Fast and secure implementation of ECC-based concealed data aggregation in WSN

In the area of wireless sensor networks (WSNs), research on secure data aggregation has grown dramatically in the past few years, since the sensors are deployed in unattended and hostile environments. For this purpose, end-to-end solutions known as concealed data aggregation have been proposed to provide privacy between the sensors and the sink. The recent experiments showed that the Elliptic Curve El Gamal (ECEG) is the most suitable cryptosystem for WSNs in order to provide the end-to-end confidentiality with a high level of security based on elliptic curve cryptography (ECC). However, the execution time remains the major drawback due to the complexity of ECC operations and the nature of sensor nodes which are equipped with limited computer resources. Therefore, an efficient implementation of such operations is then crucial. In this paper, we present a fast and secure implementation of ECEG in MicaZ mote, based on an enhanced version of TinyECC library, a fast scalar multiplication is employed, which is also secure against side channel attacks (in particular simple power analysis). Moreover, a fast point decompression algorithm needed for homomorphic operation is used to efficiently computing square roots in prime fields, satisfying p = 3 mod 4. Our results show that a secure encryption of ECEG takes only 1.29s, which is considerably better than previous software implementations on 8-bits platforms.