hpwren
HPWREN News

June 12, 2009

HPWREN-supported UCSD Master's student Edoardo Regini finished his "Resource Management in Heterogeneous Wireless Sensor Networks" thesis, and will graduate in Fall '09

By Edoardo Regini, UCSD

Heterogeneous wireless sensor networks such as HPWREN have environmental sensors located in remote and hard-to-reach locations far from the main high-bandwidth data links. The sensed data needs to be routed through multiple hops before reaching the main backbone. Some of the routing may be done by battery-powered nodes using license free radios such as 802.11. In this context, minimizing energy consumption and timely data delivery is critical to maintaining operational data links. A graduate student researcher supported by HPWREN, Edoardo Regini, recently finished his UCSD Master's degree thesis on a combined scheduling and routing mechanism for heterogeneous wireless sensor networks that addresses these issues. Tajana Simunic Rosing, UCSD professor in the Department of Computer Science & Engineering was his thesis advisor and committee chair.



The proposed solution is an energy efficient mechanism for scheduling and routing in heterogeneous wireless sensor networks such as HPWREN. It includes a distributed scheduling algorithm that allows a large portion of nodes to switch off the NIC thus saving energy. Scheduling is then combined in with the creation of a backbone of nodes in charge of delivering data to the proper destinations. By requiring no modifications to the MAC layer, the solution can be easily and quickly deployed on existing networks such as HPWREN where neither legacy devices need to be replaced nor firmware or drivers modified. Saving energy also lowers the cost of network maintenance by avoiding frequent and expensive replacement of batteries. The solution presented in this thesis achieves large power savings (50-60%) while delivering packets with low delay.

Link to thesis (PDF)

Full abstract of the Thesis

Heterogeneous wireless sensor networks (HWSNs) such as HPWREN have environmental sensors located in remote and hard-to-reach locations far from the main high-bandwidth data links. The sensed data needs to be routed through multiple hops before reaching the backbone. The routing is done by battery-powered nodes using license free radios such as 802.11. In this context, minimizing energy consumption is critical to maintaining operational data links.

This thesis presents a novel routing mechanism for HWSNs that achieves large energy savings while delivering data efficiently throughout the network. This mechanism sits on top of the unmodified MAC layer so that legacy network devices can be used, and expensive hardware/software modifications are avoided. Thus, our approach is inexpensive and easily deployable.

Our solution includes scheduling and routing algorithms. The TDMA based scheduling algorithm limits the number of active nodes and allows a large portion of nodes to sleep, thus saving energy. Since the algorithm is completely distributed, minimal (at join time) control packet exchange is required, and the nodes in sleep state can switch off the wireless network interface, thus minimizing power consumption. Simulations and measurements on a testbed network show that scheduling can achieve as much as 85 % power savings. Furthermore, results show that by limiting the number of active nodes, contention in the channel decreases and hence aggregate throughput increases by up to 10 %.

Scheduling is combined with the creation of a backbone of nodes in charge of providing connectivity to the network and delivering data to the proper destinations. Once part of the backbone, a node is required to stay awake continuously for a predefined amount of time. Since it is an energy expensive task, the nodes of the backbone are dynamically selected so that those nodes that have more energy available are more likely to become part of the backbone.

Simulations results on different scenarios show that the combined scheduling and forwarding backbone approach achieves up to 60 % energy savings per battery operated node and also have better performance when compared to existing techniques.

Main HPWREN web site (includes information for acknowledgments/disclaimers and feedback/contact)