In English

Deterministic Collision Free Communication Despite Continuous Motion

Rosa Aga
Göteborg : Chalmers tekniska högskola, 2015. 45 s.
[Examensarbete på avancerad nivå]

Mobile ad-hoc networks (MANETS) require robust and carefully designed Media Access Control (MAC) solutions to cope with all the inherent problems of wireless communications. Today, quite a lot of solutions, that promise to address these challenges, exist; nevertheless, their practical applicability to a particular situation has not been examined in real world scenarios and hence, many of them have remained on paper and in simulations. We believe it is important to evaluate the performance of the protocol in simulators and real platform. In this study the performance of deterministic collision-free despite continuous mo- tion (DCFA) algorithm has been evaluated in simulators and on a real platform. For the evaluation, the DCFA algorithm has been implemented in tinyos-2.1.1 using the nesc programming language and a special python script to handle the topology for simulation on TOSSIM simulator; and the C++ programming language for simulation on the OMNET++ and Castalia simulators. For the evaluation of the DCFA algorithm on simulators, we have used TOSSIM, OMNET++ and the Castalia simulators. To evaluate DCFA on a real platform, we have implemented it on Crossbow's MicaZ platform motes with a CC2420 radio. The performance of the DCFA algorithm has been measured in terms of throughput and success-rate. Since the evaluation of the DCFA algorithm is on mobile nodes, we have considered two mobility models to evaluate the performance of the DCFA algorithm. Two mobility models have been designed in this study in order to challenge the DCFA algorithm and evaluate it. When the nodes move in parallel and reach the boarder of the grid, the rst mobility model lets the node move to the next line of the the grid and continue in the opposite direction; the other mobility model lets the node go back to the beginning where it started the mobility and continues in same direction. From the study, we have observed that the DCFA algorithm does not have self stabilisation when unexpected things happen in the network.

Publikationen registrerades 2015-04-14.

CPL ID: 215173

Detta är en tjänst från Chalmers bibliotek