From detlef.bosau at web.de Fri Mar 13 04:54:22 2009 From: detlef.bosau at web.de (Detlef Bosau) Date: Fri, 13 Mar 2009 12:54:22 +0100 Subject: [e2e] TCP Loss Differentiation In-Reply-To: <49908411.1000800@reed.com> References: <49902B22.80403@web.de> <499058D5.9040807@reed.com> <49908411.1000800@reed.com> Message-ID: <49BA496E.2030100@web.de> David P. Reed wrote: > > My main point was that these loss processes are not characterizable by > a "link loss rate". They are not like Poisson losses at all, which > are statistically a single parameter (called "rate"), memoryless > distribution. They are causal, correlated, memory-full processes. > And more importantly, one end or the other of the relevant link > experiences a directly sensed "loss of connectivity" event. > Does anybody happen to have some good reference for this one? Something like: "The failure of poisson modelling of mobile wireless links" or something simuilar? What I have seen so far, simply assumes the contrary and uses Gilbert Markov Models and the like. Actually, this fails to model even a scenario where a user is waiting for a bus at the bus station and talking to someone else with his mobile and walks around a bit, one step to the left (broken connection) or to the right (excellent connectivity). Although this seems rather obvious to me, I don't know a paper, or a well accepted text book, which explicitely states that Gilbert Markov Models and the like are theoretically appealing, but do not reflect reality. I would appreciate any hint here. Detlef Bosau Mail: detlef.bosau at web.de Galileistrasse 30 Web: http://www.detlef-bosau.de 70565 Stuttgart Skype: detlef.bosau Mobile: +49 172 681 9937 From lachlan.andrew at gmail.com Fri Mar 13 16:00:46 2009 From: lachlan.andrew at gmail.com (Lachlan Andrew) Date: Sat, 14 Mar 2009 10:00:46 +1100 Subject: [e2e] TCP Loss Differentiation In-Reply-To: <49BA496E.2030100@web.de> References: <49902B22.80403@web.de> <499058D5.9040807@reed.com> <49908411.1000800@reed.com> <49BA496E.2030100@web.de> Message-ID: Greetings Detlef, 2009/3/13 Detlef Bosau : > David P. Reed wrote: >> >> My main point was that these loss processes are not characterizable by a >> "link loss rate". ?They are not like Poisson losses at all, which are >> statistically a single parameter (called "rate"), memoryless distribution. >> ?They are causal, correlated, memory-full processes. And more importantly, >> one end or the other of the relevant link experiences a directly sensed >> "loss of connectivity" event. >> > > Does anybody happen to have some good reference for this one? Something > like: "The failure of poisson modelling of mobile wireless links" or > something > simuilar? > > What I have seen so far, simply assumes the contrary and uses Gilbert Markov > Models and the like. Although the Gilbert model is far from perfect, it is very much better than a Poisson model for wireless. They are correlated and "memory-full", and have the notion of "loss of connectivity" (i.e., being in the bad state). It certainly can model the case you describe of periods of a broken connection interleaved with excellent connectivity. Although your work may need better wireless models, I think that for most people on this list the law of diminishing returns means that going from Poisson to Gilbert is enough. David, could you explain what it means for a stochastic process to be "causal"? My understanding was that a filtration on a random process is always causal in the sense of having being increasing in one direction, while the time reverse of the underlying random process is always another valid random process, albeit a different process except in the case of reversible processes. Cheers, Lachlan -- Lachlan Andrew Centre for Advanced Internet Architectures (CAIA) Swinburne University of Technology, Melbourne, Australia Ph +61 3 9214 4837 From kuntan at microsoft.com Sat Mar 14 03:52:26 2009 From: kuntan at microsoft.com (Kun Tan) Date: Sat, 14 Mar 2009 18:52:26 +0800 Subject: [e2e] CFP - MobiArch 2009 In-Reply-To: References: <49902B22.80403@web.de> <499058D5.9040807@reed.com> <49908411.1000800@reed.com> <49BA496E.2030100@web.de> Message-ID: <568659FFCAED224589FFA08DDC9667D815C31A191A@AA-EXMSG-C426.southpacific.corp.microsoft.com> Sorry if you have received duplicated copies. - Kun Tan Researcher | Wireless and Networking Group | Microsoft Research Asia ==================================================================================================== Call for Papers The 4th ACM International Workshop on Mobility in the Evolving Internet Architecture (MobiArch 2009) ----------------------------------------------------- Co-located with MobiSys, 22 June 2009, Krakow, Poland http://user.informatik.uni-goettingen.de/~mobiarch/2009/ With the recent development of technologies in wireless access and mobile devices, user, terminal, and network mobility has become an indispensable component of today's Internet vision, and it is likely to continue in the near future, while affecting the whole architectural design of the future Internet. Yet, issues like efficient mobility management and optimization, locator-identifier split, multi-homing, security, wireless access and related operational/deployment concerns are still in their early stages of development. Moreover, the Internet architecture, its end-to-end principles, and business models will require rethinking due to the massive penetration of mobility into the Internet. For instance, an appropriate system that allows communicating with a mobile host requires addressing several fundamental issues with the Internet architecture, such as ability to locate the mobile host/service, preserving ongoing communications upon changes of locations, as well as efficient and secure handover management. As another example, the emerging wireless technologies may pose additional challenges to the Internet architecture since they introduce design principles different from the original Internet. MobiArch 2009 welcomes submissions, from both researchers and practitioners, in exploration of recent advances in architectures, protocols, and experiences with emerging technologies on various mobility issues over the Internet, with an emphasis on wireless infrastructures and mobility patterns for mobility support, new mobility protocols, service discovery, routing and location management, mobile network performance evaluation and modelling, multi-homing, security, architectural impacts and deployment considerations. Furthermore, the potential of usability of mobility services for connecting people and devices in developing regions of the world into the Internet infrastructure will be also explored. Topics of MobiArch 2009 cover all aspects of architectural issues and system support for mobility in the Internet, including but not limited to: * Impacts of new wireless technologies/services, networking technologies, and mobility patterns on the Internet architecture * Architectures and protocols for mobility support in the Internet, ranging from approaches in link, network, transport to session/application layers and cross-layer design * Location management, routing, locator/identifier split, multi-homing and load sharing issues * Security and privacy issues in mobility networks and impacts to Internet architecture * Architectures and mechanisms for wireless/mobile connectivity in extreme environments (e.g., remote areas, developing countries) * Performance issues with mobility in the Internet * QoS and middlebox issues in mobility networks and impacts to Internet architecture * Economic and deployment issues of mobility solutions (infrastructure and devices) * Impact of social aspects on mobility architectures, mobile application and protocol design * Technologies for mobile wireless access and interactions Submissions must present original results. Selected papers will be forward-looking, describe their relationship to existing work, and have impact and implications for ongoing or future research. Submitted papers must be no more than 6 pages long, two columns, with no characters in smaller than 10 point fonts, and must fit properly on US "Letter"-sized paper (8.5x11 inches). Margins must be of 1 inch on all edges (top, bottom, left, and right) of each page. All paper submission will be handled via EDAS (http://edas.info/7188). Papers will be reviewed single blind. Important Dates: Submission Deadline: 15 March 2009 (extended) Acceptance Notification: 31 March 2009 Camera Ready Due: 15 April 2009 Workshop Date: 22 June 2009 TPC Chairs: Joerg Ott Helsinki University of Technology (TKK) Kun Tan Microsoft Research Asia For more information, contact workshop co-chairs at mobiarch-chairs at informatik.uni-goettingen.de. From dpreed at reed.com Sat Mar 14 08:51:20 2009 From: dpreed at reed.com (David P. Reed) Date: Sat, 14 Mar 2009 11:51:20 -0400 Subject: [e2e] TCP Loss Differentiation In-Reply-To: References: <49902B22.80403@web.de> <499058D5.9040807@reed.com> <49908411.1000800@reed.com> <49BA496E.2030100@web.de> Message-ID: <49BBD278.6030409@reed.com> The primary causality I was trying to reflect by that choice of words is that in all systems there are large-scale causality relationships (application layer or user layer, in most cases) that break the fundamental assumption that there are a collection of memoryless processes in the system. These are really significant in most networks, ignored by most models: - users behave differently when networks get slow (they go to the cafe or panic and start hitting keys harder and faster). - wireless network transmissions are a primary source of noise to other transmissions, so any correlations or dependence can get amplified when noise causes retransmission by other nodes, causing more noise, ... the probability that wireless networks have modes of highly synchronizing "resonances" that correlate rather than decorrelate signals is high - and can be used to increase SNR by techniques like analog network coding (zigzag, for example) if you realize that the phenomenon is not a noise process at all, since it adds no uncertainty. It's easy to create models full of independent, memoryless processes that *appear* to the student (or full professor) to be valid because the normal human reaction is to think that complexity is best modeled by lots of randomness. Complex systems are not random merely because they are complex. E.G., PRNG's are perfectly non-random. It's a form of mysticism to think that they are the same as true random processes outside a very narrow domain of applicability, where they can add insight. This is where most engineering math practitioners get it wrong. Not knowing when your modeling approach fails... because you use the "religion of your peers" (the NS2 wireless modeling toolkit, for example, which teaches nothing about actual propagation and dependence of noise). Lachlan Andrew wrote: > Greetings Detlef, > > 2009/3/13 Detlef Bosau : > >> David P. Reed wrote: >> >>> My main point was that these loss processes are not characterizable by a >>> "link loss rate". They are not like Poisson losses at all, which are >>> statistically a single parameter (called "rate"), memoryless distribution. >>> They are causal, correlated, memory-full processes. And more importantly, >>> one end or the other of the relevant link experiences a directly sensed >>> "loss of connectivity" event. >>> >>> >> Does anybody happen to have some good reference for this one? Something >> like: "The failure of poisson modelling of mobile wireless links" or >> something >> simuilar? >> >> What I have seen so far, simply assumes the contrary and uses Gilbert Markov >> Models and the like. >> > > Although the Gilbert model is far from perfect, it is very much better > than a Poisson model for wireless. They are correlated and > "memory-full", and have the notion of "loss of connectivity" (i.e., > being in the bad state). It certainly can model the case you describe > of periods of a broken connection interleaved with excellent > connectivity. > > Although your work may need better wireless models, I think that for > most people on this list the law of diminishing returns means that > going from Poisson to Gilbert is enough. > > David, could you explain what it means for a stochastic process to be > "causal"? My understanding was that a filtration on a random process > is always causal in the sense of having being increasing in one > direction, while the time reverse of the underlying random process is > always another valid random process, albeit a different process except > in the case of reversible processes. > > Cheers, > Lachlan > > From touch at ISI.EDU Sat Mar 14 10:20:07 2009 From: touch at ISI.EDU (Joe Touch) Date: Sat, 14 Mar 2009 10:20:07 -0700 Subject: [e2e] CFP test -- please ignore Message-ID: <49BBE747.9030306@isi.edu> Please ignore. From touch at ISI.EDU Sat Mar 14 10:25:38 2009 From: touch at ISI.EDU (Joe Touch) Date: Sat, 14 Mar 2009 10:25:38 -0700 Subject: [e2e] CFP test 2 -- please ignore In-Reply-To: <49BBE747.9030306@isi.edu> References: <49BBE747.9030306@isi.edu> Message-ID: <49BBE892.10208@isi.edu> Please ignore. From touch at ISI.EDU Sat Mar 14 11:05:34 2009 From: touch at ISI.EDU (Joe Touch) Date: Sat, 14 Mar 2009 11:05:34 -0700 Subject: [e2e] CFP test 3 - please ignore Message-ID: <49BBF1EE.5020901@isi.edu> From touch at ISI.EDU Sat Mar 14 11:05:49 2009 From: touch at ISI.EDU (Joe Touch) Date: Sat, 14 Mar 2009 11:05:49 -0700 Subject: [e2e] cfp test 4 - please ignore Message-ID: <49BBF1FD.1000806@isi.edu> From detlef.bosau at web.de Sat Mar 14 15:04:01 2009 From: detlef.bosau at web.de (Detlef Bosau) Date: Sat, 14 Mar 2009 23:04:01 +0100 Subject: [e2e] TCP Loss Differentiation In-Reply-To: <49BBD278.6030409@reed.com> References: <49902B22.80403@web.de> <499058D5.9040807@reed.com> <49908411.1000800@reed.com> <49BA496E.2030100@web.de> <49BBD278.6030409@reed.com> Message-ID: <49BC29D1.6090701@web.de> David P. Reed wrote: > The primary causality I was trying to reflect by that choice of words > is that in all systems there are large-scale causality relationships > (application layer or I think, "large scale" is the very point. When the drop/loss ratio is approximately constant over a certain period of time, it does not really matter whether the sender's individual response to a loss event is correct or not. As long as the sender does congestion recovery "adequately often", anything is fine. > user layer, in most cases) that break the fundamental assumption that > there are a collection of memoryless processes in the system. These > are really significant in most networks, ignored by most models: > > - users behave differently when networks get slow (they go to the cafe > or panic and start hitting keys harder and faster). This exactly reflects my "bus stop scenario". A user may move some feet to the left or to the right - and even this may result in a Rayleigh fading maximum or a Rayleigh fading minimum. And we did not even talk about other causes for fading or interference. > - wireless network transmissions are a primary source of noise to > other transmissions, so any correlations or dependence can get > amplified when noise causes retransmission by other nodes, causing > more noise, ... the probability that wireless networks have modes of > highly synchronizing "resonances" that correlate rather than > decorrelate signals is high - and can be used to increase SNR by > techniques like analog network coding (zigzag, for example) if you > realize that the phenomenon is not a noise process at all, since it > adds no uncertainty. That's, what's continously ignored by quite a lot of work. Now, the problem for me is twofold. First: If an end-to-end loss differentiation in a highly dynamic mobile networking environment is not feasible, this would be helpful for the little sketch I submitted recently. (And now, I'm eager to read the reviewer comments.) Second: I'm still thinking, how an alternative approach can be evaluated. All the simulations I know so far assume Markov models, Gilbert Markov models, Jake's model and the like, which exhibit a more or less stationary or quasi stationary behaviour, particularly on large time scales. -- Detlef Bosau Mail: detlef.bosau at web.de Galileistrasse 30 Web: http://www.detlef-bosau.de 70565 Stuttgart Skype: detlef.bosau Mobile: +49 172 681 9937 -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3351 bytes Desc: S/MIME Cryptographic Signature Url : http://mailman.postel.org/pipermail/end2end-interest/attachments/20090314/cd90631e/smime.bin From cwestphal at docomolabs-usa.com Mon Mar 23 11:13:31 2009 From: cwestphal at docomolabs-usa.com (Cedric J A. Westphal) Date: Mon, 23 Mar 2009 11:13:31 -0700 Subject: [e2e] Last Reminder: VISA'09 Paper submission deadline this Friday, 3/27 (ACM SIGCOMM workshop on Virtualized Infrastructure, Systems and Architectures) In-Reply-To: References: Message-ID: <2FD61F37AFF16D4DB46149330E4273C7019FDB1A@dcl-ex.dcml.docomolabs-usa.com> Dear Colleagues, this is a last reminder for the submission deadline for the ACM Sigcomm Workshop on Virtualized Infrastructure Architectures and Systems (VISA'09, Barcelona, August 17). Do not forget to upload your paper! Paper submission deadline: Friday March 27, 2009 Acceptance notification: Friday May 1, 2009 Camera ready final submission: Friday May 29, 2009 Workshop: Monday August 17, 2009 This is a firm deadline. Please find the final CFP below. ----------------------------------------- --- VISA 2009: The First ACM SIGCOMM Workshop on Virtualized Infrastructure Systems and Architectures --- in conjunction with ACM SIGCOMM 2009, August 17th 2009, Barcelona, Spain. http://conferences.sigcomm.org/sigcomm/2009/workshops/visa/ * FINAL CALL FOR PAPERS * Infrastructure virtualization has emerged as an important architecture and experimentation concept for the Internet infrastructure. The global computing and communication infrastructure will encompass (as it does today) a diverse and huge collection of networking, computing and storage resources. Together they need to form a coherent infrastructure and meet our society's requirements for the 21st century. Infrastructure virtualization involves creation of a virtual slice of network, computing and storage resources in support of a service, an application, or an experiment from a physical substrate of diverse resources. This allows users of a virtualized infrastructure slice to access resources on a potentially global scale without incurring the cost of building such an infrastructure. Thus infrastructure virtualization provides a platform to allow innovation on a global scale and enables new business models. As we envision and research Future Internet, there is increasing recognition that Infrastructure Virtualization will play an important role. However, there are many technical problems to solve: how to discover and advertise the resources; how to create and manage an infrastructure slice across diverse resources; how does virtualization extend to the wireless edge; how to implement virtualization across diverse resources and across layers of protocol stack; how to map an application or service to run on an infrastructure slice; what applications and capabilities are enabled by infrastructure virtualization; what kind of cross-layer protocols are possible; how does infrastructure virtualization impact the business models of network operators; and others. Many research groups in the US, Europe, Japan, and elsewhere are pursuing different aspects of infrastructure virtualization; various international funding agencies are actively supporting research in this area; and many providers and vendors are very interested in exploring how this concept and associated technologies would help solve their business problems and create new growth opportunities. The goal of the workshop is to feature recent research and developments related to infrastructure virtualization to allow exchange of ideas and help build a research and user community to explore and help realize the potential of infrastructure virtualization. * TOPICS OF INTEREST * We solicit previously unpublished work on the following, non exhaustive, list of topics: * Infrastructure virtualization architecture; * Resource allocation to virtual slices; * Management tools for infrastructure virtualization; * Implementation and transition map for infrastructure virtualization; * Isolation and slice independence in a virtualized infrastructure; * Integration of the wireless edge into a virtualized network; * Inter-operability and federation of virtualized infrastructures; * Cross-layer protocols for virtualized networks; * Applications and services enabled by virtualized infrastructure; and * Security issues with virtualized infrastructure. * EXTENDED DEADLINES * Paper submission deadline: Friday March 27, 2009 Acceptance notification: Friday May 1, 2009 Camera ready final submission: Friday May 29, 2009 Workshop: Monday August 17, 2009 * INSTRUCTIONS FOR AUTHORS * Authors should submit pdf papers exclusively, to the EasyChair conference management system: http://www.easychair.org/conferences/?conf=visa09. Please follow the format of ACM Sigcomm 2009 submissions, except anonymity is not required and the page limit is eight pages. This workshop strongly encourages the submission of exploratory results that point to new directions and challenges in the design and management of a virtualized infrastructure. Please contact visa09 at easychair.org for any inquiry regarding the workshop. * WORKSHOP ORGANIZATION * Steering Committee: - Tomonori Aoyama, Keio University and NICT - Anja Feldmann, TU Berlin and T Labs - Nick McKeown, Stanford University - Guru Parulkar, Stanford University - Larry Peterson, Princeton University - Cedric Westphal, DoCoMo Labs USA Program Chairs: - Guru Parulkar, Stanford University - Cedric Westphal, DoCoMo Labs USA Program Committee Members: - Hasan Alkhatib, Microsoft - Tomonori Aoyama, Keio University and NICT - Jack Brassil, HP Labs - Stephan Baucke, Ericsson - Simon Crosby, Citrix Systems - Christophe Diot, Thomson Labs - Lars Eggert, Nokia Research Center - Serge Fdida, UPMC - Paris 6 - Nick Feamster, Georgia Tech - Anja Feldmann, TU Berlin and T Labs - Silvano Gai, Cisco Systems - Albert Greenberg, Microsoft - James Kempf, Ericsson - Dae Young Kim, Chungnam National University - Ulas Kozat, Docomo Labs USA - Laurent Mathy, Lancaster University - Nick McKeown, Stanford University - Sue Moon, KAIST - Akihiro Nakao, The University of Tokyo - K.K. Ramakrishnan, AT&T Labs - Dipankar Raychaudhuri, Rutgers University - Jennifer Rexford, Princeton University - Robert Ricci, University of Utah - Martin Stiemerling, NEC Labs Europe - Amin Vahdat, UCSD From touch at ISI.EDU Thu Mar 26 09:00:01 2009 From: touch at ISI.EDU (Joe Touch) Date: Thu, 26 Mar 2009 09:00:01 -0700 Subject: [e2e] Last Reminder: VISA'09 Paper submission deadline this Friday, 3/27 (ACM SIGCOMM workshop on Virtualized Infrastructure, Systems and Architectures) In-Reply-To: <2FD61F37AFF16D4DB46149330E4273C7019FDB1A@dcl-ex.dcml.docomolabs-usa.com> References: <2FD61F37AFF16D4DB46149330E4273C7019FDB1A@dcl-ex.dcml.docomolabs-usa.com> Message-ID: <49CBA681.9050408@isi.edu> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Cedric J A. Westphal wrote: > Dear Colleagues, > > this is a last reminder ... This is a reminder to all subscribers that this list does not permit reminders, extension notices, or any other repeated announcements for meetings. *ALL* meeting notices must be approved in advance. This note did not follow that protocol, FYI. Joe (as list admin) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAknLpoEACgkQE5f5cImnZrtwKgCgrM+akHgpyvHPD4Ud3rSRYzbh NswAn1ZnsOqtNwoT70qnI1WfLmJDowxT =gwJu -----END PGP SIGNATURE----- From laurent at comp.lancs.ac.uk Tue Mar 31 11:05:21 2009 From: laurent at comp.lancs.ac.uk (Laurent Mathy) Date: Tue, 31 Mar 2009 19:05:21 +0100 (BST) Subject: [e2e] CFP IMC 2009 Message-ID: <200903311805.n2VI5LSj011960@gateway.comp.lancs.ac.uk> [Apologies for multiple copies] CALL FOR PAPERS Internet Measurement Conference (IMC) 2009 Sponsored by ACM SIGCOMM and in cooperation with USENIX November 4 - 6, 2009 Chicago, Illinois, USA http://www.imconf.net/imc-2009/ The Ninth Internet Measurement Conference is a two and a half day event focusing on Internet measurement and analysis, building on the success of past IMCs. We invite submissions of papers that contribute to our understanding of how to collect or analyze Internet measurements, or give insight into Internet structure and behavior. Examples of relevant topics are: - Internet traffic analysis - Internet structure and topology characteristics - Internet performance measurements - Measurement-based network management such as traffic engineering - Inter-domain and intra-domain routing - Network applications such as multimedia streaming, gaming and on-line social networks - Measurements of content distribution, peer-to-peer, overlay, and social networks - Data-centric issues, including anonymization, querying, and storage - Measurement-based inference of network properties - Design of monitoring systems, sampling methods, signal processing methods - Network anomaly detection - Network security threats and countermeasures - Software tools and environments in support of measurement - Measurement-based assessment of simulation/testbeds - Measurement-based workload generation - Measurement-based modeling - Reappraisal of previous measurement findings - Internet-oriented wireless, and mobility measurement Papers that do not in some fashion relate to measuring Internet properties are out of scope. Authors can contact the Program Chairs at imc09tpchairs at comp.lancs.ac.uk for clarification if they are unsure whether their paper is in scope. Ethical standards for measurement must be considered by all IMC authors. In particular, authors must be aware of and conform to acceptable use policies for individual domains that are probed or monitored, data privacy and anonymity for all personally identifiable information, and etiquette for using shared measurement data (see Allman and Paxson, IMC '07). If applicable, authors are also urged to notify parties of security flaws in their products or services in advance of publication. Adherence to ethical standards for measurement will be a criteria for all submissions and violations will be grounds for rejection. Important dates May 4, 2009: 10PM EDT: Registration of title and 250-word abstract May 11, 2009: 10PM EDT HARD submission deadline July 17, 2009: Notification August 22, 2009: Camera Ready Copy due