Lower than Best Effort (LBE) Transport Protocols

Lower than best effort transport protocols work by scavenging bandwidth left available by standard best effort protocols. They can be implemented at the application level, or as alternative flavors of the standard TCP suite. This is especially true for uTP and LEDBAT, whose congestion control algorithm is the same, but that are implemented with an UDP or TCP framing respectively. In this page, we offer pointers to our own NICE and LEDBAT implementations, and collect a few more relevant pointers. We point out that it is out of the scope of this page to provide a comprehensive survey of lower than best effort protocols, for which we instead refer the reader to [COMST].

In particular, [TECHREP] assesses the state of lower than best-effort transport protocols available in the Linux kernel (you'll have to read in case the meaning of the colorful picture above, taken from it, is unclear to you), and offer a comparison of their performance (similarly to [LCN10], that however limitedly employed a simulative approach). These implementations have been used in [TMA13] to assess the interoperability with active queue management techniques (see also this page for more details and dataset), and in [P2P13] to gauge their impact on the performance of BitTorrent.

Open source

  • TCP NICE [NICE]
    • The original code of NICE authors is available at this webpage
    • Our own code used in [TECHREP] is available for Linux
  • uTP [UTP]
    • A standalone uTP implementation over UDP is available at github, now used in many BitTorrent clients
  • LEDBAT [LEDBAT]
    • A LEDBAT implementation as a TCP Congestion control module is available for Mac OS X
    • Our own LEDBAT implementation as a TCP module is available for Linux (detailed instructions and LEDBAT ns2 module are available here)

References

  1. [COMST] Ros, David and Welzl, Michael, Less-than-Best-Effort Service: A Survey of End-to-End Approaches . IEEE Commuication Surveys \& Tutorial, 2013.
  2. [P2P13] C. Testa, D. Rossi, A. Rao and A. Legout, Data Plane Throughput vs Control Plane Delay: Experimental Study ofBitTorrent Performance . In IEEE P2P'XIII, 2013.
  3. [TECHREP] D. Rossi, Open Source Scavenging Transport Protocols . In Technical Report, 2013.
  4. [LEDBAT] Shalunov, S., Hazel, G., Iyengar, J. and Kuehlewind, M., Low Extra Delay Background Transport (LEDBAT) . IETF RFC 6817, Dec 2012.
  5. [P2P11] Testa, C. and Rossi, D., The impact of uTP on BitTorrent completion time . In IEEE P2P'XI, 2011.
  6. [LCN10] Carofiglio, G., Muscariello, L., Rossi, D. and Testa, C., A hands-on Assessment of Transport Protocols with Lower than Best Effort Priority , 2010.
  7. [TCP-LP] Kuzmanovic, A. and Knightly, E.W., TCP-LP: low-priority service via end-point congestion control . IEEE/ACM Transactions on Networking, 14(4):752, 2006.
  8. [NICE] Venkataramani, A., Kokku, R. and Dahlin, M., TCP Nice: A mechanism for background transfers . USENIX OSDI, 36, 2002.
  9. [uTP] Norberg, Arvid, uTorrent transport protocol (uTP) . {BitTorrent Enhancement Proposals BEP29}.