本論文針對高速通訊網路下非對稱的頻寬和資源問題，提出了一個在有限資源下的非對稱冗餘消除演算法(RCARE) ，利用多餘的下載頻寬和接收端的資源，以加速上傳的數據傳輸。該系統可以部署於客戶端或代理伺服器上。RCARE 與現有的非對稱演算法不同，它使用更加靈活的匹配機制來識別冗餘資料，並使用一個傳送端的暫存器吸收過高的下載流量。和現有的冗餘消除演算法相比，它提供了一個可根據資源與效能調整的傳送端暫存器。我們從多個伺服器和校園網路記錄了真實的流量資料，並利用這些資料評估RCARE 的效能。由我們的模擬結果顯示， RCARE 可比目前的非對稱式通訊演算法達到更高的上傳增益，以及更低的下載流量。我們也為有限資源的傳送端設計了動態調整演算法。此演算法可根據目前的樣本資料，預測並分配資源給目前的數據流，以達到最大的上傳增益。與平均分配資源的基準演算法相比較，動態調整演算法提高了高達87％ 的上傳增益。在前 10% 的實驗結果中( 以最佳的上傳增益排序) ， RCARE 平均達到了高達 40.5% 的上傳增益。

We focus on the problem of efficient communications over access networks with asymmetric bandwidth and capability. We propose a resource-constrained asymmetric redundancy elimination algorithm(RCARE) to leverage downlink bandwidth and receiver capability to accelerate the uplink data transfer. RCARE can be deployed on a client or a proxy. Different from existing asymmetric algorithms, RCARE uses flexible matching mechanism to identify redundant data, and allocates a small sender cache to absorb the high downlink traffic overhead. Compared to redundancy elimination algorithms, RCARE provides a scalable sender cache which is adaptive based on resource and performance. We evaluate RCARE with real traffic traces collected from multiple servers and a campus gateway. The trace-driven simulation results indicate that RCARE achieves higher goodput gains and reduces downlink traffic compared to existing asymmetric communication algorithms. We design an adaptation algorithm for resource-constrained senders sending multiple data streams. Our algorithm takes samples from data streams and predicts how to invest cache size on individual data streams to achieve maximal uplink goodput gain. The adaptation algorithm improves the goodput gain by up to 87% compared to the baseline. In first 10% of data streams (sorted by the optimal goodput gains), RCARE achieves up to 40.5% goodput gain on average.

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 Asymmetric Communication Algorithms . . . . . . . . . . . . . . . . . . . . . 3
1.2.2 Redundancy Elimination Algorithms . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Limitations of Current Asymmetric Communication Algorithms . . . . . 7
1.3.1 Potential of Asymmetric Communication Algorithms . . . . . . . . . . . 7
1.3.2 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 RCARE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1 A new Asymmetric Communication Algorithm: RCARE . . . . . . . . . . . 10
2.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.2 Packet Caches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1.3 Efficient Partial-Match Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2 Trace-Driven Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.1 Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.3 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3 Dynamic Adaptation Algorithm and Large-Scale Simulations . . . . . . . . 19
3.1 Data Stream Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.1 Real-Life Traffic Trace Collection . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.2 Data Stream Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.3 Simulation-Based Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2 Dynamic Adaptation Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.1 Prediction Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.2 Adaptation Formulation and Algorithm . . . . . . . . . . . . . . . . . . . . . 24
3.3 Large-Scale Simulations and Evaluations . . . . . . . . . . . . . . . . . . . . . 27
3.3.1 Evaluations on Adaptation Algorithm . . . . . . . . . . . . . . . . . . . . . . 28
3.3.2 Performance Gain of RCARE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4 Conclusion and FutureWork. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Symbol Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

[1] WAN and application optimization solution guide. http://www.cisco.com/en/US/docs/nsite/wan_optimization/WANoptSolutionGd.pdf, April 2008.
[2] G. Aceto, A. Dainotti, W. De Donato, and A. Pescape. Portload: Taking the best of two worlds in traffic classification. In Proc. of IEEE INFOCOM’10, pages 1 –5, march 2010.
[3] M. Adler and B.Maggs. Protocols for asymmetric communication channels. Journal of Computer and System Sciences, 63(4):573–596, December 2001.
[4] B. Aggarwal, A. Akella, A. Anand, A. Balachandran, P. Chitnis, C. Muthukrishnan, R. Ramjee, and G. Varghese. EndRE: An end-system redundancy elimination service for enterprises. In Proc. of USENIX Conference on Networked Systems Design and Implementation (NSDI’10), pages 419–432, San Jose, CA, June 2010.
[5] A. Anand, A. Gupta, A. Akella, S. Seshan, and S. Shenker. Packet caches on routers: The implications of universal redundant traffic elimination. In Proc. of ACM SIGCOMM’08, pages 219–230, Seattle, WA, August 2008.
[6] A. Anand, C. Muthukrishnan, A. Akella, and R. Ramjee. Redundancy in network traffic: Findings and implications. In Proc. of Joint International Conference on Measurement and Modeling of Computer Systems (SIGMETRICS/Performance’09), pages 37–48, Seattle, WA, June 2009.
[7] B. Aslam, P.Wang, and C. Zou. Pervasive Internet access by vehicles through satellite receive-only terminals. In Proc. of IEEE International Conference on Computer Communications and Networks (ICCCN’09), San Francisco, CA, August 2009.
[8] L. Atzori, A. Iera, and G. Morabito. The Internet of things: A survey. Computer Networks, 54(15):2787–2805, October 2010.
[9] H. Balakrishnan and V. Padmanabhan. How network asymmetry affects TCP. IEEE Communications Magazine, 39(4):60–67, April 2001.
[10] P. Bose, D. Krizanc, S. Langerman, and P. Morin. Asymmetric communication protocols via hotlink assignments. Theory of Computing Systems, 36(6):655–661, November 2003.
[11] IBM ILOG CPLEX Optimizer. http://www-01.ibm.com/software/integration/optimization/cplex-optimizer/.
[12] M. Dischinger, A. Haeberlen, K. Gummadi, and S. Saroiu. Characterizing residential broadband networks. In Proc. of ACM SIGCOMM Internet Measurement Workshop (IMC’07), pages 43–56, San Diego, CA, October 2007.
[13] T. Gagie. Dynamic asymmetric communication. In Proc. of International Colloquium on Structural Information and Communication Complexity (SIROCCO’06), pages 310–318, Chester, UK, July 2006.
[14] T. Gagie. Dynamic asymmetric communication. Information Processing Letters, 108(6):352–355, November 2008.
[15] E. Halepovic, C. Williamson, and M. Ghaderi. DYNABYTE: A dynamic sampling algorithm for redundant content detection. In Proc. of IEEE International Conference on Computer Communications and Networks (ICCCN’11), pages 1–8, Maui, HI, July 2011.
[16] M. Hefeeda, C. Hsu, and K.Mokhtarian. Design and evaluation of a proxy cache for peer-to-peer traffic. IEEE Transactions on Computers, 60(7):964–977, July 2011.
[17] B. Jenkins. Algorithm alley: Hash functions. Dr. Dobb’s Journal, September 1997.
[18] C. Lee and R. Yang. High-throughput data compressor designs using content addressable memory. IEE Proceedings on Circuits, Devices and Systems, 142(1):69–73, February 1995.
[19] Y. Li, C. Trang, X. Huang, C. Hsu, and P. Lin. CacheQuery: A practical asymmetric communication algorithm. In Proc. of IEEE Global Telecommunications Conference(GLOBECOM’12), Anaheim, CA, December 2012.
[20] Y.-D. Lin, I.-W. Chen, P.-C. Lin, C.-S. Chen, and C.-H. Hsu. On campus beta site: architecture designs, operational experience, and top product defects. IEEE Communications Magazine, 48(12):83 –91, December 2010.
[21] M. Mathur. Elucidation of upcoming traffic problems in cloud computing. Recent Trends in Networks and Communications: Communications in Computer and Information Science, 90(1):68–71, July 2010.
[22] G. Mazzini. Asymmetric channel cooperative compression. IEEE Communications Letters, 12(4):328–330, April 2008.
[23] I. Minei and R. Cohen. High-speed Internet access through unidirectional geostationary satellite channels. IEEE Journal on Selected Areas in Communications, 17(2):345–359, January 2004.
[24] D. Munteanu and C. Williamson. An FPGA-based network processor for IP packet compression. In Proc. of SCS International Symposium on Performance Evaluation of Computer and Telecommunication Systems (SPECTS’05), pages 599–608, Philadelphia, PA, July 2005.
[25] The network simulator. http://www.isi.edu/nsnam/ns/.
[26] B.-C. Park, Y.Won, M.-S. Kim, and J. Hong. Towards automated application signature generation for traffic identification. In Proc. of Network Operations and Management Symposium (NOMS’08), pages 160 –167, april 2008.
[27] N. Spring and D. Wetherall. A protocol-independent technique for eliminating redundant network traffic. In Proc. of ACM SIGCOMM’00, pages 87–95, Stockholm, Sweden, August 2000.
[28] C. M. Trang, X. Huang, and C.-H. Hsu. Pushing uplink goodput of an asymmetric access network beyond its uplink bandwidth. In Proc. of IEEE International Conference on Communications (ICC’12), Ottawa, Canada, June 2012.
[29] C. Tye and D. Fairhurst. A review of IP packet compression techniques. In Proc. of Annual PostGraduate Symposium on the Convergence of Telecommunications, Networking and Broadcasting (PGNet’03), Liverpool, UK, June 2003.
[30] Verizon DSL Internet Plans. http://www22.verizon.com/home/highspeedinternet/#plans.
[31] R.Wang, T. Taleb, A. Jamalipour, and B. Sun. Protocols for reliable data transport in space Internet. IEEE Communications Surveys and Tutorials, 11(2):21–32, Second Quarter 2009.
[32] J. Watkinson, M. Adler, and F. Fich. New protocols for asymmetric communication channels. In Proc. of International Colloquium on Structural Information and Communication Complexity (SIROCCO’01), pages 337–350, Catalonia, Spain, June 2001.
[33] R. Yang and C. Lee. High-throughput data compressor designs using content addressable memory. In Proc. of IEEE International Symposium on Circuits and Systems (ISCAS’94), pages 147–150, London, UK, May 1994.
[34] zlib home page, April 2010. http://zlib.net/.
[35] E. Zohar, I. Cidon, and O. Mokryn. The power of prediction: Cloud bandwidth and cost reduction. In Proc. of ACM SIGCOMM’11, pages 86–97, Toronto, Canada, August 2011.
[36] M. Zorzi, A. Gluhak, S. Lange, and A. Bassi. From today’s INTRAnet of things to a future INTERnet of things: A wireless- and mobility-related view. IEEE Wireless Communication Magazine, 17(6):44–51, December 2010.