Project Group MobiDAgg

Mobile Data Aggregation, Winter Term 2014/15 (canceled)

Motivation / Use Case:

In the project group we want to help car drivers in a big city to find a free parking lot.
The goal is to implement an App which searchs a free parking lot which is close to the current position and states also the route to it.

Scenario 1:

Consider a scenario in which we deploy stationary sensors (f.i. cameras or traffic-lights) that sense free parking lots and send the information
to a central server. Also the position-based query by a car-driver is send to the server. This is the state of the art.
A system in a big city will generate a high workload on the server thus, response time gets delayed at peak times resulting in a Big Data Problem.

Scenario 2:

Referring to the internet of things we can think of cars sensing free parking-lots itself. They can also communicate to each other and to close
stationary sensors. Queries from a user gets send through the network without usage of a central server. If there is no free lot in the closer
environment, they forward the query to their local neighbourhood.

Scenario 3:

Now think of a scenario where there are no sensors placed stationary, there are only the sensors which are placed in cars to observe
the environment around them. Free parking-lots gets observed by cars passing by and distribute the information only locally to each other.
Is a system without stationary infrastructure possible in this manner?

Bigger picture

Investigate different variants for distributed, position-based aggregation of "mobile" data.
There will be a networked system based on stationary & mobile sensors, in which mobile agents can get informations about their local environment.

GOAL: Sending correct responses to a query, which makes as less as possible use of infrastructure (stat. sensors / communication network), memory and CPU-time. 

Tasks:

  • Investigate in different variants for position-based data aggregation
  • Develop distributed Algorithms
  • Analyse correctness, communication volume and response times
  • Implement an simulation of the environment

Your choice: Working theoretically and/or practically 

Materials:

Slides from project presentation: ( pdf

Questions / Discussions?

get in contact: Manuel Malatyali, malatya@mail.upb.de