Modeling and Analysis of Data Prediction Technique Based on Linear Regression Model (DP-LRM) for Cluster-Based Sensor Networks

Modeling and Analysis of Data Prediction Technique Based on Linear Regression Model (DP-LRM) for Cluster-Based Sensor Networks

Arun Agarwal (Guru Gobind Singh Indraprastha University, India), Khushboo Jain (DIT University, India) and Amita Dev (Indira Gandhi Delhi Technical University for Women, India)
Copyright: © 2021 |Pages: 20
DOI: 10.4018/IJACI.2021100106
OnDemand PDF Download:
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Recent developments in information gathering procedures and the collection of big data over a period of time as a result of introducing high computing devices pose new challenges in sensor networks. Data prediction has emerged as a key area of research to reduce transmission cost acting as principle analytic tool. The transformation of huge amount of data into an equivalent reduced dataset and maintaining data accuracy and integrity is the prerequisite of any sensor network application. To overcome these challenges, a data prediction technique is suggested to reduce transmission of redundant data by developing a regression model on linear descriptors on continuous sensed data values. The proposed model addresses the basic issues involved in data aggregation. It uses a buffer based linear filter algorithm which compares all incoming values and establishes a correlation between them. The cluster head is accountable for predicting data values in the same time slot, calculates the deviation of data values, and propagates the predicted values to the sink.
Article Preview
Top

1. Introduction

The technological innovation in communication technology helps data collected from anywhere, including remote areas of the world, utilizing sensor networks' vast applications. Sensor networks are usually deployed in an area where conventional monitoring applications are insufficient to collect the desired data. WSNs are directed to operate in an isolated environment where recharging of sensor nodes is not possible.

A wide range of applications may be as simple as user-reminder systems or advanced monitoring agents. Sensor network has several applications that include healthcare, IoT, military applications, household safety, and disaster management. The primary objective is to collect a massive amount of data in environment monitoring applications (Kandris et al. 2020). Collecting a considerable amount of data for long durations requires battery life to be sufficiently high, or there must exist a recharging facility of sensor nodes. However, unfortunately, both of the above is not feasible with environmental monitoring of remote areas. Therefore, designing protocols for WSN is to provide prolonged network lifetime and reduced data traffic. Reducing the amount of data for transmission adds one more challenge, i.e., accuracy, which means that while minimizing the data, the accuracy of collected data must be considered, and a threshold must be set which quantifies the amount of data reduction. Sensor networks are designed with other architectural challenges, including random geographical locations and dense sensor node distribution. Random position affects the cluster formation process, and dense distribution adds to spatial correlation. Limited energy resources need a solution that overcomes the challenges mentioned above by adopting proper resource management practices in the proposed solution. Various data management issues, such as redundancy removal, compression, clustering, and classification, have been addressed (Agarwal & Dev, 2017).

Sensor networks operate in a lockstep manner, comprising three necessary steps sensing, aggregation, and transmission (Dhanda & Tyagi, 2016). The objective is to reduce data at each step of sensor network processing. Data aggregation is the main area where sensor nodes can eliminate duplicate data values and restrict extra energy dissipation. Data is collected in time series wherein each time-frame, some data value is sensed and aggregated by the sensor network. The sensed values may contain duplicate or similar data values where reduction of similar data values results in energy preservation. Similar data values may exist both in the time and space domain. A smart sensor network must be able to identify redundant values in both domains. The dynamicity occurs concerning varying network properties such as environment behavior, network architecture, target data set, users changing needs, dynamic queries (Santic et al., 2020; Carlos et al., 2016). These all require smart and intelligent protocols, which assess all the above variations and deliver a promising solution that is energy efficient and ensures minimal data loss.

Sensor nodes have their challenge, which comes with limited power. The limited capabilities in terms of energy and deployment of sensors in a harsh and hostile environment pose additional challenges when nodes are randomly arranged within a network (Elshrkawey et al., 2018). As discussed above, the sensor node has to sense nearby values and send it to the cluster head, which transmits aggregated and filtered information to the sink. Due to continuous sensor node operation, energy is drastically reduced, which results in node failure. One alternate approach is to use the predicted value in place of the actual value, minimizing energy dissipation, and providing a prolonged network lifetime (Jain & Bhola, 2018; Agarwal et al., 2020).

Complete Article List

Search this Journal:
Reset
Open Access Articles
Volume 13: 6 Issues (2022): Forthcoming, Available for Pre-Order
Volume 12: 4 Issues (2021)
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 2 Issues (2016)
Volume 6: 2 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing