Northeastern University:
Department of Electrical and Computer Engineering

Coordinating Faculty Member: Prof. A. Bruce McDonald

Applications and Use of OPNET Simulation:

(1) Undergraduate Education

(1.1) In ECE 1330 - Noise and Stochastic Processes junior-level undergraduate students in electrical and computer engineering are taught the basics on probability, random variables, stochastic processes and their applications in engineering. The honors project provides a select group of talented students and opportunity to learn the fundamental concepts of simulation modeling and experimental data analysis. Students electing to take the honors adjunct receive extra instruction and read supplementary material provided by the insructor on the topics of discrete-event simulation and queueing systems. The project consists of developing and validating models of simple queueing systems and present and analysis that demonstrates and understanding of the major issues involved in simulating queueing systems. The objective of the project is to introduce motivated students to an important design tool, namely, discrete event simulation using OPNET, and to provide them with hands on experience in developing basic simulation models and analyzing output data.

(1.2) In ECE 1458 - Computer Communications Networks, OPNET is used to re-enforce fundamental concepts to students during their initial exposure to computer networks. Simulation provides a simple environment to demonstrate aspects of performance that are otherwise difficult to capture without understanding sophisticated mathematics. OPNET is used as a tool incoroprated into several homework assignments.

(2.1) A Two-Dimensional Discrete-State Continuous-Time Stochastic Process for Modeling MAC and Physical-Layer Characteristics in Wireless Ad Hoc Networks: Simulation runtimes for large or highly dynamic ad hoc networks are notoriously long and difficult to extract statistically valid, or, general results from. In this project we are using knowledge gained through our own analytical study of contention-based MAC in ad hoc networks to develop, first, a collection of distributions for MAC delay based on IEEE 802.11 DCF that is dependent on our two-dimensional discrete-state model, and, second, build a toolkit for developing similar models for any MAC protocol/PHY-layer combination. The objective is to make available a library of these empirically derived stochastic processes and the toolkit for users of OPNET and other simulators in order to substntially reduced the complexity of ad hoc network simulation models and their associated runtimes.

(2.2) Integrated Cross-Layer QoS Architecture in Ad Hoc Networks: The objective of this research is to develop and evaluate the performance of novel routing and resource allocation strategies that provide statistical QoS guarantees in ad hoc networks. We are investigating a framework based on cross-layer principles to support robust statistical QoS communications in recongurable wireless networks. The methodology is based on soft-state resource reservations made on a statistically selected set of nodes over a sequence of statistically bounded time intervals and datagram routing using constrained relative-location based forwarding. The fundamental principle uses a cross-layer approach that develops and leverages real-time three-space localization, trajectory and link-quality prediction techniques, coupled with novel admission control and scheduling policies. Many components of this research will require simulation---much of which will involve models developed using OPNET software.