Performance Comparison of Various Wireless Sensor Network Dataset using Deep Learning Classifications

Deep Learning is the subset of artificial intelligence and various techniques are available to predict the performance of real time applications. Wireless devices are available to access the devices from multiple places based environment and coverage area. Wireless sensor network (WSN) tools are positioned in multiple places which has sense the information from the substances. The performance is the major factor to facilitate the bandwidth and power. In this paper to analyse the wireless sensor network performance using deep learning techniques. Here to measure the performance indicators such as accuracy, precision and score function of sensor dataset using TensorFlow. The deep learning models such as convolutional neural networks, recurrent neural network and k-means neural network performance factors are classified by using sensory dataset. The accuracy factor is obtained as 96% and compares this with existing models.

I. Introduction

WSN is the major real time working environment to collect the data from multiple interfaces based on objects [1], location [2], coverage area [3] and sensing environments [4]. Each collected information are recorded and processed. Major issues in WSNs are coverage constraints and energy consumptions. The amount of data is considered and attracted based on aspects, networking lifetime, capacity index and routing strategies. The main contributions of WSNs are multi-hop routing features, network topologies; resource limited access, sensory networks and data centric approach. It has large volume of dataset and dynamically moving nodes. In this case the need a self-organized routing model for handling collaboration, detection and monitoring [5]. The network coverage another issues to select the nodes such as sensor, user, sink and connectivity module which is shown in Fig 1.