International Journal of Engineering and Advanced Technology (IJEAT) ISSN: 2249 – 8958, Volume-2 Issue-3, February 2013 57 Published By: Blue Eyes Intelligence Engineering & Sciences Publication Retrieval Number: C0996022313 /2013©BEIESP Diverse Sorting Algorithm Analysis for ACSFD in Wireless Sensor Networks Jose Anand, K. Sivachandar Abstract - The progression in wireless communications and electronics has emerged the expansion of low-cost sensor networks. Sensor networks exploit large number of wireless sensor nodes to collect information from their sensing terrain. The gathered information will undergo in-network process and send to the remote sink. Sensor networks have wide range of applications including medical, military, home, etc. In the sensor networks, a fault-tolerant distributed decision fusion will occur due to the presence of sensor faults. For this fault detection, Collaborative Sensor Fault Detection (CSFD) scheme was used and this fault detection scheme is very difficult to implement because of the extensive computations. So an Approximated Collaborative Sensor Fault Detection (ACSFD) scheme was developed, which is less cost and utilizes less power than CSFD and has the same performance of CSFD. The important blocks present in the architecture of ACSFD consist of multipliers, logarithms, and sorting. In this paper, analysis has been done with various sorting algorithms and concluded the best sorting technique that can be used in ACSFD scheme to improve the performance of the fault detection scheme in the wireless sensor network. Keywords – ACSFD, fault detection, sorting algorithms, wireless sensor networks. I. INTRODUCTION Wireless Sensor Networks (WSNs) employ enormous amount of wireless sensor nodes to collect information from their sensing terrain. The gathered information will undergo in-network process and are communicated to the remote sink [6]. The wireless sensor nodes in the WSN are compact, light-weighted, and are battery-powered devices that can be used in almost any environment. With this data, simple computations are carried out and communication with other sensor nodes or controlling authorities in the network will take place [8]. WSN have recently engrossed in plenty of applications which include environment monitoring such as temperature, sound, pressure, vibration, pollutants, oxygen content, carbon dioxide content, sulphur dioxide content, etc. at different locations, mobile object tracking, and navigation applications. All these applications consist of many inexpensive wireless sensor nodes that are proficient of collecting, processing and storing various information in a prompt manner. In wireless sensor applications all the sensor nodes will periodically report the collected information to a single sink node which realizes a many-to-one communication network model [3]. The sensor nodes are organized as a cooperative network and the structure of a node is shown in figure 1. Manuscript Received on February, 2013. Jose Anand, Associate Professor, Department of Electronics and Communication Engineering, K C G College of Technology, Karapakkam, Chennai, Tamil Nadu, India. K. Sivachandar, Associate Professor, Department of Electronics and Communication Engineering, K C G College of Technology, Karapakkam, Chennai, Tamil Nadu, India. Each node has controllers for processing, multiple types of memories like program, data, and flash, RF transceivers, power source, and various sensors and actuators. The sensor nodes communicate wirelessly and often self-organize after being deployed in an ad hoc fashion. Figure 1 Structure of Sensor Node Packets transmitted in the WSN contain useful information, which can be utilized through packet-based computation and to reduce the sensor fault rate [6]. The WSN packet computation has small packet forwarding rate and the forwarding computation capability is limited. Mostly the sensor nodes are modeled with limited energy, as a result the sensor nodes lacks recharging issues. But still wireless nodes packet-based computation is preferred since it is generally known that the computation utilizes reduced energy than the communication [1]. To attain the Quality of Service (QoS) requirements, the network resources should be used in a fair and efficient manner. Moreover, techniques such as data compression, data fusion and aggregation become very useful in maintaining robustness. Due to the changes in node mobility and wireless channel or node failure, the WSN seems to be unreliable in nature. In order to efficiently use the WSN for real-time applications the issues related to the wireless protocols are reduced [6]. As a result, an efficient design of distributed fault-tolerant fusion center is important in the WSN. The sequential testing to judge the fusion center for WSN requires error probability and decisions are made on the comparison with the threshold value on the requirement [2]. The decisions taken by the fusion center are delayed and are reported at regular time interval to select the correct strategy. CSFD scheme is used for isolating the faulty nodes in the WSN and to eliminate the unreliable local decisions [7]. The CSFD scheme will make decision fusion in the WSN at regular time interval and meet the real-time requirements. The scheme identifies the faulty sensor node but the failure on the dynamic system cannot be detected. The collaborative signal processing improves the performance of decision-making process. That is, the CSFD scheme determines the faulty nodes each time with minimum error bound. The performance of CSFD