Robust dynamic reprogramming of wireless sensor networks

Rashmi Parthasarathy

A Wireless Sensor Network (WSN) is a resource constrained network with active research in progress. WSNs deployed in remote locations are comprised of a set of devices that sense the surrounding environment and route back the data to the central repository. Some of the main challenges that WSNs face are reliability, flexibility and security. Often the network needs to be updated with error free software. When WSNs deployed in remote locations need to be reprogrammed, environmental conditions often make it impossible to physically retrieve them. Over the Air Programming (OAP) plays an important role in achieving this task. OAP is a method of distributing new software updates to devices that are not physically accessible [1]. Several platform dependent protocols have been developed for OAP. These protocols are mainly dependent on hardware architecture. As a result, any change to the hardware requires modifications to the protocol. In this work we redesign part of Deluge [2] and implement it on iMote2 platform. This work has been developed specifically for our research project called OASIS. OASIS monitors volcanic activity on Mount St. Helen. As the environmental conditions on the mountain make it difficult to access the nodes, we developed an OAP protocol to reprogram the network. We faced many challenges during the development like file system management, boot loader and Java tool chain. As the environment is harsh and rugged, it is vi critical for us to ensure that OAP is robust. In the context of OAP in WSNs, a system is said to be "robust" if it is capable of successfully updating itself with the latest software over the air. We evaluate performance of our implementation of Deluge on a real test bed of sensors. Our results show that Deluge successfully reprograms the nodes in the network. We also present the results of reprogramming partial network deployed on Mt. St. Helen. We further evaluate the robustness of Deluge by having different version numbers in the system and also by performing multiple binary image downloads. Our results show that Deluge is 100% robust in the lab environment. This work also contributed to developing a GUI to monitor the reprogramming of the network in OASIS.

Robust dynamic reprogramming of wireless sensor networks

Files and links (1)

Abstract

Metrics

Details