Time-Efficient Advent for Diagnosing Flaws in Hadoop on the Big Sensor Type Data

Abstract

Hadoop is a MapReduce-based distributed processing framework used in the area of big data analytics in every organizations. Big sensor data is difficult to manage with the traditional data management tools. Thus, Hadoop challenges to manage it in high scalable amount in a time-efficient manner. In this paper, for fast detection of flaws in big sensor data sets, a different type of approach in diagnosing flaws with the time efficiency is used. Due to the wireless transfer of data across the nodes in a wireless sensor networks, there can be loss of data which will result in wrong interpretation of data at the nodes. The proposed approach of this paper is to form a group of sensors as a cluster. If any sensor detects violations, then the energy of that sensor has to be compared with the other sensors. The sensor having the highest energy will become the cluster head, and it will send the sensed data to the data center. The data center then diagnoses the flaw with respect to the sensed data in the big sensor data.