Computing Resource Provisioning at the Edge: An Overbooking-Enabled Trading Paradigm
- 9 May 2022
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications
- Vol. 29 (5) , 68-76
Abstract
The arrival of Internet of Things (IoT) era calls for cost-effective resource provisioning techniques to secure necessary computing resources in supporting dynamic and computation-intensive applications on diverse IoT devices, where resource trading in edge networks brings a viable prospect. While conventional onsite spot trading may facilitate this paradigm, the inherent extra latency of onsite decision-making and unreliable network significantly limit resource trading efficiency and the eventual resource utilization rate. More importantly, the "no show" of confirmed resource buyers incurred by unreliable network conditions, can result in negative impacts on trading performance and resource usage. To this end, resource overbooking is adopted that provides an effective solution in handling "no shows", while substantially improving resource utilization rate. In this article, an interesting overbooking-enabled resource trading paradigm in mobile edge networks is proposed, allowing prenegotiated forward contracts between the edge server (seller) and a number of overbooked smart devices (buyers), which will be fulfilled during future trading. Specifically, framework associated with the proposed trading paradigm is introduced first to provide a comprehensive view of the relevant key issues. Then, a case study is presented which elaborates how the proposed trading can be implemented in practice. Extensive simulation results and analysis demonstrate that the overbooking-enabled trading paradigm achieves both time and energy efficiency, as well as substantial resource utilization and profit improvements.Keywords
Funding Information
- National Natural Science Foundation of China (62271424,61971365)
- Natural Sciences and Engineering Research Council of Canada (RGPIN2018-06254)
This publication has 10 references indexed in Scilit:
- A Truthful Auction for Graph Job Allocation in Vehicular Cloud-Assisted NetworksIEEE Transactions on Mobile Computing, 2021
- DEFT: Decentralized Multiuser Computation Offloading in a Fog-Enabled IoV EnvironmentIEEE Transactions on Vehicular Technology, 2020
- On Provisioning Slices and Overbooking Resources in Service Tailored Networks of the FutureIEEE/ACM Transactions on Networking, 2020
- Optimal Spot Trading Integrated with Quantity Flexibility ContractsProduction and Operations Management, 2020
- Cache-Enabled Coordinated Mobile Edge Network: Opportunities and ChallengesIEEE Wireless Communications, 2020
- Futures-Based Resource Trading and Fair Pricing in Real-Time IoT NetworksIEEE Wireless Communications Letters, 2019
- Emerging Edge Computing Technologies for Distributed IoT SystemsIEEE Network, 2019
- A Multi-User Mobile Computation Offloading and Transmission Scheduling Mechanism for Delay-Sensitive ApplicationsIEEE Transactions on Mobile Computing, 2019
- Packing Groups of Items into Multiple KnapsacksACM Transactions on Algorithms, 2018
- High Performance Resource Allocation Strategies for Computational EconomiesIEEE Transactions on Parallel and Distributed Systems, 2012