Ph.D. Dissertation Defense - Deepa Phanish

TitleOptimal Clustering and Inter-cluster Routing in Large-scale Wireless Ad Hoc and Sensor Networks

Committee:

Dr. Edward Coyle, ECE, Chair , Advisor

Dr. Randal Abler, ECE

Dr. Ragupathy Sivakumar, ECE

Dr. Henry Owen, ECE

Dr. Mostafa Ammar, CoC

Abstract:

A large number of small sensing devices connected together with wireless transceivers can be used to effectively monitor various environments. However, these devices have limited energy reserves. Clustering of nodes can significantly save energy by aggregating local information to reduce the total communication cost. For energy-efficient clustering in large-scale ad hoc and sensor networks with many levels, we optimize the design variables: 1) The number of levels, and 2) The number of nodes operating at each level. By considering variable aggregated data sizes, we have a unified stochastic model for studying different classes of applications. When wireless networks are hierarchically clustered, the subsequent long-range communications from the central clusterheads cause large-scale interference and energy-hole problems around them. It is thus better to have packets forwarded via short-range multi-hop routes between the different levels. To discover the most optimal routes, we have analyzed the paths that minimize the inter-cluster routing delay. A low-delay, energy-balancing distributed algorithm for routing across clusters and between levels is developed, which outperforms shortest path routing in high throughput networks. Also, as a step towards achieving a large-scale testbed, we discuss the design, development, and deployment of an inexpensive, power-efficient, clustered, and scalable wireless sensor network. Overall, the major outcomes include: (1) A synchronized sensor network that can be remotely operated. (2) An NS3-based large-scale sensor network simulator with novel clustering and routing algorithms. (3) Innovative stochastic models for the analysis of clustered wireless networks.

Event Details

Date/Time:

  • Monday, November 13, 2017
    9:00 am - 11:00 am
Location: Room 3402, Klaus