Norm-based and commitment-driven agentification of the Internet of Things

Abstract : There are no doubts that the Internet-of-Things (IoT) has conquered the ICT industry to the extent that many governments and organizations are already rolling out many anywhere,anytime online services that IoT sustains. However, like any emerging and disruptive technology, multiple obstacles are slowing down IoT practical adoption including the passive nature and privacy invasion of things. This paper examines how to empower things with necessary capabilities that would make them proactive and responsive. This means things can, for instance reach out to collaborative peers, (un)form dynamic communities when necessary, avoid malicious peers, and be “questioned” for their actions. To achieve such empowerment, this paper presents an approach for agentifying things using norms along with commitments that operationalize these norms. Both norms and commitments are specialized into social (i.e., application independent) and business (i.e., application dependent), respectively. Being proactive, things could violate commitments at run-time, which needs to be detected through monitoring. In this paper, thing agentification is illustrated with a case study about missing children and demonstrated with a testbed that uses different IoT-related technologies such as Eclipse Mosquitto broker and Message Queuing Telemetry Transport protocol. Some experiments conducted upon this testbed are also discussed
Document type :
Journal articles
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02280905
Contributor : Mohamed Sellami <>
Submitted on : Friday, September 6, 2019 - 4:56:46 PM
Last modification on : Thursday, November 21, 2019 - 2:02:55 AM

Identifiers

Citation

Khouloud Boukadi, Noura Faci, Zakaria Maamar, Emir Ugljanin, Mohamed Sellami, et al.. Norm-based and commitment-driven agentification of the Internet of Things. Internet of Things, Elsevier, 2019, 6, pp.100042. ⟨10.1016/j.iot.2019.02.002⟩. ⟨hal-02280905⟩

Share

Metrics

Record views

49