Privacy preserving identification in home automation systems

Official URL: https://risc01.sabanciuniv.edu/record=b2579370_(Table of contents)

Home Automation System (HAS) is a set of interconnected devices in a household that are accessible via the Internet. Homeowners can monitor and control the smart home appliances using HAS, but if the system is not structured properly, intruders can gain access to one or more of the devices in the system. In this thesis, we propose a privacy preserving identification model for HAS. In this model, first, a secure key sharing and credential issuance protocol is presented. In this protocol Idemix’s Verifiable Encryption scheme is implemented. Verifiable encryption schemes ensure the authenticity of the Issuer and the User, while keeping the communication fully secure. During this protocol, the master key for the mutual verification and authentication protocol is shared and the credentials for the User are issued. Then, we explain the mutual verification and authentication protocol, which also employs the HAS Management System (HMS) as a trusted party to obfuscate the communication between the Vendors and Innovative Home Gateway (IHG)’s. We preserve the privacy of the homeowners by masking the brands, types and id’s of the appliances inside the household. HMS carries on two different conversations, one with Vendor and the other with IHG, replacing the UUID’s in the topics. The topics and messages published on these are unlinkable, thus masking the identity of the IoT devices connected to the IHG. The performance tests are performed using 4 different scenarios. Moreover, security analysis for both key sharing and credential issuance protocol and for the mutual verification and authentication protocol are given, and they are both proven to be secure according to OFMC and ATSE specifications. In conclusion, the system is scalable for actual implementation, and provides security and privacy as proposed.