may take several hours. Because of this reason many users don’t like to virus scan their computers unless so required. This article highlights the common factors that increase the scanning time. For example, apart from the increased number of files and viruses, the scanning of compressed files requires more time, a lot of time is wasted by scanning the same files again and again in different scanning sessions, poor searching and file comparing algorithms can increase the scanning time, a reliable scanning may require a combination of various methods which demand more time to scan.The article discusses some methods of reducing the duration of a typical virus scanning session. Some of the solutions are; avoiding repetitive scanning by storing the AV state of previously scanned files, selective scanning of files (e.g., scanning only those files which have been changed since last scanning), applying only selected scanning methods relevant to specific file types (e.g., scanning MS Word files for Macro viruses and not for boot sector viruses), applying improved 'string search' methods for signature scanning, applying improved method of file comparison for integrity testing, reducing disk read time by reading files in a separate threads etc.