1558-1748 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2018.2818302, IEEE Sensors Journal Abstract—Efficient and low power utilizing clock synchronization is a challenging task for a wireless-sensor network (WSN). Therefore, it is crucial to design a light weight clock synchronization protocols for these networks. An adaptive clock offset prediction model for WSN is proposed in this paper that exchanges fewer synchronization messages to improve the accuracy and efficiency. Timing information required is collected by setting a small WSN set up to investigate the soil condition to control the irrigation in agriculture. The networks investigate soils moisture, temperature, humidity and pressure content along with the sensors clock offset. First, the prediction model perceives the existing sensor clock offset to observe the clock characteristics and delay. Then, a Gaussian function is applied for adjusting the parameters weight of the observed value in the prediction model. The system results demonstrate that the proposed Adaptive Non-Linear Gaussian Regression Synchronization (ANGRS) model utilizes 20% less energy as consumed by Time Sync Protocol (TPSN) for Sensor-Network and Reference Broadcast Synchronization (RBS) Protocol. It also reduces the synchronization error with respect to Root-Mean Square Error (RMSE) by 24.85% as compared to Linear Prediction Synchronization (LPS) with RMSE 28.72% in terms of accuracy. Index Terms— Gaussian Function, Non-Linear Regression Model, Root Mean Square Error, Time-Synchronization, Wireless-Sensor Network. I. INTRODUCTION N efficient and accurate time is crucial for Wireless sensor Networks (WSN). Synchronization time is necessary for coordinating the sensing and communication Manuscript submitted on December 25, 2017, Revised on March 18, 2018. Divya Upadhyay is with the Department of Computer Science & Engineering, Amity School of Engineering & Technology, Amity University Uttar Pradesh, Noida, UP, India (e-mail: dupadhyay@amity.edu). Ashwani Kumar Dubey* is with the Department of Electronics & Communication Engineering, Amity School of Engineering and Technology, Amity University Uttar Pradesh, Noida, UP, India (e-mail: dubey1ak@gmail.com). P. Santhi Thilagam is with the Computer Science and Engineering Department, NITK Surathkal, Karnataka, India (e-mail: santhisocrates@gmail.com). between the nodes. Time synchronization is also required for tracking, computation and localization within the distributed nodes. Therefore, the experiment in this paper is conducted on a WSN designed to investigate and control the irrigation in agriculture. As it is a familiar fact that energy is the critical constraint for a WSN, therefore light-weight time synchronization algorithm is essential. Few unique characteristics of the nodes and WSN make the time synchronization a difficult task. The most essential characteristic is the limited energy for the low-end sensor nodes. There exist some areas like agriculture monitoring, fire monitoring in forest, volcano monitoring, and glacier monitoring systems where these sensors cannot be recharged [1]. To optimize the energy utilization of sensor nodes, the synchronization algorithm should be efficient and simple. Also, frequent resynchronization should not be required. Therefore, the primary challenge for clock synchronization procedures is its accuracy and simplicity. Other challenges to be taken care are: A) Delay due to unavoidable reasons at physical layer and data–link layer, B) Restriction on computation capabilities, C) Transmission range and storage space. Therefore, in this paper, an adaptive clock synchronization algorithm for WSN is proposed and analyzed based on the data received from the experimental setup as shown in Fig.1. WSN network consist of several sensor nodes that could only communicate with their neighbors. Protocols designed for wired and Adhoc networks could not be directly implemented for WSN keeping the features of WSN in mind. In past few years, several better and new time-synchronization protocols were developed such as fine-grained network time synchronization using reference broadcasts [2, 3]. Early clock synchronization protocols were based on exchange of multiple synchronization messages between nodes. But, recent synchronization protocols focuses on low message and communication overheads in WSN. It is assumed that in the existing synchronization schemes, the clock offset, skew and delay in delivery of the synchronization messages are distributed using Gaussian, exponential or gamma distribution models [4]. Non-Linear regression model is a mathematical procedure where future clock drift is predicted by means of a non-linear function from Application of Non-Linear Gaussian Regression based Adaptive Clock Synchronization Technique for Wireless Sensor Network in Agriculture Divya Upadhyay, Ashwani Kumar Dubey*, and P. Santhi Thilagam A