Research And Design Of Network Reprogramming Technology In Wireless Sensor Networks

A Wireless sensor network (WSN) is a multi-hop, self-organizing network, consisting of a large scale of wireless sensors with constrained computation, memory and communication capabilities, but interacting with the physical environment. WSNs are easy to deploy, and have the features of self-adaption and high fault tolerance rates. WSNs have been widely applied in environmental surveillance, structural health monitoring, forest protection, intelligent medical and battlefield information collection, etc.After the first deployment of a wireless sensor network, with the changes of user requirements and technology advance, network administrators need to update the configuration insensor nodes;Whereas sensor nodes can be deployed in a very harsh environmentwhere physically reaching all nodes is either impractical (e.g. volcanoes, oceans) or detrimental to original environments(e.g. birds living habits monitoring). Network Reprogramming is therefore introduced, which can update a code image through wireless communication, aiming to improve the usability of wireless sensor networks and reduce network maintenance costs. This paper studies network reprogramming for wirelesssensor networks, where main works as follows:(1)Present wireless reprogramming technologies are analyzed and discussed. The framework of wireless reprogramming has been depicted, where the challenges of existing wireless reprogramming technologies in WSN have been discussed. Classic operating systems in WSNs, have been introduced and analyzed to compare the pros and cons. In addition, the core technology of the network reprogramming, i.e. code dissemination, has been introduced with classification for appropriate analysis of current code dissemination protocols.(2)Existing code dissemination protocols transmit redundant code images when disseminating to target nodes. A Multicast Tree based Code Dissemination (MTCD) protocol is therefore designed for wireless sensor networks. The MTCD establishes a dissemination tree route between the base station and a target node to reduce the number of involved nodes. As a result, the MTCD reduces redundant data transmission and network energy consumption. Simulation results via the TOSSIM platform show that the MTCD outperforms Deluge, the de facto standard protocol for code dissemination in TinyOS, in terms of completion time and packet transmissions. (3) A link quality based code dissemination protocol is presented. To improve the performance of code dissemination in a lossy network, we measurelink quality values between nodes calculating the impact of each node. To reduce the overhead of control messages, ADV messages have been broadcast periodically to maintain the state of neighboring nodes. Simulation results via the TOSSIM show that our protocol outperforms the Deluge in terms of network completion time and message transmissions especially in the case of poor network quality.(4) A wireless reprogramming system has been established using Micaz nodes and Mib520base station from the MEMSIC. The architecture of the wireless reprogramming system has first been analyzed, consisting of the functions of the management node, the base station and sensor nodes during wireless reprogramming. An approach to data interaction between the management node and the base station node is designed, in order to implement the transmission of code images and management commands through a serial port. The solutions on code dissemination processes, such as code image storage, code image reloading, etc. are analyzed in details.