Hu, Jiejun and Reed, Martin and Thomos, Nikolaos and Al-Naday, Mays and Yang, Kun (2022) A Dynamic Service Trading in a DLT-Assisted Industrial IoT Marketplace. IEEE Transactions on Network and Service Management, 19 (4). DOI https://doi.org/10.1109/TNSM.2022.3191979
Hu, Jiejun and Reed, Martin and Thomos, Nikolaos and Al-Naday, Mays and Yang, Kun (2022) A Dynamic Service Trading in a DLT-Assisted Industrial IoT Marketplace. IEEE Transactions on Network and Service Management, 19 (4). DOI https://doi.org/10.1109/TNSM.2022.3191979
Hu, Jiejun and Reed, Martin and Thomos, Nikolaos and Al-Naday, Mays and Yang, Kun (2022) A Dynamic Service Trading in a DLT-Assisted Industrial IoT Marketplace. IEEE Transactions on Network and Service Management, 19 (4). DOI https://doi.org/10.1109/TNSM.2022.3191979
Abstract
With the increasing demand for digitalization and participation in Industry 4.0, new challenges have emerged concerning the market of digital services to compensate for the lack of processing, computation, and other resources within Industrial Internet of Things (IIoTs). At the same time, the complexity of interplay among stakeholders has grown in size, granularity, and variation of trust. In this paper, we consider an IIoT resource market with heterogeneous buyers such as manufacturer owners. The buyers interact with the resource supplier dynamically with specific resource demands. This work introduces a broker between the supplier and the buyers, equipped with Distributed Ledger Technologies (DLT) providing a service for market security and trustworthiness. We first model the DLT-assisted IIoT market analytically to determine an offline solution and understand the selfish interactions among different entities (buyers, supplier, broker). Considering the non-cooperative heterogeneous buyers in the dynamic market, we then follow an independent learners framework to determine an online solution. In particular, the decision-making procedures of buyers are modeled as a Partially Observable Markov Decision Process which is solved using independent Q-learning. We evaluate both the offline and online solutions with analytical simulations, and the results show that the proposed approaches successfully maximize players’ satisfaction. The results further demonstrate that independent Q-learners achieve equilibrium in a dynamic market even without the availability of complete information and communication, and reach a better solution compared to that of centralized Q-learning.
Item Type: | Article |
---|---|
Uncontrolled Keywords: | DLT; Dynamic Pricing; IIoT marketplace; Independent Learning; POMDP |
Divisions: | Faculty of Science and Health Faculty of Science and Health > Computer Science and Electronic Engineering, School of |
SWORD Depositor: | Unnamed user with email elements@essex.ac.uk |
Depositing User: | Unnamed user with email elements@essex.ac.uk |
Date Deposited: | 20 Jul 2022 09:16 |
Last Modified: | 30 Oct 2024 21:02 |
URI: | http://repository.essex.ac.uk/id/eprint/33174 |
Available files
Filename: IEEETranNetSerHu22.pdf