臺灣博碩士論文加值系統

流量彙整技術解決了單一波長頻道的容量與各類別連線容量無法匹配的議題，然而，在光纖流量彙整網路中，仍存在著容量公平性的議題，其解決方式是藉由實現允入控制機制。現今允入控制機制主要是來自下列不同的技術之一，分別為靜態頻寬保留、靜態臨界值設定、數學統計及馬可夫決策過程。然而，不管是採取哪一項方法，對於公平性與網路產量的取捨問題依舊存在。因此，根據網路中各連結中每一條可用波長的頻寬使用量，藉由馬可夫決策過程取得網路產量最大化的最佳策略，並依照此最佳策略決定需求連線的允入與否的Dynamic Preemption Call Admission Based on Markov Decision Process (DP-MDP)便被提出，藉以提升網路產量的同時維持容量公平性。
然而，隨著節能意識的增長，將網路硬體及通訊協定導入能源效率的考量已成為一種趨勢。特別是在多波長分波多工的網路中，基於功率感知的路由及波長配置演算法可用來降低多波長分波多工網路的基礎建設耗能。因此，基於功率感知的流量彙整光網路似乎是下一個合適的基礎建設。基於動態可搶先馬可夫決策之具備能量感知的公平允入機制 (EA- DPMDP)藉由馬可夫決策過程取得每單位能耗的網路產量最大化的最佳策略，並根據此最佳策略決定需求連線的與入與否，以確保在減少能耗的情況下維持容量公平性與網路產量。我們首先以一 雙向柱狀網路拓樸針對加入EA-DPMDP以及非能耗考量的DP-MDP進行模擬，並比較網路產量、網路能耗及容量公平性的差異。結果顯示EA-DPMDP在減少網路能耗時，依舊能維持容量公平性。接著我們使用相同的網路拓樸比較EA-DPMDP以及各種具能耗感知的路由與波長分配如RSB、LUB、TATG於網路產量、網路能耗及公平率上的差異，以及比較各種具能耗感知的路由與波長分配機制搭配EA-DPMDP對於網路節能效率與容量公平性的影響，結果顯示EA-DPMDP能夠有效的減少網路能耗同時維持容量公平性，但是路徑較長的路由分配機制如TATG在低負載下有著較差的容量公平性。為確認路由長度對於EA-DPMDP的影響，我們接著比較EA-DPMDP在不同網路拓樸下對容量公平性的影響，結果顯示高需求容量類別的連線需求在路徑長度較短的網路拓樸如NSF網路拓樸、Cost239網路拓樸和Random網路拓樸中，因阻塞機率受路由長度的影響降低，加上EA-DPMDP偏向接受高需求容量類別連線需求，使得EA-DPMDP在低負載有著較差的容量公平性。最後，我們比較不同的到達速率比例以及單一波長總容量對於容量公平性與網路能耗的影響，結果顯示EA-DPMDP在較高的到達速率比例下有著較差的容量公平性，而在較大的傳輸總容量條件下，EA-DPMDP依舊能維持良好的容量公平性。

Traffic grooming technique is seen as a solution for the issue that the bandwidth requirement of connection request does not match with total capacity of single wacelength channel. However, capacity fainess issue is still existed in traffic groomed optical network. One kind of the solutions for this issue is call admission control. Nowday, most of call admission control scheme are based on the following technique: static bandwidth reservation, static threshold setting, mathematical statics, and markov decision process formulation. Nevertheless, there is still a tradeoff between fairness and throughput. Therefore, a shceme called Dynamic Preemption Call Admission Based on Markov Decision Process (DP-MDP) which accepts or rejects the admission of connection requests based on the optimal policy decided by markov decision process based on the bandwidth utilization of each available wavelength in single link was proposed to improve throughput performance while maintain fariness.
As the growing conscious of energy conservation, it has become a trend for network hardware and telecom protocol to import energy efficiency consideration, especially in wavelength division multiplexing (WDM) network. The power-aware routing and wavelength assignment is able to reduce the energy consumption of infrastructure in WDM network. Energy-aware DP-MDP (EA-DPMDP) determines the admission of connection requests according to optimal policymade by markov decision process based on the maximum throughput per energy consumption, which maintains capacity fairness and the improvement of throughput while reduces energy consumption. A bidirectional mesh-torus network topology is first taken for simulations to compare EA-DPMDP and none energy-aware DP-MDP of network throughput, energy consumption, and capacity fairness. Simulation results show that EA-DPMDP reduces energy consumption while maintains capacity fairness. Then EA-DPMDP as well as different routing and wavelength assignments (RWAs), such as RSB, LUB, and TATG, are considered to show the effect on network throughput, energy consumption, and capacity fairness. The results show EA-DPMDP effectively reduces network energy consumption as well as maintains capacity fainess. Then EA-DPMDP with different RWAs is considered to show the effect on energy consumption and capacity fairness. The results show that EA-DPMDP with different RWAs which has longer route length, such as TATG, brings better energy reservation at different system load but worse capacity fairness at low system load. To confirm the effect of route length on EA-DPMDP, different network topologies are also considered by EA-DPMDP to evaluate impact on fairness. The results show the topology with shorter route lengths makes the decrement of blocking of high capacity requirement connection requests and brings worse capacity fainess at low system load. Finally, we evaluate EA-DPMDP with different arrival rate and total capacity of a wavelength in terms of network energy consumption and capacity fairness. The result present that EA-DPMDP with high arrival rate leads to bad capacity fainess while EA-DPMDP high total capacity of a wavelength results in good capacity fairness.

Contents VIII
List of Tables IX
List of Figures X
1. Introduction 1
2. Related Work 4
2.1. Network Model 4
2.2. Dynamic Preemption Call Admission Control Scheme Based on Markov
Decision Process 8
2.3. Energy-Aware Routing and Wavelength Assignment (RWA) 10
2.3.1 Integer Linear Programming 11
2.3.2 Time-Aware Traffic Grooming 11
2.3.3 Request Size Based and Link Utilization Based 13
2.4 Motivation 14
3. Energy-Aware DP-MDP (EA-DPMDP) 15
3.1 Markov Decision Process 15
3.2 Adjustment of Class-Based Weight Value 19
4. Simulation Results 23
4.1 DP-MDP with Energy-Aware Reward Function 26
4.2 Energy-Aware DP-MDP with Different RWA 30
4.3 Energy-Aware DP-MDP with Different Topology 36
5. Conclusion and Future Work 41
Reference: 42

[1]B. Mukherjee, C. Ou, H. Zhu, K. Zhu, N. Singhal, S. Yao, “Traffic grooming in mesh optical networks, in Proceeding of Optical Fiber Communication Conference, OFC 2004, pp. 23–27, 2004.
[2]M. Sivakumar, K. M. Sivalingam, S. Subramaniam, “On factors affecting the performance of dynamically groomed optical WDM mesh networks, in Proceeding of 2005 Workshop on High Performance Switching and Routing, HPSR 2005, pp. 411–415, 2005.
[3]K. Zhu, H. Zang, and B. Mukherjee, “A comprehensive study on next-generation optical grooming switches, IEEE Journal on Selected Areas in Communications, J-SAC, vol. 21, no. 7, pp. 1173–1186, 2003.
[4]A. K. Somani, “Survivability and traffic grooming in WDM optical networks, Cambridge University Press 2006.
[5]K. Zhu, H. Zhu, B. Mukherjee, “Traffic engineering in multigranularity heterogeneous optical WDM mesh networks through dynamic traffic grooming, IEEE Network, vol. 17, no. 2, pp. 8-15, Mar. 2003.
[6]K. W. Ross, D. H. K. Tsang, “The stochastic knapsack problem, IEEE Transactions on Communications, vol. 37, no. 7, pp. 740–747, 1989.
[7]P. Tran-Gia, F. Hubner, “An analysis of trunk reservation and grade of service balancing mechanisms in multiservice broadband network, in Proceedings of the IFIP TC6 Task Group/WG6.4 International Workshop on Performance of Communication Systems: Modelling and Performance Evaluation of ATM Technology, vol. c-15, pp. 83–97, 1993.
[8]S. Thiagarajan, A. K. Somani, “Capacity fairness of WDM networks with grooming capabilities, Optical Networks Magazine, vol. 2, no. 3, pp. 24–32, 2001.
[9]J. Choi, T. Kwon, Y. Choi, M. Naghshineh, “Call admission control for multimedia services in mobile cellular networks: a Markov decision approach, in Proceeding of 5th IEEE Symposium on Computers and Communications,ISCC 2000 , 2000, pp. 594–599.
[10]K. Mosharaf, J. Talim, I. Lambadaris, “A call admission control for service differentiation and fairness management in WDM grooming networks, in Proceeding of 1st International Conference on Broadband Networks, BroadNets 2004, pp. 162–169, 2004.
[11]K. Mosharaf, J. Talim, I. Lambadaris, “A call admission control for service differentiation and fairness management in WDM grooming networks, Optical Switching and Networking, vol. 2, no. 2, pp. 113–126, 2005.
[12]C.C. Sue, Y.B. Hsu, P.J. Ho, “Dynamic preemption call admission control scheme based on markov decision process in traffic groomed optical networks, IEEE/OSA Journal of Optical Communications and Networking, vol. 3, no. 4, pp. 300-311, Mar. 24, 2011.
[13]F. Idzikowski, E. Bonetto, L. Chiaraviglio, A. Cianfrani, A. Coiro, R. Duque, F. Jiménez, E. L. Rouzic, F. Musumeci, W. V. Heddeghem, J. L. Vizcaíno, Y. Ye, “TREND in energy-aware adaptive routing Solutions, IEEE Communicaiotn Maganize, vol. 51, no. 11, pp. 94-104, Nov. 2013.
[14]L. Chiaraviglio, M. Mellia, F. Neri, “Reducing power consumption in backbone networks, in Proceedings of the 2009 IEEE international conference on Communications, ICC 2009, pp. 1-6, 14-18 June 2009.
[15]Y. Wu, L. Chiaraviglio, M. Mellia, F. Neri, Power-Aware Routing and Wavelength Assignment in Optical Networks, in Proceedings of 35th European Conference on Optical Communication, ECOC 2009, pp. 1-2, Sep. 2009.
[16]E. Yetginer, G. N. Rouskas, “Power efficient traffic grooming in optical WDM networks,“ in Proceeding of 2009 IEEE Conference on Global Telecommunications, GLOBECOM 2009, pp. 1838-1843, 2009.
[17]S. Zhang, D. Shen, C. Chan, “Energy-efficient traffic grooming in WDM networks with scheduled time traffic, IEEE Journal of Lightwave Technology, vol. 29, no. 17, pp. 2577-2584, Sep. 2011.
[18]M. M. Hasan, F. Farahmand, J. P. Jue, J. P. Rodrigues, “A study of energy-aware traffic grooming in optical networks: static and dynamic cases, IEEE Systems Journal, vol. 7, no. 1, pp. 161-169, Mar. 2013.
[19]P. Monti, P. Wiatr, A. Jirattigalachote, L. Wosinska, “Trading power savings for blocking probability in dynamically provisioned WDM networks, in Proceeding of 12th International Conference on Transparent Optical Networks, ICTON 2010, pp. 1, 2010
[20]P. Wiatr, P. Monti, L. Wosinska, “Power savings versus network performance in dynamically provisioned WDM networks, IEEE Communications Magazine, pp. 48-55, May 2012.
[21]K. Zhu and B. Mukherjee, “A review of traffic grooming in WDM optical networks: architectures and challenges, Optical Networks Magazine, vol. 4, no. 2, pp. 55–64, 2003.
[22]Juniper Networks, T series core routers: T320, T640, T1600, TX Matrix and TX Matrix Plus. [Online]. Available : http://www.juniper.net/us/en/local/pdf/datasheets/1000051-en.pdf, May 2012
[23]Calient Networks, Diamond Wave PXC. [Online]. Available: http://www.calient.com/_docs/pxc_datasheet.pdf, accessed at Mar. 2013.
[24]J. Chabarek, J. Sommer, P. Barford, C. Estan, D. Tsiang, S. Wright, “Power Awareness in Network Design and Routing, in Proceeding of 27th International Conference on Computer Communications, INFOCOM 2008, pp.457-465, Apr. 2008.
[25]R. A. Howard, Dynamic Programming and Markov Process. MIT Press, 1960.
[26]M. L. Puterman, Markov Decision Processes.Wiley Inter-Science 1994.
[27]Jin Y. Yen, “Finding the k shortest path loopless paths in a network, Management Science, vol. 17, no. 11, pp. 712-716, July 1971.
[28]R. Albert, A. L. Barabási, “Statistical mechanics of complex networks, Reviews of Morden Physics, vol. 74, pp. 47-97, Jan. 2002