Secure Service Virtualization in IoT by Dynamic Service Dependency Verification

Virtualizing Internet-of-Things (IoT) services is a concept of dynamically building customized high-level IoT services that rely on the real-time data streams flowing from low-level standalone IoT devices. IoT service virtualization is essential when a myriads of IoT devices can get online, interact with each other, exchange data, and based on them create one's own service. Especially, when virtualization occurs across multiple externals domains, it is crucial for clients to verify the source of virtual services, i.e., whether they are built based on authentic original service sources. Also, original services' sources must be constantly aware of the identity of entities who (recursively) virtualize their services. To address these issues, this paper proposes IoT service dependency tree (SDT) validation scheme. SDT uses service dependency trees and dependency signature trees, which enable clients to validate the original sources of a virtual IoT service, verify its service dependency relationships, and have original service sources to be constantly notified of the list of entities (recursively) virtualizing their services. This paper explains SDT scheme and presents use cases for IoT service virtualization where SDT can be applied. Our experimental analysis shows that SDT is scalable for practical use.