NetNeutrality:TheTechnicalSideoftheDebate:A White PaperJon CrowcroftUniversity of Cambridge15 JJ Thomson AvenueCambridge CB3 0FD, [email protected] Neutrality is the subject of much current debate.In this white paper I try to find the signal in the noise by tak-ing a largely technical look at various definitions of networkneutrality and the feasibility and complexity of implement-ing systems that support those ideas.First off, there are a lot of emotional terms used to de-scribe various aspects of what makes the melting pot of theneutrality debate. For example, censorship or black-holing(where route filtering, fire-walling and port blocking mightsay what is happening in less insightful way); free-ridingis often bandied about to describe the business of makingmoney on the net (rather than overlay service provision);monopolistic tendencies, instead of the natural inclinationof an organisation that owns a lot of kit that they’ve sunkcapital into, to want to make revenue from it!The paper describes the basic realities of the net, whichhas never been a level playing field for many accidental andsome deliberate reasons, and then looks at the future evolu-tion of IP (and lower level) services; the evolution of overlayservices, and the evolution of the structure of the ISP busi-ness space (access, core and other); finally, I appeal to simpleminded economic and regulatory arguments to ask whetherthere is any case at all for special pleading for the Internetas a special case, different from other services, or utilities.Mutatis mutandisCategories and Subject DescriptorsA [.]: 2—Reference, C.2.1 [Packet-switching networks],C.2.4[Distributed applications], D1.3 [Distributed Programming],D4.1 [Scheduling], D4.4 [Network Communication], D.4.8[Stochastic Analysis], E.1 [Data Structures], G.1.6 [Con-strained Optimization]General TermsGeneral Terms: Performance, DesignKeywordsData Communications, Review1. INTRODUCTIONLet me try to illustrate the complexity and subtlety ofthe debate with a few, real stories from the last ten yearsof Internet Experience, each of which is chosen because itcaptures several facets of the problem space at once.Priority Rights Like many other people, I have 8Mbps In-ternet Access through an unbundled DSL broadbandprovider, which I share throughout my house using a$50 router to provide 10/100 Ethernet and Wirelessaccess to a server and the family’s laptops and mediacenters. I don’t secure the net with WEP keys and ac-cess control, since I use secure end systems with hostfirewalls and virus checkers etc etc, although the routerruns some useful filters to lower the background non-sense. When my phone line went down for 3 weeksearlier this year, my kids found 3 neighbours withopen WiFi access to their broadband lines (luckily allstill working – indeed 1 cable, and 2 different DSLproviders, one bundled, and one unbundled). We askedthem if it was OK to use their net (this is a UK legal re-quirement since recent precedents in unauthorised ac-cess to open WiFi nets being deemed an offence underthe Misuse of Computers Act). My neighbours said“sure”, but tellingly also admitted that this was be-cause our usage would have no impact on their usagesince they all used routers which implemented prior-ity queues (see, for example, http://openwrt.org/).While their nets were open, they had all independentlydiscovered that it was possible to set higher forwardingpriority for their own packets than everyone else, thusbeing socially friendly at the same time as not givingup any resource they paid for.We can unpack at least three lessons from this tale:firstly, it is literally child’s play to build communitywireless networks; secondly, it doesn’t take technicalexperts to deploy priority services; thirdly, cooperationand selfishness are not necessarily orthogonal.Content Re-Distribution In the mid 1990s, the UK Aca-demic Network provider, UKERNA, ran an experi-ment in usage charging for International Traffic. Thegoal was twofold: firstly, the charges might trickledown to real users and create a disincentive to mis-use or similar carelessness in moving large amounts ofdata around unnecessarily; secondly, the goal was toraise more revenue to pay for upgrades to the Interna-tional Links.It is one important lesson that the second of thesegoals was far more successful than the first, however,another piece of the story is interesting. A numberof national research networks in Europe provided verylarge web proxy caches to create a positive alternative(lower latency, potentially higher throughput etc). AtACM SIGCOMM Computer Communication Review 49 Volume 37, Number 1, January 2007the same time, UKERNA allowed free access to webcaches (very sensibly). However, other European coun-tries (who had not implemented the charging disincen-tive side of the story) soon noticed that users in someUK universities were using their caches (especially inwell provisioned areas with good UK connectivity suchas Scandinavia). They rapidly introduced first prior-ities (better access for IP sources in their own ASs),and eventually blocking of IP addresses outside theirown networks.Data and Digital Mobile Phones I have a cell phone.It uses around 14kbps to carry voice, and provides aglobal service which is extremely pervasive and afford-able. Indeed, there are more cell phones than InternetHosts (2.5 billion active mobile phone numbers in theworld at the time of writing). My cell phone providesdata (GPRS, EDGE and 3G as it happens). The 3Gservice runs at around 384kbps in the UK, and seemsto have pretty low latency – I do not know the ar-chitecture of the backhaul network once the wirelesssegment of a route is terminated, but it seems to sup-port pretty close to zero loss. I can run Skype or anyvanilla VoIP system on this fairly easily. However, thevolume and time tariff of the data service is set suchthat a normal pattern of voice calls made over it wouldcost more than the GSM service. This is fairly surreal(in fact, usually when I read my e-mail via my phone,I “dial-up” over GSM as it is cheaper), but you can seethat there are powerful reasons for the cellular networkproviders to stay in this regime for a while, or else haveto explain a massive loss of revenue to their sharehold-ers. Maybe? If not, what is their replacement sourceof revenue?They key lesson here is that legacy service providersresist the pressure to become merely bit pipes.Digital TV and Fibre Another UK
View Full Document
Unlocking...