Using Self-Organizing Map in Backbone Formation for Wireless Sensor Networks

Backbone formation has been used extensively in various aspects for wireless sensor networks recently. Many methods are mostly designed to minimize the size of the backbone or find small cost path. This paper propose the Artificial Neural Networks Routing (ANNR) algorithm which use self-organizing map to measure the Quality of Service supported by the networks. A method of offline process and online process is discussed to deal with constrains on sensor data computing and power consumption. Experimental results have confirmed the algorithm feasibility and the validity.