Research On Dual-Homed Fault-Tolerant Routing For Wireless Sensor Networks

Wireless sensor networks are a kind of distributed wireless network which consists of a large number of sensor nodes. Nowadays, more and more researches have focused on the Wireless Sensor Networks (WSNs). Many scientific magazines and business magazines have selected WSNs as one of the greatest technologies which will change human society greatly in the 21st century. But sensors usually use low-quality parts and it makes nodes have high possibility of failure, especially considering badly working situation of sensors. Therefore, it is necessary to research on the fault-tolerant of WSNs.This paper focuses on fault-tolerant routing for WSNs. According to the differences of fault-tolerant methods, the paper classifies the usual routing algorithms of WSNs and analyzes their characteristics and performance on fault tolerance. Then based on dual-homed fault-tolerant routing technology, we design two fault-tolerant routing algorithms, including FTLEACH protocol and grade of fault-tolerant routing algorithm.LEACH protocol is a typical two-layer routing protocol in WSNs. LEACH is easy to implement, but it isn't perfect on dealing with fault-tolerant problems. To improve it, we propose FTLEACH, a fault-tolerant routing protocol for WSNs. After simulation by NS2, results show that FTLEACH decreases loss probability, increases throughput of network and has a better fault-tolerant performance than LEACH.FTLEACH is designed for two-layer network in WSNs, which just uses a sort of dual-homed fault-tolerant routing technologies. Based on FTLEACH, the paper researches on multi-layer network in WSNs further, making uses of sorts of dual-homed fault-tolerant routing technologies, and proposes grade of fault-tolerant routing algorithm. According to different layers'different fault-tolerant requirements, grade of fault-tolerant routing algorithm employs dedicated 1+1 dual-homed routing, dedicated 1:1 dual-homed routing and shared 1:2 dual-homed routing at different layer of networks. Simulation results show that it not only achieves fault tolerance well, but also can improve resource utilization of network.