Ad-Hoc Szenario Editor

ExternalS


NS-2
NS-2 Dokumentation
TU-Braunschweig
Kommunikation und
Multimedia


XStream

CodeHaggis 



Preface


In our days mobile communication is an essential part of our environment and there is a tendency of increasing wireless independence of this technology.  In case that the environment once does not support the usual infrastructure for  using these techniques, spontaneous unstructured ad-hoc networks become necessary,  which supply us with the important services, that allow us to communicate with each other.

Ad-hoc networks often are neither planned nor tested in real environments, they are only simulated in tests to save time, manpower and money. While the simulation of wireless networks is already well developed, the random modelling of nodes is still based on very simple strategies, which have to be improved.

The NS-Mapper szenario editor for of mobile ad-hoc networks, is improved and extended by adding more realistic strategies, such as random based node placement, movement and traffic. Preexisting algorithms were used and inspected for these editor. 

Information

NS-Mapper


In contrast to many other scenario editors, which generate node motion and traffic, the NS-Mapper is not limited to only one special category. The NS-Mapper connects modelling, random node placement, random ad-hoc routing and random traffic tools in only one single application, whereas most other available tools, described in the proceeding sections, can only calculate and vary scenarios as a whole at once. In other programs a direct influence or a separation of different behaviours for different nodes is not possible or at least very difficult.

The NS-Mapper editor consists of two layers: The default GUI-based version, in which scenarios can be modelled by the user and a command line version for creating simulation files from scenario descriptions automatically. A NS-Mapper scenario contains manually modelled nodes and node mobility, as well as randomly created nodes and their movement. A user can manually define the number, the position and the mobility of nodes, he is particularly interested in, as well as their traffic pattern. Additionally, he can let the editor randomly create a given number of additional nodes to form a larger ad-hoc network. NS-Mapper scenarios can then be used to create a number of NS-2 simulation files, in which all random elements are recalculated by different seed settings. This allows the creation of scenarios with a group of fixed nodes, which behave in a predetermined way, and some random nodes, which create background traffic and provide connectivity in the ad-hoc network, according to some general settings.

Another feature of the editor are the so called MovementFields. A MovementField is a rectangle or a polygon, in which mobile nodes can be created and move. Mobile nodes are not allowed to move outside a MovementField. This feature can be used to create walls or other borders, which the nodes cannot pass. A node can only move from its own to another MovementField, when these are connected or overlap each other.



Gauss-Markov node movement

The figure above shows a screenshot of the NS-Mapper GUI with some static MovementFields (blue with black borders) and some placed nodes (green). The black lines in the simulation picture represent their movement, which is generated by the Gauss Markov algorithm.

There are two kinds of MovementFields: static and dynamic fields. Static MovementFields remain at their position during the simulation. For dynamic fields, the user can create one or more time-based movement vectors, which contain start time, stop time, speed and/or destination. All mobile nodes move with their associated MovementField which means that a movement vector of the field is added to the movement vector of its nodes. Basically, dynamic MovementFields move around with the mobile nodes sitting or moving on top of them. Examples of a static MovementField are paths, were nodes can move along on pavements, streets, or rooms in buildings. Dynamic MovementFields can be used to create moving objects, such as busses, trains or other vehicles or transportation mechanisms.

While static nodes are directly created by the user, the NS-Mapper can also create an arbitrary number of random nodes. Random nodes are placed according to a given random node placement strategy. Another strategy concerns random node movement. Any node, independently whether it was placed by the user or created randomly, can be chosen to move according to a random node movement strategy. For static nodes, this allows random movement with a fixed start point in all simulations of a scenario.

Apart from the setup and the movement of mobile nodes, also data traffic has to be modelled for a scenario. The user can create arbitrary CBR and/or several different TCP traffic patterns between static and/or random nodes. Furthermore, additional traffic can also be generated automatically by NS-Mapper.

Further information


Get the NS-Mapper

Installation


At first you have to get the NS-Mapper tar file and unpack it into your favorite directory. Then download the xstream Jar package from the link. Copy the xstream package to the NS-Mapper directory and execute:


jar i NSMapper[Version].jar xstream[Version].jar


This adds xstream to the dependencies of the NS-Mapper. Then start the NS-Mapper with:


java -jar NS-Mapper[Version].jar


The source code of the NS-Mapper is in the Jar file .


Download


Media

Screenshots



Video







CSS structure is under the Creative Commons License


2007 Emanuel Eden