為了增加軟體定義網絡（SDN）的可擴展性，過去已經提出了流聚合方案以將多個小型流合併到用於流量工程的大型聚合流中。在本論文中，我們首先注意到，在聚合流中不再保證小型流的用戶速率要求，因為由TCP 公平分配決定的流速通常不同於每個用戶的期望速率。為了解決上述問題，我們提出了一種新的架構，稱為Flexible Flow And Rate Management（F2ARM），以只有控制幾個流的速率，以增加SDN的可擴展性，同時保留由TCP管理不受控制的流。我們制定了一個新的優化問題，稱為可擴展的流量速率控制SDN（SPFRCS），其目的是找到流的最小子集作為受控流，但確保所有未控制的流的流速仍然可以滿足最小所需速率根據TCP公平分配。我們證明SPFRCS是NP-hard並且設計了一種有效的算法，稱為Joint Flow Selection and Rate Determination（JFSRD）。基於真實網絡的仿真結果表明，JFSRD在小規模網絡中執行幾乎最佳，並且受控流的數量可以在真實網絡中有效地減少50％。

To increase the scalability of software defined networks (SDNs), flow aggregation schemes have been proposed to merge multiple mice flows into an elephant aggregated flow for traffic engineering. In this thesis, we first notice that the user bit-rate requirements of mice flows are no longer guaranteed in the aggregated flow since the flow rate decided by TCP fair allocation is usually different from the desired bit-rate of each user. To address the above issue, we present a novel architecture, named Flexible Flow And Rate Management(F2ARM), to control the rates of only a few flows in order to increase the scalability of SDN,while leaving the uncontrolled flows managed by TCP. We formulate a new optimization problem, named Scalable Per-Flow Rate Control for SDN (SPFRCS), which aims to find a minimum subset of flows as controlled flows but ensure that the flow rates of all uncontrolled flows can still satisfy minimum required rates by TCP fair allocation. We prove that SPFRCS is NP-hard and design an efficient algorithm, named Joint Flow Selection and Rate Determination (JFSRD). Simulation results based on real networks manifest that JFSRD performs nearly optimally in small-scale networks, and the number of controlled flows can be effectively reduced by 50% in real networks.

[1] N. McKeown, T. Anderson, H. Balakrishnan, G. Parulkar, L. Peterson, J. Rexford, S.Shenker, and J. Turner, “OpenFlow: enabling innovation in campus networks,"SIG-COMM Computer Communication Review, vol. 38, no. 2, pp. 69–74, Mar. 2008.
[2] A. R. Curtis, J. C. Mogul, J. Tourrilhes, P. Yalagandula, P. Sharma, and S. Banerjee, “DevoFlow: scaling flow management for high-performance networks," inProc. ACMSIGCOMM, 2011.
[3] C.-Y. Hong, S. Kandula, R. Mahajan, M. Zhang, V. Gill, M. Nanduri, and R. Wattc-nhofer, “Achieving high utilization with software-driven wan," inProc. ACM SIG-COMM, 2013.
[4] “Openflowswitchspecification",[Online].Available:https://www.opennetworking.org/technical-communities/areas/specification.
[5] Y. Lu and S. Zhu, “Sdn-based tcp congestion control in data center networks," inProc.IEEE IPCCC, 2015.
[6] M. S. Seddiki, M. Shahbaz, S. Donovan, S. Grover, M. Park, N. Feamster, and Y.-Q.Song, “FlowQoS: qos for the rest of us," inProc. ACM HotSDN, 2014.
[7] Y. Li, S. H.-C. Jiang, H. Tan, C. Zhang, G. Chen, J. Zhou, and F. C.M. Lau, “Efficientonline coflow routing and scheduling," in Proc. ACM MobiHoc, 2016.
[8] S. Paris, A. Destounis, L. Maggi, G. S. Paschos, and J. Leguay, “Controlling flowreconfigurations in sdn," inProc. IEEE INFOCOM, 2016.
[9] F. Chen, C. Wu, X. Hong, Z. Lu, Z. Wang, and C. Lin, “Engineering traffic uncertaintyin the openflow data plane," inProc. IEEE INFOCOM, 2016.
[10] L.-H. Huang, H.-C. Hsu, S.-H. Shen, D.-N. Yang, and W.-T. Chen, “Multicast trafficengineering for software-defined networks," inProc. IEEE INFOCOM, 2016.
[11] Y. Liu, Y. Li, M. Canini, Y. Wang, and J. Yuan, “Scheduling multi-flow networkupdates in software-defined nfv systems," inProc. IEEE INFOCOM Workshop, 2016.
[12] H. Huang, S. Guo, P. Li, B. Ye and I. Stojmenovic, “Joint optimization of rule place-ment and traffic engineering for qos provisioning in software defined network,"IEEETransactions on Computers, vol. 64, no. 12, pp. 3488–3499, Dec. 2015.
[13] C. R. Meiners, A. X. Liu, and E. Torng, “Bit weaving: a non-prefix approach tocompressing packet classifiers in tcams,"IEEE/ACM Transactions on Networking, vol.20, no. 2, pp 488–500, Apr. 2012.
[14] S. Luo, H. Yu, and L. M. Li, “Fast incremental flow table aggregation in SDN," inProc. IEEE ICCCN, 2014.
[15] X.-N. Nguyen, D. Saucez, C. Barakat, and T. Turletti, “Officer: a general optimizationframework for openflow rule allocation and endpoint policy enforcement,“ inProc.IEEE INFOCOM, 2015.
[16] R. Bhatia, F. Hao, M. Kodialam, and T.V. Lakshman, “Optimized network traffic en-gineering using segment routing," inProc. IEEE INFOCOM, 2015.
[17] F. Hao, M. Kodialam, and T.V. Lakshman, “Optimizing restoration with segment rout-ing," inProc. IEEE INFOCOM, 2016.
[18] V. Jacobson, “Berkeley tcp evolution from 4.3-tahoe to 4.3-reno," inProc. IETF, 1990.
[19] S. Floyd and T. Henderson, “The new reno modification to tcp’s fast recovery algo-rithm," RFC 2582, 1999.
[20] L. Xu, K. Harfoush, and I. Rhee, “Binary increase congestion control (bic) for fastlong-distance networks," inProc. IEEE INFOCOM, 2004.
[21] S. Ha, I. Rhee, and L. Xu, “Cubic: a new tcp-friendly high-speed tcp variant,"ACMSIGOPS Operating System Review, vol. 42, no. 5, pp. 64–74, Jul. 2008.
[22] S. Liu, T. Basar, and R. Srikant “Tcp-illinois: a loss and delay-based congestion con-trol algorithm for high-speed networks,"Elsevier Performance Evaluation, vol. 65, no.6–7, pp. 417–440, 2008.
[23] F. Baccelli and D. Hong, “AIMD, fairness and fractal scaling of tcp traffic," inProc.IEEE INFOCOM, 2002.
[24] S. Agarwal, M.T Kodialam, and T. V. Lakshman, “Traffic engineering in softwaredefined networks," inProc. IEEE INFOCOM, 2013.
[25] Y. Guo, Z. Wang, X. Yin, X. Shi, and J. Wu, “Traffic engineering in sdn/ospf hybridnetwork," inProc. IEEE ICNP, 2014.
[26] Y. Hu, W. Wang, G. Xianhyang, X. Que, Y. Ma, and S. Cheng, “Maximizing networkutilization in hybrid software-defined networks," inProc. IEEE GLOBECOM, 2015.
[27] I. Dinur and D. Steurer, “Analytical approach to parallel repetition," inProc. ACMSTOC, 2014.
[28] “The internet topology zoo," 2014. [Online]. Available:http://www.topology-zoo.org/dataset.html
[29] M. Rifai, N. Huin, C. Caillouet, F. Giroire, D. Lopez-Pacheco, J. Moulierac and G.Urvoy-Keller, “Too many sdn rules? compress them with minnie," inProc. IEEEGLOBECOM, 2016.
[30] “IBM ILOG CPLEX," 2014. [Online]. Available:http://www-01.ibm.com/software/commerce/optimization/cplex-optimizer/
[31] Z. A. Qazi, C.-C. Tu, L. Chiang, R. Miao, V. Sekar, and M. Yu., “SIMPLE-fyingMiddlebox Policy Enforcement Using SDN," inProc. ACM SIGCOMM, 2013.
[32] Harold W. Kuhn, “The Hungarian Method for the assignment problem",Naval Re-search Logistics Quarterly, vol. 2, pp. 83-97, 1955.