Doctorate thesis defense of Haythem HAYOUNI

Committee

Abstract

A Wireless Sensor Network (WSN) is a set of nodes usually dedicated to the collection of informations, able to communicate with each other to perform various tasks. Sensors are restricted by energy consumption due to battery supply and are limited in computing capacity; therefore, energy saving technologies must be considered. Data aggregation represents an important method to remove redundancy as well as unnecessary data transmission and hence cut down the energy used in communication. However, WSNs are often deployed in hostile environment which make data aggregation face to many security and performance challenges, such as privacy protection and data integrity. In this thesis, we will address research issues related to secure data aggregation in WSNs based on homomorphic encryption. Consequently, the main contributions of the thesis are two-issues

In the first issue, we took interest to provide a new key management scheme for WSNs to establish a secure paths between sensor nodes. Firstly, we introduce a clustering algorithm allowing the election of cluster heads and the formation of different clusters. Secondly, we establish a pairwise key between two sensor nodes. Finally, we create a global group key agreement for all sensor nodes in the network. This scheme is dynamic because it allows new nodes to join the network at any time and periodically updates the established key agreement. It is shown that with this scheme, we can provide more efficient security with less energy consumption and lower storage space.

In the second issue, we propose a secure data aggregation scheme for WSNs based on homomorphic primitives. This scheme combine data aggregation, data confidentiality, data integrity and false data detection together well. We propose a symmetric-key homomorphic encryption to protect data privacy. Then, the aggregated ciphertexts is combined with homomorphic MAC (HMAC) synchronically to check the aggregation data integrity. To verify data integrity, not only in the base station, but also during data aggregation and data forwarding processes, we introduce an integrity detection process to detect false data as early as possible in each network level. The objective is to reduce the communication overhead and energy consumption, which prolong the life of the sensor nodes and network.