Model-Free Online TDD Configuration for Mobile Edge Computing
Authors:
- Pengtao ZhaoPZPengtao Zhao
- Hui TianHTHui Tian
- Kwang-Cheng ChenKCKwang-Cheng Chen
- Shaoshuai FanSFShaoshuai Fan
- Gaofeng NieGNGaofeng Nie
- 1 May 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
With tremendous computing power in the radio access network, mobile edge computing (MEC) that can support localized context awareness creates a new technological frontier for 5G and beyond. To efficiently exploit the networking and computing functionalities, Time Division Duplex Orthogonal Frequency Division Multiple Access (TDD-OFDMA) type has been considered in this paper. To take advantage of dynamic features in TDD, a model-free online TDD configuration scheme is proposed based on context analysis and Multi-Armed Bandit (MAB) optimization. The TDD configuration problem is therefore novelly modeled as a contextual MAB problem, and is solved by the contextual Upper-Confidence-Bound (C-UCB), which dynamically adjusts TDD configuration to network traffic since that the system cost can be reduced. To further reduce the energy consumption and makespan of mobile devices (MDs), a greedy resource allocation (GRA) embedded in the TDD configuration is developed to select MDs and allocate resources. The simulations demonstrate that proper TDD configuration successfully reduces the system cost, and C-UCB technique approaches the ideal TDD configuration, with significant performance gain when the GRA effectively select and allocate, to strike simultaneous efficiency for mobile networking and MEC.Keywords
This publication has 8 references indexed in Scilit:
- A Robust Regularization Path Algorithm for $\nu $ -Support Vector ClassificationIEEE Transactions on Neural Networks and Learning Systems, 2016
- Mobile Edge ComputingInformatik Spektrum, 2016
- A dynamic bandwidth allocation algorithm in mobile networks with big data of users and networksIEEE Network, 2016
- Enabling Personalized Search over Encrypted Outsourced Data with Efficiency ImprovementIEEE Transactions on Parallel and Distributed Systems, 2015
- Application optimization in mobile cloud computing: Motivation, taxonomies, and open challengesJournal of Network and Computer Applications, 2015
- Enabling Real-Time Context-Aware Collaboration through 5G and Mobile Edge Computing2015
- Application partitioning algorithms in mobile cloud computing: Taxonomy, review and future directionsJournal of Network and Computer Applications, 2015
- Reinforcement Learning: An IntroductionIEEE Transactions on Neural Networks, 1998