Abstract

Accurate estimation of end-to-end available bandwidth is essential for network-aware applications including adaptive streaming, congestion control, and traffic engineering. Existing active probing techniques either assume fluid cross-traffic models, require cooperative software at both endpoints, or exhibit high estimation error when cross-traffic conditions vary. This paper presents PathAB, a hybrid active probing method that estimates available bandwidth in two phases: a single exponentially spaced probing train provides a rapid coarse estimate, followed by multiple Poisson-distributed probing trains that refine the estimate using a linear regression model. PathAB supports both client–server and standalone modes. In standalone mode, small 28-byte echo packets are placed behind larger probe packets to minimize reverse-path interference, eliminating the need for receiver-side software. The algorithm is evaluated through NS-2 simulations at four bottleneck capacities (1.5, 5, 10, and 15 Mbps) under four utilization levels, and through network test-bed experiments in single-hop, multi-hop, and 100 Mbps configurations. PathAB is compared against PathChirp, PoissonProb, IGI, Spruce, Pathload, and a stochastic model. Results show that PathAB achieves RMS estimation error below 10% under moderate utilization in both modes, and maintains stable accuracy under high-utilization conditions where competing methods degrade substantially.

References

• R. L. Carter and M. E. Crovella, “Dynamic server selection using bandwidth probing in wide-area networks,” Boston Univ., Comput. Sci. Dept., Tech. Rep. TR-96-007, 1996.
• J. Strauss, D. Katabi, and F. Kaashoek, “A measurement study of available bandwidth estimation tools,” in Proc. 3rd ACM SIGCOMM Conf. Internet Measurement, 2003, pp. 39–44.
• M. Jain and C. Dovrolis, “Pathload: a measurement tool for end-to-end available bandwidth,” in Proc. Passive and Active Measurements (PAM) Workshop, 2002.
• D. Kiwior, J. Kingston, and S. Akhtar, “PATHMON, a methodology for determining available bandwidth over an unknown network,” in IEEE/Sarnoff Symp. Advances inWired and Wireless Communication, 2004, pp. 27–30.
• J. Cao, W. S. Cleveland, D. Lin, and D. X. Sun, “Internet traffic tends toward Poisson and independent as the load increases,” in Nonlinear Estimation and Classification. New York, NY, USA: Springer, 2002, pp. 83–109.
• M. Zhang, C. Luo, and J. Li, “Estimating available bandwidth using multiple overloading streams,” in IEEE Int. Conf. Communications (ICC), vol. 2, Istanbul, Turkey, 2006, pp. 495–502.
• L. Xin, “PoissonProb: a new rate-based available bandwidth measurement algorithm,” M.S. thesis, School of Comput. Sci., Univ. of Windsor, Windsor, ON, Canada, 2005.
• V. J. Ribeiro, R. H. Riedi, R. G. Baraniuk, J. Navratil, and L. Cottrel, “pathChirp: efficient available bandwidth estimation for network paths,” in Proc. Passive and Active Measurement Workshop, 2003.
• N. Hu and P. Steenkiste, “Evaluation and characterization of available bandwidth probing techniques,” IEEE J. Sel. Areas Commun., vol. 21, no. 6, pp. 879–894, 2003.
• A. Botta, S. D’Antonio, A. Pescap´e, and G. Ventre, “BET: a hybrid bandwidth estimation tool,” in Proc. 11th Int. Conf. Parallel and Distributed Systems (ICPADS), vol. 2, 2005, pp. 520–524.
• S. R. Kang, X. Liu, M. Dai, and D. Loguinov, “Packet-pair bandwidth estimation: stochastic analysis of a single congested node,” in Proc. 12th IEEE Int. Conf. Network Protocols (ICNP), 2004, pp. 316–325.
• E. Hemmati and S. Bhattacharyya, “Challenges of available bandwidth estimation in high-speed networks,” in Proc. IEEE Int. Conf. Communications (ICC), 2017, pp. 1–6.
• Y. Zhu et al., “Packet-level telemetry in large datacenter networks,” in Proc. ACM SIGCOMM, 2015, pp. 479–491.
• Q. Yin, J. Kaur, and F. D. Smith, “Can machine learning benefit bandwidth estimation at ultra-high speeds?,” in Proc. Passive and Active Measurement Conf. (PAM), 2020, pp. 49–64.
• C. Grigorescu, L. Lorenzi, and N. Cascarano, “Bandwidth prediction using machine learning for SDN-based networks,” in Proc. IEEE/IFIP Network Operations and Management Symp. (NOMS), 2020, pp. 1–5.
• D. Roy, “PathAB: a new method to estimate end-to-end available bandwidth of network path,” M.S. thesis, School of Comput. Sci., Univ. ofWindsor,Windsor, ON, Canada, 2009.
• A. Pasztor and D. Veitch, “The packet size dependence of packet pair like methods,” in Proc. 10th IEEE Int. Workshop Quality of Service, 2002, pp. 204–213.