University:   Technical University Braunschweig
Researcher: Prof. Lars Wolf, Sven Jaap  
Department: Department of Computing Science 

Research With OPNET:

Simulation of Wireless Ad Hoc Networks

The widespread introductions of wireless networks such as 3G/4G mobile networks, wireless LANs, mobile Ad Hoc networks, micro sensor networks, as well as DVB-T/DVB-H networks, are challenging the Internet with new communication paradigms. Especially mobile ad hoc networks are a challenging and fast growing research area. In mobile ad hoc networks, each mobile node is an independent device that moves on an individual path. Several communicating nodes may sopntaneouslly establish a network that is self-configuring. For transfering data to other nodes that are not in a node's immediate communication range, other nodes may used as relays. This way, information may take several hops to reach its destination.
The application area of these networks is huge. On the one hand mobile ad hoc networks may be used in disaster areas for pubilc safety purposes. This way, relief units are abe to communicate with each other, without the need of any fixed infrastructure. Another application in this area is the monitoring of dikes with the help of networked sensor nodes.

Our reasearch areas focus on other application areas of mobile multihop ad hoc networks. On the one hand, we deal with mechanims and algortihms to enable the use of multi-player games in ad hoc networks. These games are very delay and jitter sensitive. Players expect, that they can play a game fluenty and without any opponent that suddenly (dis-)appear.
Another focus of our research activities are vehicular ad hoc networks (VANETs), where vehicles driving on a road form an ad hoc network and exchange information about traffic or road conditions, accidents and so forth. In addition, vehicles+ may use acces points along the road to get acces to the Internet via the VANET. However, the communnication between vehicles is not limited to informational messages. Vehicles may exchange locally measured sensor data with each other, to increase a vehicle's knowledge about the local traffic situation. Moreover, vehicles may co-ordinate driving manoeuvres like overtaking manoeuvres or merging lanes in front of constructions. These mechanisms, can assist the driver in difficult traffic situations.

Tasks

The evaluation of mechanisms and algorithms for mobile ad hoc networks in real scenarios is hardly possible in most cases. In order to get compareable results, reproducable scenarios are needed. Moreover, scalability is often an important parameter for mobile ad hoc networks. Therefore, sometimes a large number of mobile nodes is needed to analyze developed protocols. Especially in the area of vehicular ad hoc networks, it is hardly possible to setup a test environment with hunderets of vehicles. Thus, simulations are needed for evaluation purposes.

In a first step we compare various network simulators including OPNET in order to compare them with respect to performance issues, scalability and functionality. On the basis of these results, we can choose the simulator that is best-suited for our purposes.
In the next step, we need to develop realistic movement scenarios. Previous simulations showed that the movement of nodes within the simulation area has a clear effect on the simulation results. Thus, realistic movement patterns that are characteristic of the application area are needed. In our case, this inculdes scenarios fpr multi-player games on schoolyards, in parks or in a train or airport terminal, as well as realistic traffic scenarios on a freeway or in city. In addition, realistic models for radio wave propagation are needed that model multipath propagation and obstacles in the modelled simulation area.
When having these components, we are able to analyze mechanisms for multi-player games in ad hoc networks of cooperative driver assistence functions with the help of network simulaitons

(Planned) Usage of OPNET

We plan use OPNET Modeler and its specialized Wireless module to simulate vehicular ad hoc networks as well as multi-player game scenarios in ad hoc networks. OPNET will be especially helpful for us to easily introduce our models into the simulated network protocol stack. Moreover, we want to integrate our mobility and propagation models in OPNET. We want to use OPNET for small- and large-scale simulations with e.g. more than 1000 vehicles driving on a road, inorder to evaluate the scalability and functioning of the developed protocols and mechanisms

Publication: