logo

SCIENCE CHINA Information Sciences, Volume 63 , Issue 4 : 140313(2020) https://doi.org/10.1007/s11432-019-2780-0

Joint time delay and energy optimization with intelligent overclocking in edge computing

More info
  • ReceivedOct 19, 2019
  • AcceptedFeb 4, 2020
  • PublishedMar 9, 2020

Abstract

With the rapid growth of user equipment (UE), the amount of data transmitted over networks has become enormous, exerting immense pressure on backbone networks and central cloud infrastructures. Simultaneously, corresponding applications requiring high energy consumption and low latency have multiplied the requirements for UE. Mobile edge computing (MEC) has been proposed to support the offloading of UE tasks to edge clouds for execution. The implementation of MEC requires fast data transmission between UE and edge servers, and the emerging 5G network appears to render this technology possible. In this paper, considering a large number of UE, a fixed MEC server, and an advanced intelligent network, we suggest an intelligent overclocking mechanism for the MEC server that operates for an intelligently calculated period to allow it to leverage more computing power without introducing additional hardware resources for a certain period of time. We jointly manage task offloading, server resource allocation, and overclocking to minimize the system-wide computation overhead and other risks. The proposed optimization problem is a mixed-integer nonlinear programming problem that is divided into three subproblems: offloading decision, resource allocation, and overclocking decision. We solve these subproblems using non-convex techniques and provide an iterative algorithm to obtain a heuristic solution for the original problem. Finally, simulation results show that the overclocked MEC server has lower system-wide computation overhead, faster task processing, and more offloaded UE as compared with the case without overclocking.


Acknowledgment

This work was supported by National Natural Science Foundation of China (Grant Nos. 61672395, 61972448, 61911540481), Fund of Hubei Key Laboratory of Inland Shipping Technology (Grant No. NHHY2019004), and National Research Foundation of Korea (NRF) Grant Funded by the Korea Government (MSIT) (Grant No. 2019K2A9A2A060-24389).


References

[1] Atat R, Liu L, Chen H. Enabling cyber-physical communication in 5G cellular networks: challenges, spatial spectrum sensing, and cyber-security. 2017, PP: 49-54 CrossRef Google Scholar

[2] Ning Z, Huang J, Wang X. Vehicular Fog Computing: Enabling Real-Time Traffic Management for Smart Cities. IEEE Wireless Commun, 2019, 26: 87-93 CrossRef Google Scholar

[3] Reddy K Y, Gandhi N V D, Balachander K. Simulation and analysis of performance models of broadband intelligent mobile networks. In: Proceedings of 2007 International Conference on Signal Processing, Communications and Networking, 2007. 573--578. Google Scholar

[4] Jiang W Strufe M, Schotten H D. Intelligent network management for 5g systems: the selfnet approach. In: Proceedings of 2017 European Conference on Networks and Communications (EuCNC) 2017. 1--5. Google Scholar

[5] Fangming Liu , Peng Shu , Hai Jin . Gearing resource-poor mobile devices with powerful clouds: architectures, challenges, and applications. IEEE Wireless Commun, 2013, 20: 14-22 CrossRef Google Scholar

[6] Chiang M, Zhang T. Fog and IoT: An Overview of Research Opportunities. IEEE Internet Things J, 2016, 3: 854-864 CrossRef Google Scholar

[7] Lin J, Yu W, Zhang N. A Survey on Internet of Things: Architecture, Enabling Technologies, Security and Privacy, and Applications. IEEE Internet Things J, 2017, 4: 1125-1142 CrossRef Google Scholar

[8] Sabella D, Vaillant A, Kuure P. Mobile-Edge Computing Architecture: The role of MEC in the Internet of Things. IEEE Consumer Electron Mag, 2016, 5: 84-91 CrossRef Google Scholar

[9] Corcoran P, Datta S K. Mobile-Edge Computing and the Internet of Things for Consumers: Extending cloud computing and services to the edge of the network. IEEE Consumer Electron Mag, 2016, 5: 73-74 CrossRef Google Scholar

[10] Sun X, Ansari N. EdgeIoT: Mobile Edge Computing for the Internet of Things. IEEE Commun Mag, 2016, 54: 22-29 CrossRef Google Scholar

[11] Zhang G, Chen Y, Shen Z. Distributed Energy Management for Multiuser Mobile-Edge Computing Systems With Energy Harvesting Devices and QoS Constraints. IEEE Internet Things J, 2019, 6: 4035-4048 CrossRef Google Scholar

[12] Abbas N, Zhang Y, Taherkordi A. Mobile Edge Computing: A Survey. IEEE Internet Things J, 2018, 5: 450-465 CrossRef Google Scholar

[13] Mach P, Becvar Z. Mobile Edge Computing: A Survey on Architecture and Computation Offloading. IEEE Commun Surv Tutorials, 2017, 19: 1628-1656 CrossRef Google Scholar

[14] Shan X, Zhi H, Li P, et al. A survey on computation offloading for mobile edge computing information. In: Proceedings of 2018 IEEE 4th International Conference on Big Data Security on Cloud (BigDataSecurity), IEEE International Conference on High Performance and Smart Computing, (HPSC), IEEE International Conference on Intelligent Data and Security (IDS) 2018. 248--251. Google Scholar

[15] Kosmides P, Lambrinos L. Intelligent routing in mobile opportunistic networks. In: Proceedings of 2018 Global Information Infrastructure and Networking Symposium (GIIS) 2018. 1--4. Google Scholar

[16] Alameddine H A, Sharafeddine S, Sebbah S. Dynamic Task Offloading and Scheduling for Low-Latency IoT Services in Multi-Access Edge Computing. IEEE J Sel Areas Commun, 2019, 37: 668-682 CrossRef Google Scholar

[17] Ning Z, Dong P, Kong X. A Cooperative Partial Computation Offloading Scheme for Mobile Edge Computing Enabled Internet of Things. IEEE Internet Things J, 2019, 6: 4804-4814 CrossRef Google Scholar

[18] Wang H, Li X, Ji H, et al. Federated offloading scheme to minimize latency in mec-enabled vehicular networks. In: Proceedings of 2018 IEEE Globecom Workshops (GC Wkshps) 2018. 1--6. Google Scholar

[19] Bi S, Zhang Y J. Computation Rate Maximization for Wireless Powered Mobile-Edge Computing With Binary Computation Offloading. IEEE Trans Wireless Commun, 2018, 17: 4177-4190 CrossRef Google Scholar

[20] Li S, Tao Y, Qin X. Energy-Aware Mobile Edge Computation Offloading for IoT Over Heterogenous Networks. IEEE Access, 2019, 7: 13092-13105 CrossRef Google Scholar

[21] Cui L, Xu C, Yang S. Joint Optimization of Energy Consumption and Latency in Mobile Edge Computing for Internet of Things. IEEE Internet Things J, 2019, 6: 4791-4803 CrossRef Google Scholar

[22] Bu S, Yu F R. Green Cognitive Mobile Networks With Small Cells for Multimedia Communications in the Smart Grid Environment. IEEE Trans Veh Technol, 2014, 63: 2115-2126 CrossRef Google Scholar

[23] Pham Q V, Le L B, Chung S H. Mobile Edge Computing With Wireless Backhaul: Joint Task Offloading and Resource Allocation. IEEE Access, 2019, 7: 16444-16459 CrossRef Google Scholar

[24] Siddique U, Tabassum H, Hossain E. Wireless backhauling of 5G small cells: challenges and solution approaches. IEEE Wireless Commun, 2015, 22: 22-31 CrossRef Google Scholar

[25] Ge X, Cheng H, Guizani M. 5G wireless backhaul networks: challenges and research advances. IEEE Network, 2014, 28: 6-11 CrossRef Google Scholar

[26] Tran T X, Pompili D. Joint Task Offloading and Resource Allocation for Multi-Server Mobile-Edge Computing Networks. IEEE Trans Veh Technol, 2019, 68: 856-868 CrossRef Google Scholar

[27] Zhang J, Hu X, Ning Z. Energy-Latency Tradeoff for Energy-Aware Offloading in Mobile Edge Computing Networks. IEEE Internet Things J, 2018, 5: 2633-2645 CrossRef Google Scholar

[28] Thomas D, Shanmugasundaram M. A survey on different overclocking methods. In: Proceedings of 2018 Second International Conference on Electronics, Communication and Aerospace Technology (ICECA) 2018. 1588--1592. Google Scholar

[29] Wu F, Chen J, Dong Y, et al. Improve energy efficiency by processor overclocking and memory frequency scaling. In: Proceedings of 2018 IEEE 20th International Conference on High Performance Computing and Communications, IEEE 16th International Conference on Smart City, IEEE 4th International Conference on Data Science and Systems (HPCC/SmartCity/DSS) 2018. 960--967. Google Scholar

[30] Leveraging Process Variation for Performance and Energy: In the Perspective of Overclocking. IEEE Trans Comput, 2014, 63: 1316-1322 CrossRef Google Scholar

[31] Short M, Sheikh I. Dual-rate overclocking in can networks: a soft-core controller prototype. In: Proceedings of 2010 Seventh International Conference on Networked Sensing Systems (INSS) 2010. 314--317. Google Scholar

[32] Kai Zhao , Jiangpeng Li , Jun Ma . Overclocking NAND Flash Memory I/O Link in LDPC-Based SSDs. IEEE Trans Circuits Syst II, 2014, 61: 885-889 CrossRef Google Scholar

[33] Wang C, Yu F R, Liang C. Joint Computation Offloading and Interference Management in Wireless Cellular Networks with Mobile Edge Computing. IEEE Trans Veh Technol, 2017, 66: 7432-7445 CrossRef Google Scholar

  • Figure 1

    (Color online) System model.

  • Figure 2

    Proposed framework for solving the problem (12).

  • Figure 3

    (Color online) (a) Loss function $L(t)$ vs. time $t$; (b) task allocation resource $f^r_n$ vs. processing time $t^r_n$.

  • Figure 4

    (Color online) (a) Computation overhead vs. the number of iterations; (b) computational overhead vs. the number of UEs.

  • Figure 5

    (Color online) (a) Computational overhead in different states; (b) benefit difference between overclocking and non-overclocking.

  • Figure 6

    (Color online) (a) Processing time of MEC servers in two states; (b) computational overhead in different states.

  • Figure 7

    (Color online) (a) The number of UEs changes with time; (b) computational resource overhead vs. time.

  •   

    Algorithm 1 Joint optimization for offloading decision, overclocking decision, and computation resource allocation (JOOC)

    Initialization:

    Make $\boldsymbol{x}=\{x_n=1~|~\forall~n~\in~\mathcal{N}$, $\mathcal{N}_{\mathrm{Off}}=\mathcal{N}$.

    Iteration:

    for $a$ $\leftarrow$ 0 to 1

    Get $\boldsymbol{f}$ according to (16), (17) or (39).

    if $\exists~n~\in~\mathcal{N}_{\mathrm{Off}}$, $U^r_n~>~U^l_n$ then

    Repeat

    $i^{\#}~\leftarrow~\mathop{\arg}~\min~\nolimits_{n~\in~\mathcal{N}_{\mathrm{Off}}}~\{x_n(U^l_n-U^r_n)\}$;

    $\mathcal{N}_{\mathrm{Off}}~\leftarrow~\mathcal{N}_{\mathrm{Off}}~-~\{i^{\#}\}$;

    Update $\boldsymbol{x}$ by setting $x_{i^{\#}}~\leftarrow~0$;

    Update $\boldsymbol{f}$ according to (16), (17) or (39);

    Until $U^r_n~<~U^l_n$, $\forall~n~\in~\mathcal{N}_{\mathrm{Off}}$ or $\mathcal{N}_{\mathrm{Off}}~=~\phi$;

    end if

    if $\exists~n~\in~\mathcal{N}_{\mathrm{Off}}$, $(t^{p}_n+t^{r}_n)>~T^{\rm~max}_n$ then

    Repeat

    $i^{\#}~\leftarrow~\mathop{\arg}~\min~\nolimits_{n~\in~\mathcal{N}_{\mathrm{Off}}}~\{x_n(T^{\rm~max}_n~-~(t^{p}_n+t^{r}_n))\}$;

    $\mathcal{N}_{\mathrm{Off}}~\leftarrow~\mathcal{N}_{\mathrm{Off}}~-~\{i^{\#}\}$;

    Update $\boldsymbol{x}$ by setting $x_{i^{\#}}~\leftarrow~0$;

    Update $\boldsymbol{f}$ according to (16), (17) or (39);

    Until $(t^{p}_n+t^{r}_n)~<~T^{\rm~max}_n$, $\forall~n~\in~\mathcal{N}_{\mathrm{Off}}$ or $\mathcal{N}_{\mathrm{Off}}~=~\phi~$;

    end if

    end for

    Update $a$ according to (40);

    Output: the optimal solution $\boldsymbol{x}$, $a$, $\boldsymbol{f}$.

  • Table 1   The simulation parameters
    Parameter name Parameter value Unit
    Bandwidth $W$ $2.~5~\times~10^7$ Hz
    Transmission power $P_n$ 20 dBm
    Nose power $N_0$ $-$85 dBm
    CPU frequency of UEs $f^l_n$ 0.5 GHz
    Total CPU frequency F 20 GHz
    Task CPU cycles $C_n$ [0.5,~0.7] Gigacycle
    Task date size $D_n$ [20,~350]$\times~10^3$ kB
    Task QoS $T^{\rm~max}$ [1.0,~1.1] s
    Time weighting factor $\lambda^t_n$ 0.5
    Energy weighting factor $\lambda^e_n$ 0.5
    Growth rate of loss function $\alpha$ 0.3

Copyright 2020  CHINA SCIENCE PUBLISHING & MEDIA LTD.  中国科技出版传媒股份有限公司  版权所有

京ICP备14028887号-23       京公网安备11010102003388号