International Journal of Science and Research (IJSR) ISSN (Online): 2319-7064 Index Copernicus Value (2013): 6.14 | Impact Factor (2013): 4.438 Volume 4 Issue 4, April 2015 www.ijsr.net Licensed Under Creative Commons Attribution CC BY Efficient Technique for Network Lifetime Enhancement by Cleaning Dirty Data Komal V. Shiyale 1 , Pranay D. Saraf 2 1, 2 Computer Science & Engineering, G. H. Raisoni College of Engineering, Nagpur, India Abstract: In the current scenario, wireless sensor and network (WSN) is used in substantial number of applications. It co-operatively sends the sensed data to the base station. This continuous sensing may give rise to the problem of dirty data i.e. missing data, non- ordered data, outlier, redundant data etc. This may lead to energy consumption to large extend and causes inaccuracy in the application where real time and accurate data is most essential factor. Therefore, efficient utilization of power and maintaining the accuracy in data is must in order to use networks for long duration. To address it, an integration of Redundant transmission of sensed data technique and Belief based data cleaning technique has been proposed for improving efficiency in communication and reducing time delay as well as communication overhead in WSN. It also aims at regenerating the missing data, cleaning dirty data, non ordered data, outliers, etc to gather the necessary information which also enhances network lifetime by efficient use of battery power. Keywords: dirty data, energy consumption, network lifetime, cleaning 1. Introduction In recent years it is seen that real-time requirements for timely actions has become the basic need in any network. Therefore, many new techniques are been developed in pervasive computing, communication and sensing technologies and give rise to the emergence of Wireless sensor and actor networks (WSANs). It is a group of sensors (mobile or static) and actors (mostly mobile, e.g. Sub- Kilogram Intelligent Tele-robots (SKITs), Autonomous Battlefield Robot designed for the Army, etc) which are wirelessly connected with each other and perform distributed sensing and actuation tasks [1]. Figure 1: Physical architecture of WSAN Fig. 1 shows the physical architecture of the WSAN. Here sensors are passive elements with limited energy, low cost small device, and low processing and communication capabilities which gather information from physical environment. Were as, actors are active elements with higher energy, longer battery life, processing and communication capabilities, and is responsible for taking decisions and later performing appropriate actions on the environment. In this way actors remotely interact with the environment. WSAN is the combination of sensor and actor nodes, which is illustarted in equation (1) Sensors + Actors = WSANs (1) There are many applications of WSAN like 1) Environmental Applications (e.g. detecting and extinguishing forest fire). 2) Climate control in buildings (e.g. for detection of temperature by the sensors and then trigger the audio alarm actors in that area) 3) Distributed Robotics & Sensor Network (e.g. Mobile robots in sensor network) 4) Battlefield Applications (e.g. detection of mines or explosive substances) There are many factors because of which WSAN is chosen over WSN. They are as follows:  Many applications need real time data as well as data must be valid at the time of action.  Network deployment can be heterogeneous i.e. sensors nodes can be densely deployed and actors nodes can be loosely deployed  Co-ordination between sensors and actors is the main requirement. There are basically two types of coordination a) sensor-actor and b) actor-actor coordination. a) Sensor-Actor coordination: here sensors sense the data from the environment and then transmit the data to actor nodes. b) Actor-Actor coordination: after receiving the data from the sensors actor may or may not coordinate with the nearby actors and make decision to perform appropriate action. Energy efficiency and accurate data is an important concern in Wireless Sensor Networks. We have studied various data cleaning approaches and their effectiveness in enhancing lifetime of Wireless Sensor. Improved Cross-Redundant Data Cleaning Algorithm (ICRDC) [8] and Redundancy Elimination for Accurate Data Aggregation (READA) [9] are used to eliminate redundant data. READA also compresses the data before sending and it can also behave as an event detection system and report if an unusual behavior is noted Adaptive Filter-based Data Cleaning [10] and Online Data Cleaning method [11] are used to clean the outliers, missing Paper ID: SUB153674 2525