Implementation of Homomorphic Encryption for Wireless Sensor Networks Integrated with Cloud Infrastructure

With the current explosion of cloud services, it is easy for everyone to store the prodigious amount of data on remote servers. On the other hand, service providers are given full access to our information, or in the worst case, can be intercepted by malicious people. The solution is to encrypt the data to make it completely secret. But then, we can no longer manipulate these data remotely (retouching the photos, looking for words in a text, performing calculations ...). A new form of cryptography, which is just beginning, promises precisely to offer this security while allowing the encrypted data to be manipulated by the authorized users, it named “Homomorphic Encryption”. This paper presents the implementation of a homomorphic encryption to ensure the confidentiality of aggregated data from wireless sensor networks and also demonstrates how to use it through the cloud infrastructure in order to perform arithmetic operations directly on the encrypted data without decrypting them. After a theoretical comparative study on homomorphic encryption algorithms, we picked out the Domingo-Ferrer’s cryptosystem on the basis of criteria mentioned during the survey. This cryptosystem requires to split the message into several fragments (at least two fragments) before the encryption process. For each implementation case performed with two, three and four fragments of the message, 10,000 messages were sent so that the battery levels of the sensor nodes dropped by 6%, 7% and 8% respectively, which led us to conclude that as long as the number of fragments increases, the power consumption in the network is increasingly significant because of excessive processing and a large amount of data to be transmitted over the wireless network, generated by the Domingo-Ferrer’s cryptosystem.