Odysseus

We want to turn every citizen of a city into a mobile environmental sensor. This project will design a simulator to prove the basic premise of the idea. A network of intelligent mobile sensor nodes, well-connected ‘drop-off’ nodes, storage and control nodes combine to record, disseminate and control data flow within the system. The system has built in feedback  mechanisms that react to detected phenomena, with sophisticated mechanisms for sharing and reporting of data, intelligently spreading global queries to drive local operations, such as requesting more data in a particular area, causing nodes to increase sample rate or alerting users to a phenomenon.

Create a simulator to include:

• Simulated Movement patterns based on real data (see ZebraNet, CAR & sCAR,
Random Walk, Seal2Seal)
◦ must be pluggable: i.e. future developers must be able to plug in a new
movement patterns without affecting the other aspects of the project
◦ Evaluate the quality of data from UCD WaveLAN access points use, as a
source to derive movement patterns from.
• Programmable nodes - (see OMNET++, NS2, NS3)
◦ nodes must be pluggable, as above, so that nodes with different features
can be modelled
◦ must support multiple node types
◦ must support simulation of different wireless comms types (Bluetooth,
WiFi, Zigbee)
◦ for this particular project, nodes must emulate some type of
environmental sensing, such as particulate levels (pollution), which is the
data disseminated about the network
• Implement communications algorithms:
◦ Existing delay tolerant algorithms include: Flooding, Epidemic (various),
Data Mule, Data Ferries, CAR and sCAR, Spray and Wait, etc.
◦ Some or all of these should be implemented to test the system, and can
be compared against previous results using these algorithms.
◦ Develop a simple reactive algorithm for nodes, that reacts to
sensed environmental data and instructions from controlling
nodes, by increasing or decreasing level and frequency of
sensing.
Much of the simulation software is available in the form of OMNET++, NS2 and NS3, one
of these could be adapted to account for movement patterns, or an existing adaption
could be used. The emphasis of this project is on the realistic movement of nodes and
the reactive nature of the system to environmental phenomenon.