Research output: Contribution to Journal/Magazine › Journal article › peer-review
Research output: Contribution to Journal/Magazine › Journal article › peer-review
}
TY - JOUR
T1 - A Hybrid Computing Solution and Resource Scheduling Strategy for Edge Computing in Smart Manufacturing
AU - Li, Xiaomin
AU - Wan, Jiafu
AU - Dai, Hong Ning
AU - Imran, Muhammad
AU - Xia, Min
AU - Celesti, Antonio
PY - 2019/7/1
Y1 - 2019/7/1
N2 - At present, smart manufacturing computing framework has faced many challenges such as the lack of an effective framework of fusing computing historical heritages and resource scheduling strategy to guarantee the low-latency requirement. In this paper, we propose a hybrid computing framework and design an intelligent resource scheduling strategy to fulfill the real-time requirement in smart manufacturing with edge computing support. First, a four-layer computing system in a smart manufacturing environment is provided to support the artificial intelligence task operation with the network perspective. Then, a two-phase algorithm for scheduling the computing resources in the edge layer is designed based on greedy and threshold strategies with latency constraints. Finally, a prototype platform was developed. We conducted experiments on the prototype to evaluate the performance of the proposed framework with a comparison of the traditionally-used methods. The proposed strategies have demonstrated the excellent real-time, satisfaction degree (SD), and energy consumption performance of computing services in smart manufacturing with edge computing.
AB - At present, smart manufacturing computing framework has faced many challenges such as the lack of an effective framework of fusing computing historical heritages and resource scheduling strategy to guarantee the low-latency requirement. In this paper, we propose a hybrid computing framework and design an intelligent resource scheduling strategy to fulfill the real-time requirement in smart manufacturing with edge computing support. First, a four-layer computing system in a smart manufacturing environment is provided to support the artificial intelligence task operation with the network perspective. Then, a two-phase algorithm for scheduling the computing resources in the edge layer is designed based on greedy and threshold strategies with latency constraints. Finally, a prototype platform was developed. We conducted experiments on the prototype to evaluate the performance of the proposed framework with a comparison of the traditionally-used methods. The proposed strategies have demonstrated the excellent real-time, satisfaction degree (SD), and energy consumption performance of computing services in smart manufacturing with edge computing.
KW - Edge computing
KW - industry 4.0
KW - resource scheduling
KW - smart manufacturing
U2 - 10.1109/TII.2019.2899679
DO - 10.1109/TII.2019.2899679
M3 - Journal article
AN - SCOPUS:85063689140
VL - 15
SP - 4225
EP - 4234
JO - IEEE Transactions on Industrial Informatics
JF - IEEE Transactions on Industrial Informatics
SN - 1551-3203
IS - 7
M1 - 8643392
ER -