Design And Implementation Of Management And Fault Detection In Wireless Sensor Networks

ABSTRACT:Wireless Sensor Networks (WSNs) are composed of a large number of power-limited sensor nodes which distributed randomly by self-organization distributed protocols. With the rapid development of modern microelectronic technology, wireless communication technology and computer network technology, WSNs have become one of the most important technologies in information acquisition. WSNs can support a wide range of applications of our daily life in future, such as environmental monitoring, public safety, medical, transportation and military. However, as the constraints of the real environment, applications of WSNs share similar disadvantages such as high mobility, interference and prone to failure. The network management of WSN is an important requisite to guarantee the stability, reliability and high efficiency in WSN. The operation of a wireless sensor management system is greatly affected by different interrelated factors such as network traffic flows, network topologies, and communication protocols. These natures make the management of WSNs a more difficult task than that of traditional network.In this paper, firstly, the existing management frameworks and their typical systems are summarized, such as centralized structure, distributed structure and hierarchical structure. Then taking the feature of energy conserving, reliability and expansibility into consideration, a lightweight simple network management protocol (LW-SNMP) for the wireless sensor network was designed, including its reference model, message formats and detailed protocol design. Then a lightweight hierarchical wireless sensor networks management system (LHWMS) was proposed, including a PC manager, cluster proxies, node agents. And then, the detail of design and the key mechanism was given.Secondly, LHWMS was implemented in micro sensor node MSR6680, which is a platform designed by our lab. The functions and performances of LHWMS were verified and analyzed by building an indoor test environment. The test results show that LHWMS can supply the demands of wireless sensor networks and have a good performance in stability, effectiveness of memory, expansibility and integration with traditional networks. This paper was concluded by summarizing the completed works and proposing future works. The work of this paper is supported by the National 863 Program (No.2007AA01Z241) and CNGI Program (No.CNGI-09-03-05).