Home'."\n"; } else echo "\t\t".''."\n"; if ($title=="Interactive" || $title=="SceneGenerator" || $title=="Applet" || $title=="mobile") { echo "\t\t".''."\n"; } else echo "\t\t".''."\n"; if ($title=="Downloads") { echo "\t\t".''."\n"; } else echo "\t\t".''."\n"; if ($title=="People") { echo "\t\t".''."\n"; } else echo "\t\t".''."\n"; if ($title=="Contact") { echo "\t\t".''."\n"; } else echo "\t\t".''."\n"; if ($title=="Documentation" || $title=="GUI" || $title=="General") { echo "\t\t".''."\n"; } else echo "\t\t".''."\n"; ?> '; echo "\t\t".''."\n"; echo "\t".''; echo "\t".''."\n"; echo "\t\t".''."\n"; echo "\t".''; } // Untermenue 2 if ($title=="GUI") { echo "\t".''; echo "\t\t".''."\n"; echo "\t".''; echo "\t".''."\n"; echo "\t\t".''."\n"; echo "\t".''; } ?>


1. Introduction

Today we see an increasing number of portable computers as notebooks and PDAs which are equipped with wireless interfaces. These wireless interfaces like IEEE 802.11(WLAN) or IEEE 802.15.1(Bluetooth) offer an Ad-Hoc mode that is independent of any infrastructure. Wireless networks which are spontaneous, not planned, self organized and built up as required are in reach. Computer scientists name this type of network Ad-Hoc network. The most important technical challenges include the wireless communication at a restricted frequency range, battery life time, and limited CPU power.

2. Ad-Hoc network

According to Charles Perkins an Ad-Hoc network is an infrastructureless, wireless computer network whose end systems are mobile in many cases and in which only certain subsets of end systems can reach each other directly at a time. We name these end systems network nodes or simply nodes. To transport data between arbitrary nodes, all nodes within the network cooperate and some of them relay data. Relays are nodes that forward packets based on the destination address. The result is that nodes which are not in a direct reach of each other communicate by forming longer paths via available relays. In figure 1 the end systems 33, 31 and 43 relay for the communication partners colored in blue and green. Ideally Ad-Hoc networks organize themselves on-demand without a further interaction of the user. Compared to the infrastructure mode Ad-Hoc Networks offer cost advantages and operate without deployment and planning.

fig.1: Illustration of an Ad-Hoc network with links and relays

However the question is, when does it pay off to try communication over such networks? It is obvious, that it makes little sense to form a long-lasting communication connection if on the one hand the probability for the successful set-up is very small and on the other hand established connections were constantly interrupted by the mobility of the intermediate stations. Estimations of connectivity are helpful to assess Ad-Hoc networks for a certain application in a specific geographical surrounding field.

2.1 Types of Ad-Hoc networks

We distinguish between two types of Ad-Hoc networks.

On the one hand, there are static networks in which all nodes are stationary. Stationary Networks can be planned or random. An example of a planned Ad-Hoc Network is the placing of nodes for sensing an area or object by individual placed network nodes. The placing of the sensors with a wireless interface is driven by the necessity to measure interesting data. Random placement of nodes happens when we throw sensors out of an airplane for the sensing purpose in an area that is hardly accessible.

If at least a part of the network nodes move, then the network is described as mobile. Mobile means that nodes stay connected to the network during the movement and an existing communication remains unchanged. The end systems are able to initiate a communication during the movement and remain reachable.

3. ANSim

ANSim simulates Ad-Hoc networks and examines connectivity and is especially suited for statistical simulation with long simulation times and large networks. Therefore, ANSim abstracts from the lower layers and implements above the transport layer a mechanism similar to RPC(Remote Procedure Call) for the exchange of data between the nodes. ANSim checks if the location of the nodes Ni and Nj allows to send data from Ni to Nj. Whether such transmissions will suffer from collisions or transmission errors is not taken into account. Also transmission delays caused by busy transmission media in the real world are ignored. New links and paths between nodes can be used to one hundred percent and without any delay. ANSim supports the simulation of static Ad-Hoc networks and Ad-Hoc networks with mobile nodes by two different operational modes:

3.1 Static mode

According to the input parameters of the user ANSim generates snapshots using a random process by placing nodes within the field. Single snapshots are evaluated as one experiment. After completion of the evaluation and data elevation a new snapshot is generated. A number of such snapshots create a random sample.

3.2. Mobile mode

In mobile mode snapshots are generated periodically where nodes move according ot the mobility model. The result of these snapshots are accumulated as a random sample over time. The choice of the mobility model plays an important role and simulation results vary very much with changing mobility model.

3.3. The GUI

3.3.1. Simulation Input

N nodes are distributed on a grid with the help of a random function (scenario). Two nodes are selected and examined whether they are connected. With rising number of scenarios the accuracy of the result rises.
On the left side you can see the input block for the simulation parameters. The right side shows the text output block with the numerical calculated result values. Within the lower area (graphical output block) single scenarios are visualized.

The left side contains the input block of the simulation parameters:

The buttons to control the simulation:

The additional input parameters of the static mode are:

The additional input parameters of the mobile mode are:

fig.2: ANSim 4.00

3.3.2. Numerical Output

The output block illustrate the numerical calculated result values. The result values of the started or stoped simulation are:

3.3.3. Graphical Output

The graphical output area shows a single snapshot. The source node (or sender) is coloured in green and the destination node (or recipient) is coloured in blue. The green source node tries to find the shortest possible path to the blue destination node. The algorithm ends, as soon as the target node is reached (connectivity) or the tree cannot be completed further (no connectivity). The resulting spanning tree is displayed in black (SPT from green to blue). All available links between nodes are coloured in grey. In the example the length of the path is 3 hops. The two circles of the source and destination node illustrate the transmission range

To get insight information of an object during simulation runs you can click into the graphical output area, for example on a node. A new window will pop up, that conains detailed information about that object.

fig.3: Window that opens, after clicking on the connection between node 1 and 8.

3.4. Configuration files

Besides the manual configuration with the help of the GUI, it is possible to use configuration files as input for the simulations. Configuration files have the extension .anm. Compared to the previous versions the configuration syntax and options are extended substantially. Configuration files specify the fields as polygons. An unlimited number of obstacles blocking or being transparent to the wireless transmission may be placed within the field. Additionally hotspots which serve as final destinations for the nodes movement can be added. For further information see the EBNF notation and an example the IU campus.

fig.4: Graphical Output of the configuration file IUcampusHotSpot

If you like to provide us with some ANSim configuration files, we will be happy to add them to the applet. Just send us the .anm file by e-mail.

3.5. Trace files

All built-in movement models in ANSim like Random Direction or Gauss Markov move nodes in a pseudo random fashion. To use recorded real life traces ANSim supports trace files specifying start position, target position, begin of movement and the velocity of the node. The number of position changes for a node is unlimited and limited only through memory constraints. A Trace File is an xml file following the sppmobtrace schema notation and is used ideally together with a Configuration File. The syntax of the configuration file is defined in EBNF notation too. If you are planning to use a configuration file a sample trace with some runners on the Hamburg Marathon might help (If you plan to use it locally change the Url of the tracefile entry in the configuration file to a local directory where the tracefile is located).

fig.5: Two runners on the Hamburg Marathon

4. Questions

If any questions are left, you are welcome to contact us via the contact form or via e-mail