Secure data aggregation protocol for wireless sensor network

Abstract

In this work, the relationship between security and data aggregation protocols has been investigated. Providing security causes more energy consumption due high computation requirements. Therefore, a method is needed to facilitate energy consumption. Combining security with data aggregation is beneficial because data aggregation protocols decrease the amount of communications that reduce energy expenditure of redundant transmissions. In this project, Data Aggregation and Authentication protocol (DAA) is simplified called SDAA and implemented in order to detect false data along with data aggregation and confidentiality. The main goal of this particular algorithm is to find false data as early as possible to decrease data transmissions. In DAA there are three steps which are: Monitoring node selection, forming pair of nodes and integration of false data detection along with data aggregation and confidentiality as SDFC algorithm. In order to simplify DAA the first two steps have been predefined prior to network deployment. Besides, monitoring nodes are assumed to be sensing nodes which sense plain data. Eight different experiments are performed where in each experiments the number of data aggregators or number of monitoring nodes are varied. TelosB motes were deployed and all of results that relate to this particular sensor node with its constraints are analyzed. Consequently, with the aid of false data detection along with the data aggregation total amount of transmission compared to traditional methods decreased tremendously. Although the number of computations increased, the energy conserved by applying data aggregation and false data detection is more than other similar methods that detects false data at the base station. In this work, the number of transmissions of packets decreases from approximately 4000 to 2500 packets which shows the efficiency of SDAA protocol. However, when number of monitoring nodes increased the amount of packet loss and end-to-end delay also increased. In this project, to facilitate packet loss and end-to-end delay synchronization of monitoring nodes has been utilized. Moreover, with the aid of synchronization the end-to-end delay decreased from 17583 (ms) to 679 (ms) and the amount of packet loss improved from 85% to only 18%.