Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
en:projects:msp:licos6 [2008/11/25 11:48] behn |
en:projects:msp:licos6 [2009/09/11 13:03] (current) bardet |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| - | Master Semester Project: 2008-2009 | + | Master semester project: |
| - | ==== Active Sensing ==== | ||
| - | \\ | ||
| - | |||
| - | **Description:** | ||
| - | |||
| - | Sensor networks can be used for detection tasks. If the sensor nodes are battery driven, they are subject to severe power constraints. Wireless communication is using most of the power of such sensor nodes. | ||
| - | |||
| - | Detection tasks (detecting a certain state of the environment) can be modelled using statistics. In the past, various distributed detection problems have been formulated and studied. However, in most of these settings, all the nodes are constantly communicating. However, one can imagine that different users of the network have different requirements concerning the precision of detection. In such a case, it is more meaningful to let the user request messages from the individual nodes. This way, the user can tune himself the tradeoff between communication cost and precision of detection. | ||
| - | |||
| - | In this project, we will assume that the user is actively requesting information from the sensor network in order to estimate the state of its environment. One can imagine that the user can request a piece of information from every node in the network. Every request has a cost (modelling the power used for the communication) and a benefit (modelling the statistical use of the piece of information). The student will be asked to model such a scheme mathematically, to analyze different requesting strategies and decision rules and to look for an optimal strategy. | ||
| - | \\ | ||
| - | |||
| - | **Objective:** | ||
| - | | ||
| - | Find schemes for detecting events when the user is actively querying the sensors, while the sensors do not take action on their own. | ||
| - | \\ | ||
| - | \\ | ||
| - | |||
| - | **Prerequisites:** | ||
| - | |||
| - | * Solid background in probability and statistics | ||
| - | * Hypothesis testing / detection | ||
| - | |||
| - | \\ | ||
| - | ** Benefit ** | ||
| - | |||
| - | The flavour of the project will be more theoretical than practical. You will learn about existing results in distributed detection and probability. You will have a fair amount of freedom and will be expected come up with own ideas. | ||
| - | \\ | ||
| - | \\ | ||
| - | |||
| - | |||
| - | **References:** | ||
| - | \\ | ||
| - | |||
| - | **Contact:** | ||
| - | |||
| - | |||
| - | Etienne Perron (LICOS) * Email: etienne.perron@epfl.ch * Office: INR-033 * Tel: 36457 | ||
| - | |||
| - | \\ | ||
| - | |||
| - | **Supervisor:** | ||
| - | Prof. Suhas Diggavi | ||
| - | |||
| - | \\ | ||
| - | |||
| - | [[en:projects:mastersemester:msp|back to master semester projects]] | ||