The Research On Tolerant Routing Algorithms In Sensor Networks

Wireless integrated sensor networks, which include collecting, managing data and communication, can be used widely in many areas and this makes it a new researching area. But the low price makes that sensors usually use low-quality parts and this makes nodes has high possibility of failure, especially considering badly working situation of sensors. Therefore, when we research the routing algorithms, the tolerance of routing is necessary. At the same time, this thesis emphasizes the tolerance-routing algorithms.This thesis classifies the usual routing algorithms and analysis their tolerance. These routing algorithms are classified into four groups and We analyse the characters of these algorithms. These results are helpful to design more effective routing algorithms. Based on low complexity and no-aim of flooding, we propose a new fault-tolerance routing algorithms based on flooding. New algorithm limits flooding in a finite area and packets record the node-sequence forwarding the packets. Sink receives all the packets and selects several paths as the forwarding path. Theoretic analysis and simulation results show that new algorithm has low complexity and has better performance. But no-aim makes it difficult to improve the performance. Therefore, this thesis proposes a new algorithm with load-balance. Each node sends packets including its position information to sink and sink collects all these packets. When sink knows all the information of network, it uses more effective method to construct multipath and select suitable paths to forward data. Simulation results show that new algorithm has better fault-tolerance ability. Based on the concept of mesh, this thesis proposes the new algorithms when sub-mesh is connected. The algorithm divides the whole area into some sub-mesh and these sub-meshes are adjacent. The algorithm has linear complexity and it's very easy to find a path and simulation results show that the algorithm has better performance.