Research output: Contribution to Journal/Magazine › Journal article › peer-review
Research output: Contribution to Journal/Magazine › Journal article › peer-review
}
TY - JOUR
T1 - Computation Offloading and Service Caching for Intelligent Transportation Systems with Digital Twin
AU - Xu, Xiaolong
AU - Liu, Zhongjian
AU - Bilal, Muhammad
AU - Vimal, S.
AU - Song, Houbing
PY - 2022/11/1
Y1 - 2022/11/1
N2 - Mobile edge computing (MEC) provides a novel computing paradigm to satisfy the increasing computation requirements of mobile applications. In MEC-enabled intelligent transportation systems (ITS), the latency-sensitive computing tasks are offloaded to RSUs for execution, reducing the transmission latency compared with the cloud solutions. However, the repetitive executions of the same tasks whose outputs are dependent on the inputs lead to the extra system latency, an alternative is to cache the required services on RSUs in advance. The service requirements of latency-sensitive computing tasks are satisfied by jointly considering computation offloading and service caching. Besides, the digital twin (DT) is utilized to construct the virtual world reflecting the physical world in real-time to efficiently make offloading strategies. In this paper, a computation offloading and service caching method using decision theory in ITS with DT, named CODT, is proposed. Specifically, the computation offloading and service caching in ITS is modeled first with DT. Then, a mixed-integer nonlinear programming (MINLP) problem is formulated to minimize the system latency. Afterward, the decision theory is used to analyze the utilities of offloading strategies in different states of RSUs and make the optimal strategy. Finally, extensive simulations based on the real-world datasets demonstrate that the proposed CODT outperforms other baselines.
AB - Mobile edge computing (MEC) provides a novel computing paradigm to satisfy the increasing computation requirements of mobile applications. In MEC-enabled intelligent transportation systems (ITS), the latency-sensitive computing tasks are offloaded to RSUs for execution, reducing the transmission latency compared with the cloud solutions. However, the repetitive executions of the same tasks whose outputs are dependent on the inputs lead to the extra system latency, an alternative is to cache the required services on RSUs in advance. The service requirements of latency-sensitive computing tasks are satisfied by jointly considering computation offloading and service caching. Besides, the digital twin (DT) is utilized to construct the virtual world reflecting the physical world in real-time to efficiently make offloading strategies. In this paper, a computation offloading and service caching method using decision theory in ITS with DT, named CODT, is proposed. Specifically, the computation offloading and service caching in ITS is modeled first with DT. Then, a mixed-integer nonlinear programming (MINLP) problem is formulated to minimize the system latency. Afterward, the decision theory is used to analyze the utilities of offloading strategies in different states of RSUs and make the optimal strategy. Finally, extensive simulations based on the real-world datasets demonstrate that the proposed CODT outperforms other baselines.
KW - Computation offloading
KW - decision theory
KW - digital twin
KW - intelligent transportation systems
KW - service caching
U2 - 10.1109/TITS.2022.3190669
DO - 10.1109/TITS.2022.3190669
M3 - Journal article
AN - SCOPUS:85135241447
VL - 23
SP - 20757
EP - 20772
JO - IEEE Transactions on Intelligent Transportation Systems
JF - IEEE Transactions on Intelligent Transportation Systems
SN - 1524-9050
IS - 11
ER -