An SRN Based Approach for Performance Evaluation of Network Layer in Mobile Ad hoc Networks

Abstract

The application of mobile ad hoc networks (MANET) in emergency and critical cases needs a precise and formal performance evaluation of these networks. Traditional simulation-based performance evaluators like NS-2 and OPNET usually need a considerable time for producing high level performance metrics. Also there is no theoretical background for mentioned simulators, too. In this research, we propose a framework for performance evaluation of mobile ad hoc networks. The presented framework points to the network layer of MANETs using SRN (Stochastic Reward Nets) modeling tool as variation of generalized stochastic Petri net (GSPN). Based on decomposition technique it encompasses two separate models: one for analysis of data flowing process and the other for modeling routing process ; supposing AODV as a routing protocol that is worked out. To verify the presented model, an equivalence-based method is applied. The proposed SRN model has been quantified by deriving two performances metrics as Packet Delivery Ratio (PDR) and End-to-end Delay. Both metrics are also compared to the value obtained from NS-2 simulator versus different number of nodes and four packet generation rates. The results show the obtained values from presented SRN model well matched to the values generated from NS-2 simulator with a considerable lesser execution time.