Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement Learning
Nowadays, mobile application services face the challenges of high speed, low latency and high reliability. The combination of digital twin (DT) technology and mobile edge computing (MEC) network can effectively solve these challenges. DT technology can help MEC network monitor and predict the networ...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10070781/ |
| _version_ | 1797861351879606272 |
|---|---|
| author | Siyu Yuan Zhenyu Zhang Qin Li Weiyuan Li Yong Zhang |
| author_facet | Siyu Yuan Zhenyu Zhang Qin Li Weiyuan Li Yong Zhang |
| author_sort | Siyu Yuan |
| collection | DOAJ |
| description | Nowadays, mobile application services face the challenges of high speed, low latency and high reliability. The combination of digital twin (DT) technology and mobile edge computing (MEC) network can effectively solve these challenges. DT technology can help MEC network monitor and predict the network states. In this paper, we propose a DT-aided MEC network scenario with deep neural network (DNN) inference as the computing task of end devices (EDs). ED can offload part of DNN layers to MEC server. To allocate communication resources, we propose an algorithm based on asynchronous advantage actor-critic (A3C), which manages the transmission power and channel selection of EDs. Since DNN inference is continuous in real scenes, we consider the continuous DNN inference tasks. We convert the DNN optimal partition point solving problem to a min st-cut problem, and propose a graph theory based DNN optimal partition point solving algorithm to minimize the inference latency. Simulation results show that the proposed algorithm can effectively reduce the inference latency. Compared with actor-critic (AC) and deep Q network (DQN), the proposed algorithm has faster convergence speed and better convergence value. Compared with the traditional one-time DNN model partition algorithm, the proposed algorithm is more suitable for DNN continuous task arrival scenario. |
| first_indexed | 2024-04-09T22:00:51Z |
| format | Article |
| id | doaj.art-63a94166224047eeb4a56fbb7e33cbe3 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-09T22:00:51Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-63a94166224047eeb4a56fbb7e33cbe32023-03-23T23:00:22ZengIEEEIEEE Access2169-35362023-01-0111270992711010.1109/ACCESS.2023.325734210070781Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement LearningSiyu Yuan0https://orcid.org/0000-0001-8151-0130Zhenyu Zhang1https://orcid.org/0000-0001-5589-7538Qin Li2Weiyuan Li3Yong Zhang4https://orcid.org/0000-0003-4997-698XSchool of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing, ChinaSchool of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing, ChinaChina Mobile Research Institute, Beijing, ChinaChina Mobile Research Institute, Beijing, ChinaSchool of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing, ChinaNowadays, mobile application services face the challenges of high speed, low latency and high reliability. The combination of digital twin (DT) technology and mobile edge computing (MEC) network can effectively solve these challenges. DT technology can help MEC network monitor and predict the network states. In this paper, we propose a DT-aided MEC network scenario with deep neural network (DNN) inference as the computing task of end devices (EDs). ED can offload part of DNN layers to MEC server. To allocate communication resources, we propose an algorithm based on asynchronous advantage actor-critic (A3C), which manages the transmission power and channel selection of EDs. Since DNN inference is continuous in real scenes, we consider the continuous DNN inference tasks. We convert the DNN optimal partition point solving problem to a min st-cut problem, and propose a graph theory based DNN optimal partition point solving algorithm to minimize the inference latency. Simulation results show that the proposed algorithm can effectively reduce the inference latency. Compared with actor-critic (AC) and deep Q network (DQN), the proposed algorithm has faster convergence speed and better convergence value. Compared with the traditional one-time DNN model partition algorithm, the proposed algorithm is more suitable for DNN continuous task arrival scenario.https://ieeexplore.ieee.org/document/10070781/DTMEC networkcontinuous DNN inference taskA3Cmin st-cut |
| spellingShingle | Siyu Yuan Zhenyu Zhang Qin Li Weiyuan Li Yong Zhang Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement Learning IEEE Access DT MEC network continuous DNN inference task A3C min st-cut |
| title | Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement Learning |
| title_full | Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement Learning |
| title_fullStr | Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement Learning |
| title_full_unstemmed | Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement Learning |
| title_short | Joint Optimization of DNN Partition and Continuous Task Scheduling for Digital Twin-Aided MEC Network With Deep Reinforcement Learning |
| title_sort | joint optimization of dnn partition and continuous task scheduling for digital twin aided mec network with deep reinforcement learning |
| topic | DT MEC network continuous DNN inference task A3C min st-cut |
| url | https://ieeexplore.ieee.org/document/10070781/ |
| work_keys_str_mv | AT siyuyuan jointoptimizationofdnnpartitionandcontinuoustaskschedulingfordigitaltwinaidedmecnetworkwithdeepreinforcementlearning AT zhenyuzhang jointoptimizationofdnnpartitionandcontinuoustaskschedulingfordigitaltwinaidedmecnetworkwithdeepreinforcementlearning AT qinli jointoptimizationofdnnpartitionandcontinuoustaskschedulingfordigitaltwinaidedmecnetworkwithdeepreinforcementlearning AT weiyuanli jointoptimizationofdnnpartitionandcontinuoustaskschedulingfordigitaltwinaidedmecnetworkwithdeepreinforcementlearning AT yongzhang jointoptimizationofdnnpartitionandcontinuoustaskschedulingfordigitaltwinaidedmecnetworkwithdeepreinforcementlearning |