Performance Evaluation of OpenID Connect for an IoT Information Marketplace

Abstract

The continuous global growth of Internet of Things devices and deployments constantly produce an equally growing volume of streaming data. This volume puts high pressure on the cloud infrastructure that is expected to manage the storage and processing of such streams. Special purpose IoT data management platforms such as the IoT Framework which is the system under study in this paper, transform raw streaming data into actual information products thus providing value added services to interested users. The purpose of this paper is to evaluate and analyse the impact of the OpenID Connect protocol on the performance of the IoT Framework. The OpenID protocol was introduced in order to enable the IoT framework to be transformed into an IoT Information marketplace. The obtained results show that the introduction of an AA mechanism has a considerable impact on the performance of the IoT Framework. The reason is that the AA server is remote with respect to the IoT platform and therefore the network delay impacts the experienced end to end delay by the user. A possible means to alleviate this problem is to create a local AA server closer to the IoT Platform which either maintains local user AA credentials or federates with other public AA servers.

Improving the Privacy of IoT with Decentralised Identifiers (DIDs)

“…Future work in this area includes, e.g., designing and evaluating a proxy solution for DIDs, an evaluation of the benefits and drawbacks of DIDs compared to other identifier solutions, including roles/attributes-based solutions (e.g., [44]), token-based solutions (e.g., OpenID as used in [45]), capabilities-based solutions (e.g., [46]), and others. Another area for future work is studying how the use of DIDs could enhance the general IoT authorisation problem, i.e., how the user gets access to some device, which is managed by some party.…”

A Survey of Internet of Things (IoT) Authentication Schemes

“… Authentication factor Identity: An information presented by one party to another to authenticate itself. Identity-based authentication schemes can use one (or a combination) of hash, symmetric or asymmetric cryptographic algorithms.Context: which can be: Physical: Biometric information based on physical characteristics of an individual, e.g., fingerprints, hand geometry, retinal scans, etc.Behavioral: Biometric based on behavioral characteristics of an individual, e.g., keystroke dynamics (pattern of rhythm and timing created when a person types), gait analysis (method used to assess the way we walk or run), voice ID (voice authentication that uses voice-print), etc. Use of tokens Token-based Authentication: Authenticates a user/device based on an identification token (piece of data) created by a server such as OAuth2 protocol [65,66] or open ID [67]. Non-Token based authentication: Involves the use of the credentials (username/password) every time there is a need to exchange data (e.g., TLS/DTLS [12,68,69]). Authentication procedure One-way authentication: In a scenario of two parties wishing to communicate with each other, only one party will authenticate itself to the other, while the other one remains unauthenticated.Two-way authentication: It is also called mutual authentication, in which both entities authenticate each other.Three-way authentication: Where a central authority authenticates the two parties and helps them to mutually authenticate themselves. Authentication architecture Distributed: Using a distributed straight authentication method between the communicating parties.Centralized: Using a centralized server or a trusted third party to distribute and manage the credentials used for authentication.Whether centralized or distributed, the authentication scheme architecture can be: Hierarchical: Utilizing a multi-level architecture to handle the authentication procedure.Flat: No hierarchical architecture is used to deal with the authentication procedure. IoT layer : Indicates the layer at which the authentication procedure is applied.…”

Survey on Identity and Access Management for Internet of Things

“…Since this work uses OpenID, it inherits well-know challenges of this platform. For example, if we consider the latency and throughput of the network and the use of computational resources of the devices, the OpenID contributes to highload environments, resulting in a performance loss [105]. Furthermore, since the access token mechanism of the OpenID Connect protocol utilizes the same token across different requests, a malicious user can acquire this token to use in a man-in-the-middle attack and get unauthorized access to data.…”