Research On Wireless Sensor Network Location Based On Node Density Distribution

Sensor localization is a basic but important branch in the study of wireless sensor networks. The path-based localization scheduling for mobile beacons is a hot research topic nowadays. In such algorithms, an ordinary sensor node to be localized computes its location by using the messages received from the beacons. That’s why it is important to study where and when beacons transmit messages.In practical applications, sensor nodes are usually deployed in the monitoring area randomly, leading a non-uniform node distribution in the network. Existing localization algorithms fail to take the density of nodes in the network as a reference for localization, that is, they are insensitive to the distribution density. If an algorithm uses the same localization strategy in all the areas with different node density, many problems will arise, such as low localization accuracy in high-density areas, low localization ratio in low-density areas and low efficiency in using beacon energy. The article proposes a novel algorithm Generate Beacon-based Tree (GBT) based on node distribution density in the network to solve these problems. By virtue of the concept of Depth-First Traversal (DFT), the algorithm generates a Depth-First Beacon-based Tree (DFBT) by using a beacon group in the network and comparing the number of one-hop neighbors not localized of the sensor nodes within the communication area where beacons transmit messages. The beacon group traverses all the nodes in the network along a planned efficient path and finally guarantees full localization. Compared with other path-based localization algorithms by using mobile beacons, our algorithm has better performance in terms of localization time, localization accuracy and efficiency in using beacon energy.When monitoring area becomes larger and high density of nodes is deployed in it, one beacon group in the network will be insufficient to meet the localization requirement. Hence, Multiple Generate Beacon-based Trees (MGBT) is developed on the basis of the algorithm GBT. MGBT uses multiple beacon groups for localization in the network and each beacon group has its own Local Cartesian Coordinate System (LCCS). Unify Local Cartesian Coordinate Systems (Uni-LCCSs) will be used for different LCCSs, so that all the nodes in the network will be eventually localized by negotiated LCCS. To further improve the performance of algorithm, an optimization strategy called Pruning is come up for MGBT, which leads an algorithm Min-MGBT. Experimental study reveals that the proposed algorithm MGBT and Min-MGBT have improvement in localization time and localization accuracy.