Genetic algorithm application in information security systems

Information infrastructure protection against malicious software, such as vi-ruses, Internet worms and botnets is a crucial task that requires development and implementation of both reactive, such as patches, removal tools, antivirus soft-ware updates and proactive, such as countermeasure planning in advance, meas-ures. Success of these measures application highly depends on reaction time and understanding of malware evolution trends respectively. Reaction time should be directly proportional to the risk level posed by malware. Currently risk evalua-tion task is based on expert knowledge and no automatic evaluation systems ex-ist. Existing malware models mainly concentrate on malware epidemic conse-quences modelling, i.e. forecasting the number of infected computers, simulating malware behaviour or economic propagation aspects and are based only on cur-rent malware propagation strategies.Present study presents an innovative genetic algorithm based malware risk level evaluation and malware evolution forecasting model. Genetic algorithm approach was selected taking into consideration its efficiency while solving tasks with large solution space and ability to model the evolution process which is the case for malware, often considered as a form of artificial life, evolution forecasting. The model presented covers the malware feature representation de-scription in the genetic algorithm suitable format, evolution evaluation fitness functions for propagation and survivability... [to full text]