車(chē)聯(lián)網(wǎng)中一種基于軟件定義網(wǎng)絡(luò)與移動(dòng)邊緣計(jì)算的卸載策略
doi: 10.11999/JEIT190304 cstr: 32379.14.JEIT190304
-
1.
重慶郵電大學(xué)通信與信息工程學(xué)院 重慶 400065
-
2.
武漢大學(xué)電子信息學(xué)院 武漢 430000
An Offloading Mechanism Based on Software Defined Network and Mobile Edge Computing in Vehicular Networks
-
1.
School of Communication and Information Engineering, Chongqing University of Posts andTelecommunications, Chongqing 400065, China
-
2.
School of Electronic Information, Wuhan University, Wuhan 430000, China
-
摘要:
在新興的車(chē)聯(lián)網(wǎng)絡(luò)中，汽車(chē)終端請(qǐng)求卸載的任務(wù)對(duì)網(wǎng)絡(luò)帶寬、卸載時(shí)延等有著更加嚴(yán)苛的需求，而新型通信網(wǎng)絡(luò)研究中移動(dòng)邊緣計(jì)算(MEC)的提出更好地解決了這一挑戰(zhàn)。該文著重解決的是汽車(chē)終端進(jìn)行任務(wù)卸載時(shí)卸載對(duì)象的匹配問(wèn)題。文中引入了軟件定義車(chē)載網(wǎng)絡(luò)(SDN-V)對(duì)全局變量統(tǒng)一調(diào)度，實(shí)現(xiàn)了資源控制管理、設(shè)備信息采集以及任務(wù)信息分析?；谟脩羧蝿?wù)的差異化性質(zhì)，定義了重要度的模型，在此基礎(chǔ)上，通過(guò)設(shè)計(jì)任務(wù)卸載優(yōu)先級(jí)機(jī)制算法，實(shí)現(xiàn)任務(wù)優(yōu)先級(jí)劃分。針對(duì)多目標(biāo)優(yōu)化模型，采用乘子法對(duì)非凸優(yōu)化模型進(jìn)行求解。仿真結(jié)果表明，與其他卸載策略相比，該文所提卸載機(jī)制對(duì)時(shí)延和能耗優(yōu)化效果明顯，能夠最大程度地保證用戶的效益。
-
關(guān)鍵詞:
- 車(chē)聯(lián)網(wǎng) /
- 軟件定義網(wǎng)絡(luò) /
- 移動(dòng)邊緣計(jì)算 /
- 卸載機(jī)制
Abstract:In the emerging vehicular networks, the task of the car terminal requesting offloading has more stringent requirements for network bandwidth and offload delay, and the proposed Mobile Edge Computing (MEC) in the new communication network research solves better this challenge. This paper focuses on matching the offloaded objects when the car terminal performs the task offloading. By introducing the Software-Defined in-Vehicle Network (SDN-V) to schedule uniformly global variables, which realizes resource control management, device information collection and task information analysis. Based on the differentiated nature of user tasks, a model of importance is defined. On this basis, task priority is divided by designing the task to offload the priority mechanism. For the multi-objective optimization model, the non-convex optimization model is solved by the multiplier method. The simulation results show that compared with other offloading strategies, the proposed offloading mechanism has obvious effects on delay and energy consumption optimization, which can guarantee the benefit of users to the greatest extent.
-
表 1 任務(wù)卸載優(yōu)先級(jí)機(jī)制
(1) 輸入：車(chē)輛$i$的請(qǐng)求信息為$\{ {C_i},{S_i},t_{{Q_i}}^{\max }\} $，定義$\zeta $的取值，$i \in \{ 1\; 2\; ··· \; n\} $, ${\rm{Im}}{{\rm{p}}_{\rm{i}}}{\rm{ = \{ im}}{{\rm{p}}_{\rm{1}}}{\kern 1pt} {\kern 1pt} {\rm{im}}{{\rm{p}}_{\rm{2}}}\; ···\; {\rm{im}}{{\rm{p}}_{{n}}}\; {\rm{\} }}$ (2) 輸出：降序排列的重要度${\rm{im}}{{\rm{p}}_i}$ (3) for $i = 1;i < n;i + + $ (4) 將${C_i},t_{{Q_i}}^{\max }$代入式(9)求出${\rm{im}}{{\rm{p}}_i}$ (5) ${\rm{Im}}{{\rm{p}}_{\rm{i}}}={\rm{\{ im}}{{\rm{p}}_{\rm{1}}}{\kern 1pt} {\kern 1pt} {\rm{im}}{{\rm{p}}_{\rm{2}}}{\kern 1pt} {\kern 1pt} ···\; {\rm{im}}{{\rm{p}}_{{i}}}{\rm{\} }}$ (6) for $i = 1:n$ do (7) if ${{{\rm Imp}(i) < {\rm Imp}(i + 1)}}$; ${{\rm temp} = {\rm Imp}(i + 1)}$; ${{{\rm Imp}(i + 1) = {\rm Imp}(i)}}{\kern 1pt} {\kern 1pt} {\kern 1pt} ;{{{\rm Imp}(i) = {\rm temp}}}$ (8) end 
下載: 導(dǎo)出CSV
表 2 基于Q-學(xué)習(xí)的任務(wù)卸載策略機(jī)制
(1) 輸入：車(chē)輛$i$的請(qǐng)求信息$\{ {Q_i},{T_i}\} $, ${\tau _{\rm{1}}},{\tau _2},({\rm{0 < }}{\tau _{\rm{1}}} < {\tau _{\rm{2}}})$, $i \in \{ 1\; 2\; ··· \; n\} $, ${\rm{Im}}{{\rm{p}}_{{i}}}{\rm{ = \{ im}}{{\rm{p}}_{\rm{1}}}{\kern 1pt} {\kern 1pt} {\rm{im}}{{\rm{p}}_{\rm{2}}}\; ···\; {\rm{im}}{{\rm{p}}_{{i}}}{\rm{\} }}$ (2) 輸出：${x_i}$, ${\psi _i}$ (3) if ${\rm{im}}{{\rm{p}}_i} < {\tau _{\rm{1}}}$：${x_i}=0$；${\kern 1pt} {\kern 1pt} {\rm{im}}{{\rm{p}}_i} > {\tau _2}$：${x_i}{\rm{ = 1}}$ (4) elif ${\tau _{\rm{1}}} < {\rm{im}}{{\rm{p}}_i} < {\tau _{\rm{2}}}$：初始化$g$, ${x_{ij}} = 1$, $\varsigma $, $p$, $\hat Q\left( {{a_i}} \right) = 0,\; {\kern 1pt} t = 0$最大收斂時(shí)間${t_{c - \max }}$ (5) while ${\kern 1pt} t < {t_{c - \max }} + 1$：按照時(shí)延約束對(duì)車(chē)輛用戶排序 (6) for $i = 1:N\; {\kern 1pt} {\kern 1pt} $ do (7) 根據(jù)貪婪方法選擇行為${a_i}$、根據(jù)式(15)求出用戶獎(jiǎng)勵(lì) (8) 更新$\hat { Q}$數(shù)值矩陣通過(guò)${\hat Q_{t + 1}}\left( {s,a} \right) \leftarrow \left( {1 - \varsigma } \right){\hat Q_t}\left( {s,a} \right) + \varsigma \left( {g + \eta \mathop {\max }\limits_{a'} {{\hat Q}_t}\left( {s',a'} \right)} \right)$, $p \leftarrow \left( {p/\sqrt t } \right)$ (9) end for；$t = t + 1$；end while (10) 利用${\psi _i}$更新目標(biāo)優(yōu)化式(7) (11) end
下載: 導(dǎo)出CSV
表 3 模擬參數(shù)表
參數(shù) 數(shù)值 計(jì)算任務(wù)${Q_i}$ 1～50 MB 傳輸帶寬$W$ 100 MHz 汽車(chē)用戶發(fā)射功率${p_i}$ 0.2 W 任務(wù)所需CPU周期數(shù)${C_i}$ 0.1～1 GHz MEC服務(wù)器CPU周期頻率${f_{\rm b}}$ 6 GHz 車(chē)輛用戶的CPU周期頻率${f_v}$ 0.5～1 GHz 高斯噪聲${\sigma ^2}$ –100 dBm 信道傳輸距離${d_{mn}}$ 5～500 m 汽車(chē)CPU能耗功率系數(shù)${p_{{v} } }$ 80 W/GHz 電池最大容量 20 kWh
下載: 導(dǎo)出CSV
-
程剛, 郭達(dá). 車(chē)聯(lián)網(wǎng)現(xiàn)狀與發(fā)展研究[J]. 移動(dòng)通信, 2011, 35(17): 23–26. doi: 10.3969/j.issn.1006-1010.2011.17.004CHENG Gang, and GUO Da. Research on the status and development of internet of vehicles[J]. Mobile Communications, 2011, 35(17): 23–26. doi: 10.3969/j.issn.1006-1010.2011.17.004 齊彥麗, 周一青, 劉玲, 等. 融合移動(dòng)邊緣計(jì)算的未來(lái)5G移動(dòng)通信網(wǎng)絡(luò)[J]. 計(jì)算機(jī)研究與發(fā)展, 2018, 55(3): 478–486. doi: 10.7544/issn1000-1239.2018.20170801QI Yanli, ZHOU Yiqing, LIU Ling, et al. MEC coordinated future 5G mobile wireless networks[J]. Journal of Computer Research and Development, 2018, 55(3): 478–486. doi: 10.7544/issn1000-1239.2018.20170801 LI Yong, JIN Depeng, HUI Pan, et al. Optimal base station scheduling for device-to-device communication underlaying cellular networks[J]. IEEE Journal on Selected Areas in Communications, 2015, 34(1): 27–40. doi: 10.1109/JSAC.2015.2452415 AMEMIYA K, AKIYAMA Y, KOBAYASHI K, et al. On-site evaluation of a software cellular based MEC system with downlink slicing technology[C]. The 7th IEEE International Conference on Cloud Networking, Tokyo, Japan, 2018: 1–7. doi: 10.1109/CloudNet.2018.8549380. WANG Hansong, LI Xi, JI Hong, et al. Dynamic offloading scheduling scheme for MEC-enabled vehicular networks[C]. 2018 IEEE/CIC International Conference on Communications in China, Beijing, China, 2018: 206–210. doi: 10.1109/ICCChinaW.2018.8674508. YU Rong, DING Jiefei, HUANG Xumin, et al. Optimal resource sharing in 5G-enabled vehicular networks: A matrix game approach[J]. IEEE Transactions on Vehicular Technology, 2016, 65(10): 7844–7856. doi: 10.1109/tvt.2016.2536441 鮑楠, 左加闊, 胡晗, 等. 基于SDN的網(wǎng)絡(luò)資源選擇多目標(biāo)優(yōu)化算法[J]. 通信學(xué)報(bào), 2019, 40(2): 51–59. doi: 10.11959/j.issn.1000?436x.2019031BAO Nan, ZUO Jiakuo, HU Han, et al. SDN based network resource selection multi-objective optimization algorithm[J]. Journal on Communications, 2019, 40(2): 51–59. doi: 10.11959/j.issn.1000?436x.2019031 CHEN Min and HAO Yixue. Task offloading for mobile edge computing in software defined ultra-dense network[J]. IEEE Journal on Selected Areas in Communications, 2018, 36(3): 587–597. doi: 10.1109/JSAC.2018.2815360 ZHANG Ke, MAO Yuming, LENG Supeng, et al. Optimal delay constrained offloading for vehicular edge computing networks[C]. 2017 IEEE International Conference on Communications, Paris, France, 2017: 1–6. doi: 10.1109/ICC.2017.7997360. HU Bo, CHEN Jianye, and LI fengcun. Dynamic service allocation algorithm in mobile edge computing[C]. 2017 International Conference on Information and Communication Technology Convergence, Jeju, South Korea, 2017: 104–109. doi: 10.1109/ICTC.2017.8190951. LI Baozhu, ZHAO Xuhui, HAN Shiyuan, et al. New SDN-based architecture for integrated vehicular cloud computing networking[C]. 2018 International Conference on Selected Topics in Mobile and Wireless Networking, Tangier, Morocco, 2018: 1–4. doi: 10.1109/MoWNet.2018.8428935. HUANG Xumin, YU Rong, KANG Jiawen, et al. Distributed reputation management for secure and efficient vehicular edge computing and networks[J]. IEEE Access, 2017, 5: 25408–25420. doi: 10.1109/ACCESS.2017.2769878 TRAN T X and POMPILI D. Joint task offloading and resource allocation for multi-server mobile-edge computing networks[J]. IEEE Transactions on Vehicular Technology, 2019, 68(1): 856–868. doi: 10.1109/TVT.2018.2881191 JAIN R and PAUL S. Network virtualization and software defined networking for cloud computing: A survey[J]. IEEE Communications Magazine, 2013, 51(11): 24–31. doi: 10.1109/MCOM.2013.6658648 WILHELMI F, BELLALTA B, CANO C, et al. Implications of decentralized Q-learning resource allocation in wireless networks[C]. The 28th IEEE Annual International Symposium on Personal, Indoor, and Mobile Radio Communications, Montreal, Canada, 2017: 1–5. doi: 10.1109/PIMRC.2017.8292321. -
計(jì)量
- 文章訪問(wèn)數(shù): 4182
- HTML全文瀏覽量: 2191
- PDF下載量: 255
- 被引次數(shù): 0
下載:
