Stochastic Modeling of the TCP Protocol

PhD Candidacy

Exam Paper List

Eli Brosh

1. TCP Protocol Specifications

Many TCP variants exist and are distinguished by the particular way the sources react to packet loss events. I will focus on the most-widely used variants, TCP-Tahoe [1], TCP-Reno [2], TCP-NewReno[3], TCP-SACK[4], and TCP-Vegas [5].

2. Renewal Theory Models

Renewal theory based modeling has received a lot of attention in the literature. This approach assumes a single source model and aims to compute TCP’s throughput and latency as a function of the network characteristics. I will give a description of the work (listed in chronology order) in: [6, 7, 8, 9, 10, 11].

3. Fixed Point Methods

In this approach the interaction between separate TCP sources is studied in order to find the operating regime of a network. I will discuss the following papers: [12, 13, 14, 15].

4. Fluid Models

In this framework TCP packets are approximated as a fluid flowing through the network. A typical setup includes a system of stochastic or ordinary differential equations which describe the changes in TCP’s transmission rate. I will cover the following papers: [16, 17, 18, 19]

5. Processor Sharing Models

The processor sharing framework has been traditionally used to model the interaction between flows sharing a server. In this model, performance metrics (e.g., download times and number of concurrent connections) are studied on a flow level, and the complexities of the TCP protocol are hidden. I will cover the following papers: [20, 21]
 

6. Control Theoretic Models

Methods from the mature field of control theory have been successfully applied to TCP-network modeling and the design of flow-control mechanisms. I will briefly cover this area by the following papers: [22, 23].

7. Experimental Enhancement and Future Trends

I will discuss recent suggestions to enhance or replace TCP, such as [24, 25]; and will present a methodology to infer and track the key parameters of TCP in large scale networks such as the internet [26, 27]. Multimedia congestion control, particularly the idea of TCP-friendliness, represents one of the current trends in Internet congestion control research. I will discuss some representative results based on [28, 29].

 

References

[1] V. Jacobson, “Congestion avoidance and control,” in ACM SIGCOMM ’88, Stanford, CA, Aug. 1988, pp. 314–329.

[2] M. Allman, V. Paxson, and W. Stevens, “TCP Congestion Control.” RFC 2581, Apr. 1999.

[3] S. Floyd and T. Henderson, “The NewReno Modification to TCP’s Fast Recovery Algorithm,” RFC 2582, Apr. 1999.

[4] M. M. Floyd S., Mahdavi J. and P. M., “An Extension to the Selective Acknowledgement (SACK) Option for TCP,” RFC 2883, July 2000.

[5] L. S. Brakmo and L. L. Peterson, “TCP Vegas: End to end congestion avoidance on a global internet,” IEEE Journal on Selected Areas in Communications, vol. 13, no. 8, pp. 1465–1480, 1995.

[6] M. Mathis, J. Semke, and J. Mahdavi, “The macroscopic behavior of the TCP congestion avoidance algorithm,” Computer Communications Review, vol. 27, no. 3, 1997.

[7] J. Padhye, V. Firoiu, D. Towsley, and J. Kurose, “Modeling TCP throughput: A simple model and its empirical validation,” in ACM SIGCOMM, 1998, pp. 303–314.

[8] A. Kumar, “Comparative performance analysis of versions of TCP in a local network with a lossy link,” IEEE/ACM Transactions on Networking, vol. 6, no. 4, pp. 485–498, 1998.

[9] N. Cardwell, S. Savage, and T. Anderson, “Modeling TCP latency,” in IEEE INFOCOM, Tel-Aviv, Israel, 2000.

[10] B. Sikdar, S. Kalyanaraman, and K. S. Vastola, “An integrated model for the latency and steady-state throughput of TCP connections,” Performance Evaluation, vol. 46, no. 2-3, pp. 139–154, 2001.

[11] C. B. Samios and M. K. Vernon, “Modeling the throughput of TCP Vegas,” in ACM SIGMETRICS, San Diego, CA, USA, 2003.
  

[12] A. Misra and T. J. Ott, “The window distribution of multiple TCPs with random loss queues,” in IEEE GLOBECOM, 1999.

[13] V. Firoiu and M. Borden, “A study of active queue management for congestion control,” in IEEE INFOCOM, Tel-Aviv, Israel, 2000.

[14] T. Bu and D. Towsley, “Fixed point approximations for TCP behavior in an AQM network,” in ACM SIGMETRICS, Boston, MA, 2001.

[15] C. Casetti and M. Meo, “A new approach to model the stationary behavior of TCP connections,” in IEEE INFOCOM, Tel-Aviv, Israel, 2000.

[16] V. Misra, W. Gong, and D. Towsley, “Stochastic differential equation modeling and analysis of TCP-Windowsize behavior,” in Proceeding of PERFORMANCE, Istanbul, Turkey, 1999.

[17] V. Misra, W.-B. Gong, and D. F. Towsley, “Fluid-based analysis of a network of AQM routers supporting TCP flows with an application to RED,” in ACM SIGCOMM, 2000, pp. 151–160.

[18] E. Altman, K. Avrachenkov, and C. Barakat, “A stochastic model of TCP/IP with stationary random losses,” in ACM SIGCOMM, 2000.

[19] B. T. Eitan Altman, D. Barman and M. Vojnovic, “Parallel TCP sockets: Simple model, throughput and validation,” in IEEE INFOCOM, Barcelona, Spain, 2006.

[20] A. P. G. R. S. Ben Fred, T. Bonald and J. W. Roberts, “Statistical bandwidth sharing: a study of congestion at flow level,” in ACM SIGCOMM, San Diego, California, US, August 2001.

[21] R. K. J Beckers, I Hendrawan and R. van der Mei, “Generalized processor sharing performance models for internet access lines,” in 9th IFIP Conference on Performance Modelling and Evaluation of ATM and IP Networks, Budapest, 2001.

[22] D. M. Chiu and R. Jain, “Analysis of the increase and decrease algorithms for congestion avoidance in computer networks,” in Journal of Computer Networks and ISDN, vol. 17, 1989.

[23] C. V. Hollot, V. Misra, D. F. Towsley, and W. Gong, “A control theoretic analysis of RED,” in IEEE INFOCOM, 2001, pp. 1510–1519.

[24] A. Aggarwal, S. Savage, and T. Anderson, “Understanding the performance of TCP pacing,” in IEEE INFOCOM, Tel-Aviv, Israel, 2000.

[25] Thomas Anderson, Andrew Collins, Arvind Krishnamurthy, and John Zahorjan, “PCP: Efficient Endpoint Congestion Control,” in NSDI, San Jose, CA, USA, 2006.

[26] S. Jaiswal, G. Iannaccone, C. Diot, J. Kurose, and D. Towsley, "Measurement and Classification of Out-of-Sequence Packets in a Tier-1 IP backbone",  in IEEE INFOCOM, San Francisco, USA, 2003.

 

[27] S. Jaiswal, G. Iannaccone, C. Diot, J. Kurose, and D. Towsley, “Inferring TCP connection characteristics through passive measurements,” in IEEE INFOCOM, Hong Kong, 2004.

 

[28] S. Floyd, M. Handley, J. Padhye, and J. Widmer, “Equation-based congestion control for Unicast applications,” Stockholm, Sweden, September 2000.

[29] T Bu, Y Liu, D Towsley, “On the TCP-Friendliness of VoIP Traffic,” in IEEE INFOCOM, Barcelona, Spain, 2006.