Wireless sensor network (WSN) is a collection of sensor nodes deployed in the sensing

The Internet of Things aims to connect each existing things in the world with internet. It is sweeping all the things to a world like a garland where each flower is connected by a sting forming connectivity. IoT can be considered as a big bucket where everything's, every data in the world can be poured to form a live-like connectivity, and hence needs data computation for prediction of the unknown data. Data computation in the internet of things is incorporated to return the value data from the huge collected data collected from the different sources device of the IoT. There are various algorithms for computation of data. This paper focus on comparing supervised learning algorithms i.e. K-NN, Naive Bayes and Cased Based Reasoning (CBR) Classifier.The effects of the mentioned algorithms are based on the following parameters i.e. size of the dataset, performance, processing time and accuracy.

Abstract

Our planet is abundant with raw data and to monitor the available data properly, processing of the enormous raw data is very vital. One of the key things in development of mankind and the nature is to acquire as much data as possible and to react appropriately in accordance with the studied data. It's nothing but diagnosis of the physical world by studying the data acquired from them in order to take proper measures that can help in treating them better. Large volume of data incurs high energy consumption for its transmission and thus results in decrease of overall network lifetime. Wireless Sensor Network (WSN) is a collection of multiple sensor nodes that all together forms a network for transmitting data acquired by each sensor node to sink known as Base Station (BS). In hierarchical routing acquired data are sent via relay agents like Cluster Heads (CH). The Cluster Heads must be customised with computations and formulations, which will help in aggregating the gathered data, in order to reduce energy consumption while transmitting the data further in the network while maintaining the data integrity to withhold the significance of every single value in a data set. The proposed work devised a data aggregation technique to reduce the size of data frame in Wireless Sensor Network which will have the following benefits: • Reduce the amount of data traffic between sensor nodes. • Reduce overall energy consumption of the network. • Communicating more data each round between Sensor Heads and finally to Base Station. • Increase in network lifetime. The work performs data aggregation at bit level which produces better aggregated data as significance of all the bits are included in the data. The consensus based data aggregation algorithm is further used to modify Low-Energy Adaptive Clustering Hierarchy (LEACH) protocol for conserving energy and improving the network performance. Various parameters that are taken into consideration for evaluating Proposed Work protocol:

No abstract