REVIEWER B ---------- ===================== Guidelines for Reviewers ------------------------------------- Summary of guidelines for review of manuscripts submitted to IEEE Network for consideration for publication: 1.) Please provide a one-paragraph description of the content of this manuscript. The authors propose an Ipv6 extension that requires routers to report the departure times of processed packets, which should empower bandwidth estimation tools. 2.) Please identify and discuss the contribution of this manuscript. Please include in your discussion items such as the following: In general, I do not think that the authors have a significant contributions. Ipv6 and Ipv4 extensions requiring routers to report timestamps have been proposed for a while but none was adopted for various reasons. Also, the authors did not make a strong argument that timestamps at routers can provide bandwidth estimates that end-to-end techniques cannot at least approximately provide. The simulation section supporting their argument is very weak as illustrated below. 2.a) Does the paper have significant tutorial content? That is, is there enough background provided so that the generalist in networking can understand its main contributions? Elaborate. The description of the Ipv6 extension, while not technically deep in my opinion, provides enough background about Ipv6. Bandwidth estimation overview is neither thorough nor up-to-date. 2.b) Does the paper contain original contributions? What is the nature of the contributions? I do not think the authors have an original contribution whether as Ipv6 extension or in bandwidth estimation. 2.c) Is there a description of lessons learned that are given to the reader to help the reader avoid pitfalls in his own work? The case that timing information at routers will make a significant difference needs to be much stronger. The simulations justifying this claim is too simplistic and some important details like the bandwidth of the 20 links are not provided. Cartouche probes were not defined even though the INFOCOM reference is listed. Also, cartouche probes are used to infer capacity estimates and not available bandwidth estimates. The use of r=2 may affect the accuracy of the cartouche estimates. The larger r, the more accurate the estimates. If a larger r is used and more accurate estimates are obtained, does this mean that router timestamps are not useful? Probably not. But the authors need to pinpoint applications where routers’ timestamps accomplish what end-to-end estimates cannot. 2.d) Is there a need for a paper such as this in the Networking community? For example, are there articles that are already available which cover more or less the same topic at about the same depth? Measurement research is important but again, in my opinion, the contribution of this paper is minimal. 3) Please discuss the quality of the citations in this manuscript. If you think the citations should be improved, please provide specific references or sources of articles, such as journals or magazines, that should be consulted. There are many bandwidth estimation references that are missing. For example, refer to the related work section of Loguinov’s 2005 ICNP paper. 4) Please comment on the organization of the paper, and offer any suggestions that you think will improve the paper and its readability. -- 5) Please comment on the technical correctness of the manuscript in general, identify any specific technical inaccuracies that you find, and make suggestions for correcting those. The way available bandwidth estimates are computed from routers’ timestamps is not detailed and does not seem straightforward. The amount of queuing of the second probe packet at a router does not necessarily represent all the cross traffic that intervened between the probe packet-pair. This approximation did not show up in the simulation results, in which the accuracy was 100%. 6) If the manuscript does not require major revision, please provide a list of minor changes, such as spelling or grammatical errors, that need to be made. Please use the format   "p. 7., l. 18 somth ==> smooth" to mean  "on line18 of page 7, correct the spelling from somth to smooth." The paper requires major revision. 7) Please provide a summary comment on the overall suitability of the paper for publication in IEEE Network, assuming the recommended revisions are made. For example, if this is an outstanding contribution, please so state. If a major revision is needed, please so state. If the manuscript requires major editing, please so state. In its current shape and with the current content my recommendation is a “reject”.