Research On Distributed Dynamic Navigation Method In Wireless Sensor Networks

Wireless Sensor Networks (WSNs) are multi-hop networks comprising large amount of WSN nodes with sensing and wireless communication capability, also limited energy. With the development of semiconductor electronic technology, WSNs have been seen as a novel way for people to perceive the physical world and widely used in production, medical treatment, environmental monitoring and military fields and so on. But just monitoring is not enough for system in most cases that also need feedback and control. Combined with mobile robot technology, the system can make WSN provide global map information and navigation service for mobile robot, which will take advantages of both technology and enhance operation efficiency. It is always a research focus that how WSN would monitor environment and adjust to dynamic dangerous zone to guarantee the short distance and security of navigation path, while balance the energy of network and save communication cost.This paper focuses on mobile robot dynamic navigation in WSN without location information. Considering deficiencies of existing potential field algorithms on dynamic adjustment, a distributed dynamic path planning algorithm is proposed with a gradient function including factors of path length and security. When the network detects dangerous events, it can adjust the potential field to provide optimal path planning for every node based on their local information. Experimental results show that when environment changes, the proposed algorithm can make WSN plan shorter path, effectively save communication cost and deal with path security flexibly according to different system requirements.To reduce the flooding messages during gradient potential field construction and communication cost for adjustment to dynamic dangerous zone, a backbone-based roadmap algorithm for path planning in WSNs is proposed, which restricts distributed dynamic path planning running only on the roadmap network. The performance and characters of the roadmap network constructed by the proposed algorithm is analyzed by theoretical analysis and simulations. Results show that it can effectively reduce redundant nodes in roadmap network while keep the connectivity between minimum dominating nodes within three hops, save communication cost and have a good dynamic respond.At last, based on RSSI measuring, two approaching methods with adjustable step length are proposed to solve the problem that how mobile robot would approach to the nodes on navigation path. The first one is a head-searching method. The second one is a fuzzy control method assisted by last-hop node which also considers the situation that mobile robot detects dangerous nodes during its approaching course. Experiments on software and hardware platform verified the validity and feasibility of the proposed methods.