Tools for Bandwidth Estimation

Tools for Bandwidth Estimation (for available and bottleneck bandwidth)

• Yaz:
  Yaz measures end-to-end available bandwidth. 2007.
  Joel Sommers, Paul Barford, Walter Willinger, Laboratory-based Calibration of Available Bandwidth Estimation Tools, Microprocessors & Microsystems, V.31 N.4, 2007.
  "We designed a calibrated available bandwidth estimation tool called Yaz that builds on the insights of Pathload."

• IGI:
  "This is an available bandwith measurement tool using active probing... For packet pair probing to get accurate network measurement, we need carefully tune the probing parameters."
  Nangning Hu and Peter Steenkiste, Evaluation and Characterization of Available Bandwidth Probing Techniques. IEEE JSAC Special Issue in Internet and WWW Measurement, Mapping, and Modeling, August, 2003.

• netest:
  LBNL's Network Test netest measures physical and available bandwidth. From 2003. (Netest might be a successor to pipechar, I can't quite tell...)

• pathChirp is an active tool for estimating the available bandwidth. From 2003.
  STAB, a successor to pathChirp, locates low-bandwidth links on a network path. From 2003.

• Pathload measures end-to-end available bandwidth. Last release in 2003.

• Pathrate measures end-to-end capacity (aka bottleneck bandwidth). Last release in 2003.

• pchar (based on pathchar) characterize the bandwidth, latency, and loss of links along an end-to-end path. From 1999-2001.
  Brian Tierney's TCP tuning guide for distributed application on wide area networks has a comparison of pchar and pipechar.

• Spruce:
  Jacob Strauss, Dina Katabi, Frans Kaashoek, Balaji Prabhakar, Spruce: A Lightweight End-to-End Tool for Measuring Available Bandwidth.
  Jacob also has a paper comparing Spruce, Pathload, and IGI.

• Dina Katabi and Charles Blake, Inferring Congestion Sharing and Path Characteristics from Packet Interarrival Times, LCS Technical Report, 2001.
  "This paper presents new non-intrusive measurement techniques to ... discover bottleneck router link speeds."

• 2003 Bandwidth Estimation Workshop.

• R. S. Prasad, M. Murray, C. Dovrolis, K. Claffy, Bandwidth Estimation: Metrics, Measurement Techniques, and Tools, IEEE Network, November/December 2003.
  "Existing bandwidth estimation tools measure one or more of three related metrics: capacity, available bandwidth, and bulk transfer capacity (BTC)."

Tools for Estimating Loss Rates

• badabing:
  J. Sommers, P. Barford, N. Duffield, and A. Ron, Improving Accuracy in End-to-end Packet Loss Measurement, SIGCOMM, 2005. A later version of this paper appeared in IEEE/ACM Transactions on Networking, 2008.
  "We developed a one-way active measurement tool called BADABING. BADABING sends fixed-size probes at specified intervals from one measurement host to a collaborating target host."

• sting:
  Stefan Savage's sting is "a TCP-based network measurement tool that measures end-to-end network path characteristics. sting ... can estimate one-way properties, such as loss rate." From 1999.

• Mark Allman, Wesley Eddy, and Shawn Ostermann, Estimating Loss Rates With TCP, May 2003.
  "Estimating loss rates along a network path is a problem that has received much attention within the research community. However, deriving accurate estimates of the loss rate from TCP transfers has been largely unaddressed."

• T-RAT:
  "We attempt to determine the cause of the rates at which flows transmit data by developing a tool, T-RAT, to analyze packet-level TCP dynamics."

• tulip:
  Ratul Mahajan, Neil Spring, David Wetherall, and Thomas Anderson.
  "Tulip is a tool to localize the occurrence of packet reordering, loss, and queueing delay along the path from the source to an arbitrary destination." Tulip uses ICMP timestamps and sequential IP identifiers.

• L. Qiu and V. N. Padmanabhan, Server-Centric View of Internet Performance: Analysis and Implications, Microsoft Research Technical Report MSR-TR-2001-78, September 2001.
  Given packet traces taken at the web server, this paper measures the packet drop rates to the clients. Given in addition the network topology (e.g., from traceroutes), the paper presents an algorithm for estimating the packet loss rates of the links on the tree.

• P. Benko, A. Veres, A Passive Method for Estimating End-to-End TCP Packet Loss, Proc. IEEE Globecom 2002, Taiwan.
  "Based on the seen sequence number pattern the loss ratios are estimated for the two segments of end-to-end path divided by the monitor." The tool considers timeouts, clustered losses, multipathing, and reordering.

Tools for Estimating Delay

• king:
  Krishna P. Gummadi, Stefan Saroiu, Steven D. Gribble, King: Estimating Latency between Arbitrary Internet End Hosts, SIGCOMM Internet Measurement Workshop, 2002.
  "We present King, a tool that accurately and quickly estimates the latency between arbitrary end hosts by using recursive DNS queries in a novel way."

• cing:
  Anagnostakis, Greenwald, and Ryger, Measuring Network-internal Delays using only Existing Infrastructure, Infocom 2003.
  The cing tool uses ICMP timestamp probes to estimate network-internal queueing delays.

Tools for Estimating Statistics from Individual Flows

• Tcptrace:
  Shawn Ostermann's Tcptrace. With an example of their time-sequence graphs.
  "tcptrace can take as input the files produced by several popular packet-capture programs, including tcpdump, snoop, etherpeek, HP Net Metrix, and WinDump. Tcptrace can produce several different types of output containing information on each connection seen, such as elapsed time, bytes and segments sent and recieved, retransmissions, round trip times, window advertisements, throughput, and more."

• Tstat - TCP Statistic and Analysis Tool:
  Tstat started as an evolution of TCPtrace, and monitors TCP anomalies, i.e., packet retransmission by RTO, FR, unnecessary retransmission, network reordering, duplicates, etc.

ISP-based Tools

• Network Measurement Tools from AT&T Labs-Research.
  These include PacketScope and GigaScope, the Traffic Aggregation Probe (TAP), OSPF and BGP monitors, and Netdb for looking at router configuration information.

• The IP Monitoring Project. from Sprint.
  "The IP Monitoring Project (IPMON) began with a goal of building a general purpose measurement system for IP networks capable of collecting both detailed packet-level traffic statistics as well as delay, loss, and other network performance statistics."

Other Tools

• Netperf includes tests for both unidirectional throughput and end-to-end latency.

• Scriptroute is a system for distributed Internet debugging and measurement.

• Kenjiro Cho's notes on tcpdpriv for removing user data, tcpdstat for getting summary info from a tcpdump file, and other tools. From 2000.

• Brian Tierney's TCP tuning guide for distributed application on wide area networks includes pointers to other tools besides pchar and pipechar, including iperf for measuring end-to-end performance, and NLANR's TCP testrig as a wrapper for tcpdump and tcptrace tools.

• Grenville Armitage's SIFTR is "a FreeBSD kernel module that logs statistics on active tcp connections to a file."

• Other tools: nping; Van Jacobson's pathchar; ntrace; xplot; pingroute.