Modeling and Statistical Analysis of Scenario Metric Parameters of Ad Hoc on Demand Distance Vector Routing Protocol

The goal of our work is to establish chosen scenario metric parameters for ad hoc on demand distance vector (AODV) routing protocol by both simulation and statistical analysis. In first part of our work, we have carried out simulation of AODV on NS2 under different topological conditions. AODV’s performance for end to end delay, energy consumption and packet delivery rate as a function of area, packet rate and packet size is recorded. Based on evaluation of statistical data & graphs, range of scenario metric parameters at which AODV performs best is chosen. Also, Random topology with mobility is considered as chosen topology for AODV after evaluating performance. Performance of MANET is highly influenced by parametric settings for speed, area, packet rate, packet size. Based on our analysis of goodness of fit, residual and prediction bounds we conclude that the regression analysis equation for performance parameters is acceptable predictive empirical model for the range of values obtained from the experimental data. The simulation results show that our empirical model is capable of producing good estimates as statistical parameters values are well within limits. Therefore, we may infer that for the experimental set up under consideration, the chosen scenario metric parameters are packet rate of 35 packets per second, area 500 m2, packet size 512 bytes. With the range chosen for scenario metric, the performance metric parameters range obtained was, Energy Consumption 0.1–3.2 J, end to end delay 3.2–4.8 ms, PDR 90–100 which demonstrated the capability of predictable and repeatable performance.