Mobile Node Tracer (MoNoTrac)

The Mobile Node Tracer (MoNoTrac) is a simulation environment written in Java to generate mobility traces of mobile entities considering geographic data supplied by OpenStreetMap. Currently, we are working on a beta version of the software to make it available under the GPL on this site.

Motivation

The simulation of wireless networks with mobile nodes such as mobile ad-hoc networks and vehicular networks requires the specification of the movement of these nodes. For this, mobility traces are needed during the simulation. MoNoTrac is a mobility trace generator based on OpenStreetMap. The simulation area is selected from a real city map and the movement of nodes of different kinds, such as cars and pedestrians, is simulated on this geographical data. Common mobility models are modified in a way that they respect the geographical and structural restrictions. MoNoTrac produces an XML-based meta-format which may be converted to formats required by popular network simulators such as ns-2 and OMNeT++. 

The Frontend and Workflow

MoNoTrac comes with a graphical frontend, that may be used by researchers to create the desired scenarios. This is achieved interactively as depicted in the picture below. In order to create a scenario description, the researcher selects the simulation area using drag and drop features from the rendered OpenStreetMap data. Then, the desired number of nodes are added and configured. The used mobility model as well as global simulation options, such as the simulation runtime is specified. Further scenario configuration, such as the position of the nodes, the simulation runtime, and the number of repetitions, may be set.

Upon the completion of the scenario description, the MoNoTrac core starts the simulation and generates the mobility traces. After the simulation, the traces are available to the user in an XML-based meta-format. This meta-format may be translated to simulator specific trace files using MoNoTrac's translators. Currently, translators for ns-2 and OMNeT++ are available for MoNoTrac. 

Mobility Models for Geo-Data

Two different mobility models that consider the OpenStreetMap geographical data have been developed for MoNoTrac so far. Geo-Random-Walk is an adaption of the Random Walk model, which considers for the movements paths the restrictions of the geographical data. Routed-Geo-Waypoint is an adaption of the makes use of an routing algorithm to find paths to destination waypoints in the simulation area.

Results

In order to evaluate MoNoTrac and the developed mobility models, we created scenarios with simulation areas from the Berlin map. After the simulation, a python script plots the generated mobility traces in order to visualize the spatial distribution of the mobile nodes during the simulation. The pictures shows the simulation area of the scenario on the left and the resulting spatial distribution on the right. A total of 50 nodes were simulated with the Geo-Random-Walk model. The plot shows that only streets have been visited by the nodes and therefore the geographical data was respected by the mobility model.

Future Development

Currently, we are working on a beta version of MoNoTrac with the described features to make it available on this website. For this, the frontend has to be elaborated and the developed mobility models have to be investigated in depth stochastically and analytically to get an better understanding of them. Next, we will implement additional mobility models. We will focus on typical movement patterns of vehicles and pedestrians in city areas in scenarios, in which wireless interaction is common. Of interest to us are group mobility models that describe movement patterns for larger groups of nodes, such as a guided tourist tour. We will also adapt mobility models that simulate mobile nodes in artificial city areas, such as the popular Manhatten Grid and the City Section models. We will also extend MoNoTrac by adding a simulation animation function. This way, generated mobility traces can be visualized on the simulation area.