楊志軍 孫洋洋
摘 要:針對提高輪詢控制模型工作效率和區(qū)分網(wǎng)絡優(yōu)先級的問題,提出了區(qū)分站點忙閑狀態(tài)的完全門限兩級輪詢控制模型(ETTPSS)。模型以兩級優(yōu)先級為基礎,依據(jù)站點的忙閑狀態(tài)采用并行處理方式只對忙站點進行信息分組發(fā)送服務。模型既能區(qū)分傳輸服務優(yōu)先級又能避開對無信息分組的空閑站點的查詢,從而提高了模型資源利用率和工作效率。運用概率母函數(shù)與馬爾可夫鏈相結(jié)合的方法對該模型進行理論分析研究,精確解析了模型各個重要性能參數(shù)。仿真實驗結(jié)果表明,仿真值與理論值近似相等,說明理論分析正確合理。與普通輪詢模型相比,該模型性能大幅度提高。
關鍵詞:優(yōu)先級;忙站點;輪詢模型;利用率;工作效率
中圖分類號:TN911
文獻標志碼:A
Abstract: To improve the work efficiency of polling control model and distinguish network priorities, an ExhaustiveThreshold Twostage Polling control model based on Site Status (ETTPSS) was proposed. Based on two levels of priority, parallel processing was used to only send information to busy sites according to busy and idle states of sites. The model could not only distinguish the priorities of transmission services but also avoid the queries to the idle sites without information packets, thereby improving model resource utilization and work efficiency. The method of probabilistic generating function and Markov chain was used to analyze the model theoretically, and the important performance parameters of the model were analyzed accurately. The simulation results show that the simulation values and the theoretical values are approximately equal, indicating that the theoretical analysis is correct and reasonable. Compared with normal polling model, the model performance is greatly improved.
英文關鍵詞Key words: priority; busy site; polling model; utilization; work efficiency
0 引言
輪詢控制模型具有服務質(zhì)量保障的優(yōu)點,一直是通信網(wǎng)絡中媒體訪問控制(Media Access Control, MAC)一種重要的調(diào)度方式,使其在現(xiàn)代網(wǎng)絡中應用非常普遍[1]。文獻[2]分析研究了輪詢控制模型在大數(shù)據(jù)流式計算平臺Apache Storm中的應用;文獻[3-5]分析研究了輪詢控制模型在計算機網(wǎng)絡異構(gòu)無線網(wǎng)絡以及信息采集的應用。
輪詢控制模型中,對無信息分組的空閑站點的查詢會浪費模型資源。文獻[6]通過對有信息分組發(fā)送需求的忙站點分配信道避免空閑查詢,且服務器完成對當前站點的信息發(fā)送后需要經(jīng)過一個轉(zhuǎn)換查詢時間才能對下一個需要信息發(fā)送的站點進行服務,而采用并行調(diào)度控制方式[7],就是把查詢和服務過程進行并行處理,不再消耗模型的轉(zhuǎn)換查詢時間。不過,輪詢表的生成與站點忙閑狀態(tài)相互獨立,特別是當站點空閑時間較長時,接收者每次輪詢都要對空閑站點進行查詢監(jiān)聽,造成模型的工作效率和資源利用率大幅度降低,并且也不能區(qū)分網(wǎng)絡業(yè)務優(yōu)先級。文獻[8]構(gòu)建“完全+門限”輪詢服務兩級模型以區(qū)分業(yè)務優(yōu)先級,但是該模型查詢服務包括空閑站點在內(nèi)的所有站點,信道利用率受到限制。文獻[9]提出區(qū)分站點狀態(tài)的限定(K=2)服務方式,文獻[10] 提出區(qū)分站點狀態(tài)的完全服務方式。雖然文獻[9-10]基于不同的輪詢服務方式來區(qū)分站點的忙閑狀態(tài),以降低系統(tǒng)的平均等待時間和能耗來提高系統(tǒng)網(wǎng)絡資源利用率,但并未設置中心站點和普通站點來區(qū)分網(wǎng)絡業(yè)務的傳輸優(yōu)先級。
針對上述問題,本文依據(jù)輪詢模型的動態(tài)性[11],提出了區(qū)分站點忙閑狀態(tài)的完全門限兩級輪詢控制模型(ExhaustiveThreshold Twolevel Polling control model based on Site Status, ETTPSS)。該模型算法與文獻[9-10]相比,最大的創(chuàng)新是進行中心站點與普通站點的兩級設置,中心站點傳輸高優(yōu)先級業(yè)務,普通站點傳輸?shù)蛢?yōu)先級業(yè)務,解決了網(wǎng)絡業(yè)務傳輸優(yōu)先級的問題。該模型算法與文獻[9-10]模型算法相同之處就是同樣根據(jù)站點的忙閑狀態(tài),對有發(fā)送需求的忙站點進行信息分組的發(fā)送服務,且服務過程與查詢過程采用并行處理方式,節(jié)省了轉(zhuǎn)換查詢時間,提高了模型工作效率。運用概率母函數(shù)[12]與馬爾可夫鏈[13]相結(jié)合的方法對該模型進行分析研究,仿真實驗表明該模型理論分析的正確合理性。
4 結(jié)語
本文提出了一種采用CNN算法進行UWB信道環(huán)境分類的方法,直接對信道統(tǒng)計特性進行特征提取,識別信道環(huán)境。實驗結(jié)果表明, 將CNN用于信道環(huán)境分類具有較高識別率,并且模型穩(wěn)定性比較高。
參考文獻 (References)
[1]CERNOHORSKY J, JANDURA P, RYDLO P. Real time ultrawideband localisation [C]// ICCC 2018: Proceedings of the 2018 19th International Carpathian Control Conference. Piscataway, NJ: IEEE, 2018: 445-450.
[2]LING P, SHEN C, ZHANG K, et al. An improved NLOS error elimination algorithm for indoor ultrawideband localization [C]// Proceedings of the 2017 IEEE SENSORS. Piscataway, NJ: IEEE, 2017: 1-3.
[3]LI X, ZHANG T. Research on improved UWB localization algorithm in NLOS environment [C]// Proceedings of the 2018 International Conference on Intelligent Transportation, Big Data and Smart City. Washington, DC: IEEE Computer Society, 2018:707-711.
[4]孫希延,劉健,紀元法,等.基于似然比檢驗的超寬帶信道檢測與定位算法[J].電子與信息學報,2017,39(3):590-597.(SUN X Y, LIU J, JI Y F, et al. UWB channel detection and location algorithm based on likelihood ratio test [J]. Journal of Electronics and Information Technology, 2017,39(3): 590-597.)
[5]ALBAIDHANI A, MORELL A, VICARIO J L. Ranging in UWB using commercial radio modules: experimental validation and NLOS mitigation [C]// Proceedings of the 2016 International Conference on Indoor Positioning and Indoor Navigation. Piscataway, NJ: IEEE, 2016: 1-7.
[6]MUQAIBEL A H, LANDOLSI M A, MAHMOOD M N. Practical evaluation of NLOS/LOS parametric classification in UWB channels [C]// Proceedings of the 2013 1st International Conference on Communications, Signal Processing and their Applications. Piscataway, NJ: IEEE, 2013:1-6.
[7]張浩,梁曉林,呂婷婷,等.一種新穎的基于偏度的非視距區(qū)分算法[J].電訊技術(shù),2015,55(5):484-490.(ZHANG H, LIANG X L, LYU T T, et al. A novel nonlineofsight identification algorithm based on skewness [J]. Telecommunication Engineering, 2015, 55(5): 484-490.)
[8]LIU M, SHI J, LI Z, et al. Towards better analysis of deep convolutional neural networks [J]. IEEE Transactions on Visualization and Computer Graphics, 2016, 23(1):91-100.
[9]李彥冬,郝宗波,雷航.卷積神經(jīng)網(wǎng)絡研究綜述[J].計算機應用,2016,36(9):2508-2515.(LI Y D, HAO Z B, LEI H. Survey of convolutional neural network [J]. Journal of Computer Applications, 2016, 36(9): 2508-2515.)
[10]王勇,周慧怡,俸皓,等.基于深度卷積神經(jīng)網(wǎng)絡的網(wǎng)絡流量分類方法[J].通信學報,2018,39(1):14-23.(WANG Y, ZHOU H Y, FENG H, et al. Network traffic classification method basing on CNN [J]. Journal on Communications, 2018, 39(1): 14-23.)
[11]YAZDANDOOST K Y. Channel modeling subcommittee report [R/OL].[2018-03-20]. Networks https://doi.org/IEEE P802.1502/490r1SG3a.
[12]WANG X. A UWB positioning scheme based on BP neural network in NLOS environment [C]// Proceedings of the 2011 7th International Conference on Wireless Communications, Networking and Mobile Computing. Piscataway, NJ: IEEE, 2011: 1-3.
[13]XIONG Z, CHEN K, GU C. An algorithm of image classification based on BP neural network [C]// Proceedings of the 2012 IEEE International Conference on Computer Science and Automation Engineering. Piscataway, NJ: IEEE, 2012: 523-526.
[14]LIU H, LIANG Z, LIU D, et al. Improved UWB indoor positioning algorithms based on BP neural network model [C]// Proceedings of the 12th International Conference on Communications and Networking in China. Berlin: Springer, 2017: 114-124.