Evaluating Buffer Management Algorithms for Delay Tolerant Wireless Sensor Networks
(English)Manuscript (preprint) (Other academic)
One of the challenges in Delay Tolerant Wireless Sensor Networks (DT-WSN), is to handle situations where the available buffer space is insufficient. This is the buffer management problem. In this work, we the buffer management problem, with focus on algorithms to decide what data to discard when buffers are full. We simulate a wide range of network configurations, using a full factorial experiment design. The simulations let us investigate both the relative performance of four buffer management algorithms and the effect of different factors on buffer management performance. The four algorithms we investigate are the two well-known algorithms FIFO and Random Discard, and two Quality of Information based algorithms: an exponential moving average forecasting-based algorithm and the SmartGap algorithm. Our results indicate that certain factors, including mobility model and routing algorithm, have larger influence on relative performance than factors such as link speed and lifetime of data. We also find that the SmartGap buffer management algorithm provides significantly better QoI compared to the alternative algorithms.
DTN, WSN, DT-WSN, SmartGap, Simulation, ANOVA
Research subject Computer Science
IdentifiersURN: urn:nbn:se:kth:diva-184497OAI: oai:DiVA.org:kth-184497DiVA: diva2:916209
QCR 201605302016-04-012016-04-012016-05-30Bibliographically approved