Bilkent University
Department of Computer Engineering
S E M I N A R

Constructing Efficient Bluetooth Scatternets

 

Tağmaç Topal  

Master Thesis Presentation

 Supervisor: Asst. Prof. Dr. İbrahim Körpeoğlu

 

Among various technologies for short-range wireless networking, Bluetooth has received a particular attention from users as well as from vendors. It is the main technology that supports wireless personal area networking. Bluetooth Special Interest Group (SIG), which is a consortium established to develop and promote the technology, produces the specifications of Bluetooth standards. The Bluetooth standards specify the building blocks to construct Bluetooth networks of arbitrary size, i.e. scatternets, but they do not specify the policies and algorithms that can be used in constructing these scatternets. There may be various approaches for forming Bluetooth scatternets which will result in different topologies for the same set of nodes. In this thesis, we first define and provide some performance metrics that can be used to evaluate various scatternet topologies that can be the output of different scatternet formation algorithms. Then, we provide a new Bluetooth scatternet construction algorithm that differs from other algorithms in that it also considers the traffic pattern of users (i.e. traffic requirements of nodes among themselves) in establishing a scatternet. Then we evaluate the performance of our algorithm through simulations by observing the properties of the constructed scatternets. In a scatternet that is the result of our algorithm we particularly look to the weighted average shortest path lengths that traffic flows follow, the ratio of satisfied users, and the utilization of the scatternet capacity. The results show that we can achieve a good ratio of satisfied users, a high network utilization, and a reasonably small value for average path lengths using our algorithm. The algorithm is currently centralized, but can be extended to a distributed one in the future.


 

DATE: January 19, 2003, Monday @ 16:00
PLACE: EA-502