STUDENT THESES
Title: A Bit-Map-Assisted Energy-Efficient MAC Scheme for Wireless Sensor Networks (presentation slides | full text | code)
Name: Jing Li
Advisor: Dr. Georgios Y. Lazarou
Defense Date: March 30th, 2004
AbstractThe low-energy characteristics of Wireless Sensor Networks pose a great design challenge for MAC protocol design. The cluster-based scheme is a promising solution. Recent studies have proposed different cluster-based MAC protocols. We propose an intra-cluster communication bit-map-assisted (BMA) MAC protocol. BMA is intended for event-driven applications. The scheduling of BMA can change dynamically according to the unpredictable variations of sensor networks. In terms of energy efficiency, BMA reduces energy consumption due to idle listening and collisions. In this study, we develop two different analytic energy models for BMA, conventional Time Division Multiple Access (TDMA) and energy efficient TDMA (E-TDMA) when used as intra-cluster MAC schemes. Simulation experiments are constructed to validate the analytic models. Both analytic and simulation results show that in terms of energy efficiency, BMA performance heavily depends on the sensor node traffic offer load, the number of sensor nodes within a cluster, the data packet size and, in some cases, the number of sessions per round. BMA is superior for the cases of low and medium traffic loads, relatively few sensor nodes per cluster, and relatively large data packet sizes. In addition, BMA outperforms the TDMA-based MAC schemes in terms of average packet latency.
Title: On the Modeling of TCP Latency and Throughput (presentation slides | full text)
Name: Dong Zheng
Advisor: Dr. Georgios Y. Lazarou
Defense Date: July 8th, 2002
AbstractIn this thesis, a new model for the slow start phase based on the discrete evolutions of congestion window is developed, and we integrate this part into the improved TCP steady state model for a better prediction performance. Combining these short and steady state models, we propose an extensive stochastic model which can accurately predict the throughput and latency of the TCP connections as functions of loss rate, round-trip time (RTT), and file size. We validate our results through simulation experiments. The results show that our models predictions match the simulation results better than the Padhye and Cardwell's stochastic models, about 75% improvement in the accuracy of performance predictions for the steady state and 20% improvement for the short-lived TCP flows.
Title: Scheduling for Proportional Differentiated Services On the Internet (presentation slides | full text)
Name: Manimaran Selvaraj
Advisor: Dr. Georgios Y. Lazarou
Defense Date: Oct. 25th, 2002
AbstractProportional Differentiated Services can be provisioned in terms of bandwidth, delay, or packet loss. Several studies contributed schedulers and packet droppers that achieved proportional bandwidth, delay, or loss differentiation. However, all these schemes differentiated in terms of only one of the three metrics. A simple, unified, scalable, and robust scheme to simultaneously control all three metrics was felt important.
By controlling just delay and packet loss, proportional differentiation can be achieved in terms of all three metrics. A robust adaptive scheduler for proportional delay differentiation services is presented. Proportional services are further policed by a class based packet dropper. The combination of the adaptive scheduler and the packet dropper treats different traffic classes proportionally in terms of all three metrics. Simulation experiments show that regardless of the network traffic characteristics, our scheme can effectively differentiate services in terms of bandwidth, delay, and loss simultaneously.
Updated: December 16, 2003