研究生: |
陳詠舜 Yong-Shun Chen |
---|---|
論文名稱: |
具有多種服務速率與省電機制之發布/訂閱系統研究 A Study on the Pub/Sub System with Multiple Service Rates and Power Saving Mechanism |
指導教授: |
鍾順平
Shun-Ping Chung |
口試委員: |
林永松
Yeong-Sung Lin 王乃堅 Nai-Jian Wang |
學位類別: |
碩士 Master |
系所名稱: |
電資學院 - 電機工程系 Department of Electrical Engineering |
論文出版年: | 2022 |
畢業學年度: | 110 |
語文別: | 英文 |
論文頁數: | 245 |
中文關鍵詞: | 客戶端到服務器系統 、發布/訂閱系統 、時變距離 、服務速率 、省電機制 、中介器 、間歇排隊模型 、中間件層 、應用層 、存活期限 、到達速率 |
外文關鍵詞: | client–server system, pub/sub system, time-varying distance, service rate, power saving mechanism, broker, intermittent queueing model, middleware layer, application layer, lifetime limit, arrival rate |
相關次數: | 點閱:312 下載:0 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
在傳統緊密耦合的客戶端到服務器系統(client–server system)中,只要其中任一物件方不能正常運行,就無法接收或發布消息。而發布/訂閱系統(Pub/Sub system)則用於替換客戶端到服務器系統。發布/訂閱系統由發布器(publisher)、中介器(broker)和訂閱器(subscriber)所組成。因此,發布者與訂閱者鬆散耦合。在我們的研究中,為了表示不同物件之間的時變距離,我們研究了具有兩個以上服務狀態的間歇佇列,其中每個服務狀態都有相對應的服務速率。此外,具有兩種服務狀態,也就是ON和OFF的間歇排隊模型用來表示省電機制。中介器由M/M/1/K 佇列和間歇佇列串連而成。中間件層(middleware layer)的發布/訂閱系統由兩個 M/M/1/K 佇列和兩個間歇佇列相連。最外層的應用層(application layer)是通過在中間件層的頭和尾部分別添加一個ON和OFF的間歇排隊模型來連接的。另外,為了表達事件的及時性,發布的事件有一個存活期限。當佇列中的事件過期時,它會立即離開系統。除此之外,當事件到達每個佇列時,它會重新生成一個新的生命週期限制。首先,我們推導出所考慮系統的解析模型。其次,使用迭代算法來尋找穩態機率分佈和感興趣的效能指標。第三,我們研究了系統參數對我們感興趣的效能指標的影響,例如服務狀態的數量和到達速率。最後,我們編寫模擬程式來驗證解析結果。
In the traditional tightly coupled client–server system, as long as one object does not function properly, it will not be possible to receive or post messages. The Pub/Sub system is used to replace the client-server system. The Pub/Sub system is comprised of publishers, brokers, and subscribers. Thus, publishers are loosely coupled to subscribers. In our research, to characterize the time-varying distance between different objects, we study the intermittent queue with more than two service states, where each service state has a corresponding service rate. In addition, the intermittent queuing model with two service states, i.e., ON and OFF is used to represent a power saving mechanism. The broker is comprised of an M/M/1/K queue connected with an intermittent queue. The pub/sub system in the middleware layer is comprised of two M/M/1/K queues connected with two intermittent queues. The outermost application layer is connected by adding an intermittent queuing model with ON and OFF to the head and tail of the middleware layer, respectively. In addition, in order to express the event’s timeliness, the published event has a lifetime limit. When the event expires in the queue, it will immediately leave the system. Apart from this, when the event arrives at each queue, it will regenerate a new lifetime limit. First, we derive the analytical models for the considered systems. Second, an iterative algorithm is used for finding the steady state probability distribution and the interested performance measures. Third, we studied the effects of system parameters, such as the number of service states and arrival rate, on the performance measures of interest to us. Finally, we write the simulation programs to verify the analytical results.
[1] K. Birman and T. Joseph, "Exploiting Virtual Synchrony in Distributed Systems," SIGOPS Oper. Syst. Rev. 21, 1987, pp. 123–138, doi: 10.1145/37499.37515.
[2] G. Banavar, T. Chandra, B. Mukherjee, J. Nagarajarao, R. E. Strom, and D. C. Sturman, "An Efficient Multicast Protocol for Content-Based Publish-Subscribe Systems," Proceedings. 19th IEEE International Conference on Distributed Computing Systems (Cat. No.99CB37003), 1999, pp. 262-272, doi: 10.1109/ICDCS.1999.776528.
[3] P. Nguyen and K. Nahrstedt, "Resource Management for Elastic Publish Subscribe Systems: A Performance Modeling-Based Approach," 2016 IEEE 9th International Conference on Cloud Computing (CLOUD), 2016, pp. 561-568, doi: 10.1109/CLOUD.2016.0080.
[4] Andrew Banks and Rahul Gupta. MQTT Version 3.1. 1. OASIS standard.
[5] Oasis advanced message queuing protocol. AMQP Version 1.0. OASIS standard.
[6] "RabbitMQ," https://www.rabbitmq.com. 2018.
[7] "JMS," http://www.oracle.com/technetwork/java/jms/index.html. 2018.
[8] "Apache Kafka," http://kafka.apache.org/. 2018.
[9] T. H. Nguyen, M. Forshaw, and N. Thomas, "Operating Policies for Energy Efficient Dynamic Server Allocation," Electronic Notes in Theoretical Computer Science, 2015, pp.159-177, doi: 10.1016/j.entcs.2015.10.025
[10] P.D. Tuan, "Server Farms with Batch Arrival and Staggered Setup," Proceedings of the Fifth Symposium on Information and Communication Technology, 2014, doi: 10.1145/2676585.2676613.
[11] G. Bouloukakis, N. Georgantas, A. Kattepur, and V. Issarny, "Timeliness Evaluation of Intermittent Mobile Connectivity over Pub/Sub Systems," ACMISPEC ICPE, 2017, pp. 275-286, doi: 10.1145/3030207.3030220.
[12] G. Bouloukakis, A. Kattepur, N. Georgantas, and V. Issarny, "Queueing Network Modeling Patterns for Reliable and Unreliable Publish / Subscribe Protocols," MobiQuitous, 2018, pp. 176-186, doi: 10.1145/3286978.3287002.
[13] G. Bouloukakis, I. Moscholios, and N. Georgantas, "Probabilistic Event Dropping for Intermittently Connected Subscribers over Pub/Sub Systems," IEEE ICC, 2019, pp. 1-6, doi: 10.1109/ICC.2019.8761557.
[14] Y. M. Gu, "A Study on the Pub/Sub System with More Than Two Service Rates," Master thesis, NTUST, 2021.
[15] T. M. C. Chu, H. Phan, and H. J. Zepernick, "Adaptive Modulation and Coding with Queue Awareness in Cognitive Incremental Decode-and-forward Relay Networks," 2014 IEEE International Conference on Communications (ICC), 2014, pp. 1453-1459, doi: 10.1109/ICC.2014.6883526.
[16] D. Gross, J. F. Shortle, J. M. Thomposon, and C. M. Harris, "Fundamentals of Queueing Theory".
[17] D. Yue, W. y. Yue, and G. Xu, " Analysis of Customers' Impatience in an M/M/1 Queue with Working Vacations," Journal of Industrial and Management Optimization, 2012, pp. 895-908, doi: 10.3934/jimo.2012. 8.895.
[18] H. Takagi, "Waiting Time in THE M/M/m/(m + c) Queue with Impatient Customers," International Journal of Pure and Applied Mathematics, 2014, pp. 519-559, doi: 10.12732/ijpam.v90i4.13.