A technical research survey on bio-inspired intelligent optimization grouping algorithms for finite state automata in intrusion detection system

Abstract

Network Security plays an essential role in the modern world. Current network services mainly rely on processing of payload in packets.  Deep Packet Inspection (DPI) is a key factor in examining the packet payload which uses the signatures to identify the packet that carries  any viruses, worms, malicious traffic, unauthorized access and attacks. DPI uses regular expression matching as a core operator to  examine the packet payload. Finite State Automata (FSA) are natural representations for regular expression. FSA is usually too large to be  constructed or deployed and has a huge overhead. Finite State Automata frequently leads to state explosion problem which require more  storage space, high bandwidth and more computational time. To overcome this problem, Intelligent Optimization Grouping  Algorithms (IOGA) can be used to distribute the regular expressions into various groups and for each group the Deterministic Finite  Automata (DFA) are built independently. Grouping the regular expression efficiently solves the state explosion problem by achieving  large-scale best tradeoff among the memory utilization and computational time. This paper reviews the various Intelligent Optimization  Grouping Algorithms like Genetic Algorithm, Ant Colony Optimization, Particle Swarm Optimization, Bacterial Foraging Optimization,  Artificial Bee Colony Algorithm, Biogeography Based Optimization, Cuckoo Search, Firefly Algorithm, Bat Algorithm and Flower Plant  Optimization. The discussions states that by effectively using these grouping algorithms along with finite state automata can reduce the  number of states by solving the state explosion blow up problem, providing a balance between the memory consumption, number of   groups and provide faster convergence.