Gradient Descent Localization Algorithm Based on Received Signal Strength Technique in a Noisy Wireless Sensor Network

Abstract

In this article we propose an improved RSS-based localization technique in a wireless sensor network based on the gradient descent optimization algorithm in a noisy propagation model, usually RSS is susceptible to the noise factor affecting the estimated distance of the sensor network. Generally, to localize an unknown node, a specific number of anchors (reference nodes) is required, this number of needed anchors will increase in the presence of noise factor. Results have shown that, using our improved technique, the number of used anchors is reduced despite the existence of the noise aspect. A comparison with other techniques is made to show the effectiveness of our proposed approach.