From detlef.bosau at web.de Mon Sep 5 13:47:00 2011 From: detlef.bosau at web.de (Detlef Bosau) Date: Mon, 05 Sep 2011 22:47:00 +0200 Subject: [e2e] TCP Performance over WWAN In-Reply-To: <4E47FB37.5000708@freedesktop.org> References: <4E40684D.6060808@web.de> <4E470F9C.2040903@web.de> <4E47FB37.5000708@freedesktop.org> Message-ID: <4E653544.3060208@web.de> On 08/14/2011 06:43 PM, Jim Gettys wrote: > So attempting to retransmit a packet makes sense. > > The problem is, how many times? And when do you give up and try > something different? And eventually, you really must drop a packet (or > mark with ECN) to slow down the transmitter). Could we just hang on on this very question. How many times shall a base station retransmit a packet? And when shall the base station give up? Basically, this as implications for fairness and stability as well. Fairness, because in WWAN a base station may serve more than one mobiles. Hence, there is immediately a problem of resource allocation in downlink direction and vice versa a resource allocation and MAC problem in uplink direction. Stability, because a base station doing unlimited retransmission causes network load on the physical link. Hence, a "limited" input, e.g. a single packet, may cause an "unlimited" output, i.e. an unlimited number of retransmissions. This may immediately violate the BIBO stability, which is typically pursued in the Internet. We've well heard creeds on this issue, a very prominent one is the Berkeley SNOOP protocol. However, there is hardly anything concrete on this issue. Most of the statements I've read so far, stay astonishingly vague. Detlef From detlef.bosau at web.de Tue Sep 6 15:54:46 2011 From: detlef.bosau at web.de (Detlef Bosau) Date: Wed, 07 Sep 2011 00:54:46 +0200 Subject: [e2e] Local Recovery, TCP Performance over WWAN In-Reply-To: <4E653544.3060208@web.de> References: <4E40684D.6060808@web.de> <4E470F9C.2040903@web.de> <4E47FB37.5000708@freedesktop.org> <4E653544.3060208@web.de> Message-ID: <4E66A4B6.9030509@web.de> Hi to all. I'm curious, why it appears to be difficult, to get in touch with somebody to discuss these issues. Perhaps, my questions are not clear enough? I'm a bit lost on this one. Detlef -- ------------------------------------------------------------------ Detlef Bosau Galileistra?e 30 70565 Stuttgart Tel.: +49 711 5208031 mobile: +49 172 6819937 skype: detlef.bosau ICQ: 566129673 detlef.bosau at web.de http://www.detlef-bosau.de ------------------------------------------------------------------ From detlef.bosau at web.de Fri Sep 9 04:49:40 2011 From: detlef.bosau at web.de (Detlef Bosau) Date: Fri, 09 Sep 2011 13:49:40 +0200 Subject: [e2e] ARQ/HARQ Message-ID: <4E69FD54.2010408@web.de> In some private communication, my peer pointed out that in many cases a packet sent via a wireless link would reach the receiver either in only some few trials, say one or two, or would fail even in quite a lot of trials. Hence, the number of retransmission attempts could be set rather low, because either only one or few attempts would finally suffice - or even 20 or 30 attempts would not achieve the transmission. Unfortunately, my peer did not had hard facts to argue for this. However, one reason for this behaviour could be the burstiness of wireless errors. Another observation from actual BLER/SNR measurements is, refer to http://www.detlef-bosau.de/wrapper.php?file=from-symbols-to-packets/R4-020612_html_14c9b9cd.gif which is taken from http://www.3gpp.org/ftp/tsg_ran/WG4_Radio/TSGR4_22/Docs/R4-020612.zip that the transition from a nearly error free SNR range to a unusable SNR range is rather steep. Actually, in this HSDPA related example the difference from an SNR where an appropriately coded packet is conveyed nearly error free and an SNR where the transmission will fail is 1 dB or even less. Actually, I think this problem is addressed (at least in part) by HARQ approaches, which avoid the dumb repetition of inappropriately coded packets in case of transmission failure. If so, this would support a "binary" Model (like e.g. a Gilbert Markov Model, although a real wireless channel will hardly be Markovian) which only discriminates "link on" and "link off". For HSDPA, one would surely extend this in order to reflect several codings, were finaly a coding could be appropriate for an actual SNR, hence the transmission would be successful, or inappropriate, hence the transmission would fail. When I think of standards like the GPRS standards, this would leave QoS profiles which provide for a BLER of 10^(-9) rather questionable, because these are achieved, particularly in GPRS, by local retransmission. Are these thoughts reasonable? Detlef -- ------------------------------------------------------------------ Detlef Bosau Galileistra?e 30 70565 Stuttgart Tel.: +49 711 5208031 mobile: +49 172 6819937 skype: detlef.bosau ICQ: 566129673 detlef.bosau at web.de http://www.detlef-bosau.de ------------------------------------------------------------------ From Jon.Crowcroft at cl.cam.ac.uk Fri Sep 9 06:20:05 2011 From: Jon.Crowcroft at cl.cam.ac.uk (Jon Crowcroft) Date: Fri, 09 Sep 2011 14:20:05 +0100 Subject: [e2e] best e2e-type papers in the last 2 years? Message-ID: so i'm a bit out of touch these days. what would people here say are the best papers they've read in the past 2 years? stuff you'd give to someone to read first? stuff that could make a difference - i mean stuff like this...just for example http://www.cs.ucl.ac.uk/staff/m.handley/papers/mptcp_nsdi.pdf thanks j. From matta at bu.edu Wed Sep 14 13:50:51 2011 From: matta at bu.edu (Matta, Abraham I) Date: Wed, 14 Sep 2011 16:50:51 -0400 Subject: [e2e] CCW 2011 Call for Participation (cutoff date: This Friday) Message-ID: Apologies for multiple copies. Note that the cutoff date for registration and hotel is this Friday, Sep 16. ======= IEEE Computer Communications Workshop (CCW) 2011 Resort and Conference Center at Hyannis Cape Cod, MA http://www.ieee-ccw.org/ CCW is the annual flagship workshop of the IEEE Communications Society's Technical Committee on Computer Communications (TCCC). It is a panel-based workshop with informal, interactive sessions exploring emerging issues and trends in networking and computer communications. This year's program includes ten sessions on: - Security in Future Internet Architectures - Smart Grid - Addressing and Routing - Data Centers - Energy-efficient Networking - Information Access on the Internet - Network Control and Configuration - Home Networking - Social Networking - Wireless Networking Please kindly note: - The workshop program spans 2.5 days from Mon Oct 10 to Wed Oct 12 noon - Early registration deadline: Sep 16, 2011 - Hotel reservation cut-off due: Sep 16, 2011 We look forward to seeing you soon in Cape Cod! Lixin Gao, Umass and Ibrahim Matta, BU CCW 2011 Chairs From bltierney at lbl.gov Mon Sep 26 10:05:12 2011 From: bltierney at lbl.gov (Brian Tierney) Date: Mon, 26 Sep 2011 10:05:12 -0700 Subject: [e2e] Nationwide 100Gbps testbed available to researchers In-Reply-To: <4CAB6D7F.7040002@isi.edu> References: <4CAB6D7F.7040002@isi.edu> Message-ID: Hi all: This may be of interest to those of you working on protocols for high-speed networks. ESnet will be deploying a nationwide 100Gbps testbed by the end of this year. This testbed is available to anyone. For more information see: https://sites.google.com/a/lbl.gov/ani-testbed/proposal-process