Homomorphic Asymmetric Encryption applied to analysis of IoT Communications

Abstract

In this paper, we describe the use of homomorphic encryption techniques in order to not only ensure the data are transmitted in a confidential way, but also to use the encrypted data to provide the manager with statistics that allow them to detect the incorrect functioning of a sensor node or a group of sensors due to either malicious data injection, data transmission, or simply sensor damage (miscalibration, faulty sensor functioning). Obtaining these statistical values does not need decryption, so the process is sped up and can be developed in real time. Operating the data in this way ensures privacy and removes the need to maintain a shared key infrastructure between the sensor nodes and the manager nodes that are part of the blockchain infrastructure. In this work, we focus on operations with the sensor nodes that provide data that will be, later, treated as part of the business logic in the agribusiness sector (for example), hence the importance of having fast checking mechanisms in terms of data quality. The results obtained on conventional configurations of sensor nodes encourage the use of this technique in the aforementioned infrastructure.