Modeling of Routing Protocols for a MANET Using Neuro Fuzzy Petri Nets

Abstract

In the recent advancements of the mobile technology, the ad hoc networks technology opens new paths for new class of applications. The popularity of the mobile devices with ad hoc networks requiring no infrastructure has great contribution in the fields of data transfer and in mobile computing. In the recent advancements of discrete and continuous event system modeling, Petri Nets are the versatile tools for modeling and analysis. In short Petri Nets are the super sets in discrete event modeling. Petri Nets provide the compact form of multiple interacting components that are equivalent to automaton models. A single Petri Net structure can describe the behaviour of several concurrent components modeled as tokens marking the structure. In this thesis, we design, implement and analyze a proactive routing protocol and reactive protocol using the Neuro Fuzzy Petri Net (NFPN). Individually ANN, FL and PN are great tools for problem-solving control system methodology or information processing or discrete or non-linear statistical data modeling systems. But the combination of these three makes more powerful than that of they were individually. So combination of these three makes it a super set having all the advantages of the three and becomes a hybrid modeling tool. Petri Nets and Artificial Intelligence techniques (FL and ANN) are combined to achieve some goals which are difficult to achieve by using only either of the two. NFPN modeling language and supporting computer tools are powerful enough to specify and analyze a real world communication. The feasibility of the proposed approach can be estimated using the network simulation tools.