Upgrade to Pro — share decks privately, control downloads, hide ads and more …

1997 - "IP Dialtone" Keynote - Networld+InterOp...

1997 - "IP Dialtone" Keynote - Networld+InterOp Las Vegass

Keynote presentation slides for "IP Dialtone" by Tom Lyon, Founder and CTO of Ipsilon Networks.
Video available at https://www.youtube.com/watch?v=lHPFZ4xyuBE&ab_channel=TomLyon

Tom Lyon

May 07, 1997
Tweet

More Decks by Tom Lyon

Other Decks in Technology

Transcript

  1. 1 IP Dialtone: Access for the New Internet www.ipsilon.com 1

    888 IPSILON Tom Lyon C.T.Q. lpsilon Networks, Inc. J) |P§,!§LQ§
  2. The Telephone Dialtone I Dialtone - the sound we’ve come

    to love, trust, and depend on I lt’s always there for you I Means the network is ready to place or receive calls I If not there - phone must be broken I Taken for granted by users
  3. Dialtone as Service I For phone companies, dial tone is

    not trivial, its a SERVICE I Many kinds of dialtone sen/ice: ~ LifeLine, Residential, Business, Party Line, PBX I Incoming, outgoing, local only calls I # of sen/ices limited by regulation, lack of competition I Long Distance Sen/ices: competition, many types - Friends &family, etc.
  4. The Busy Signal I Hardly heard anymore, due to ~

    answering machines, voice mail, fon/varding, cell phones, pagers I Busy == Failure I Seems like we only get it trying to reach the Internet I Widely reported AOL busy problems I ATT Reliability Challenge - 94% success and climbing
  5. The Busy Network I Féatgt Busy signal means the network

    is busy, not the I Even without ISP problems, network has problems I The phone network is not engineered to meet Internet requirements I PacBeIl - 27% of residential Minutes Of Use in 1996 were for Internet access I People who depend on the Internet avoid dial-up problems by being connected all the time... I REAL INTERNET USERS DON’T GET BUSY SIGNALS
  6. The Consumer (Dial-Up) Internet I TODA Y’S CONSUMER INTERNET IS

    NOT WHAT ANYONE WANTS I Users don’t like it I Internet Service Providers don’t like it I Phone Companies don’t like it
  7. Users don ’t like Dial-Up I Not reliable - 94%

    reliability is not enough for any consumer product I Too much time for short transactions ~ 2-3 minutes to dialup, listen to modems ~ all to find no email I Misconfiguration can cause big billing problems - Automated transparent dialing I Consumers need reliability, instant gratification, safety net
  8. ISPs don ’t like Dial-Up i I Harder to support

    than full time connections I Configuration / Authentication issues I No quality control over connection I Hard to compete with un-differentiated sen/ice I Bizarre effect - well established |SPs have fewer dialup customers than younger ones _
  9. Phone companies don ’t like Dial-Up I Most Internet calls

    are free, local, long duration calls - (free in US, often subsidized in other countries) I Costs a lot of money to maintain and enhance phone network I No revenue from fastest growing use I Economic reality will soon overthrow current regulations I Are lSPs customers or competitors of phone companies?
  10. Dial-Up is Dead All 3 interest groups involved in dial-up

    - users, lSPs, phone companies - don’t like it. The lSPs and phone companies lose money, the users lose patience with dial-up For acceptable sen/ice & sustainable growth, must abandon dial-up model of Internet access DIAL-UP IS DEAD - WHAT THE CONSUMER NEEDS IS IP DIALTONE -
  11. IP DialTone n IP Dialtone - a permanent open channel

    to the internet it The network becomes only a packet away, not a phone call away It Lets you get to the network, but also lets the network come to you it Like telephone dialtone, IP Dialtone is a single interface to many different services - Internet, intranet, extranet, etc. l 1 _
  12. My House I To get my house on the Internet...

    I bought CAT5 cable I ran it to several rooms I ran ISDN line, got ISDN modem & Ethernet Hub I Ran AC wiring - recommend professionals I Very cool wiring closet I Lots of friends with Ethernet in their homes
  13. What does Dilbert’s Mom say? 5 é § rw :emu~wiAs<£aau

    - zmwwwwzwuo A £1“~~°‘*‘EE5 T9 *5 ‘ 1 sazuam manna aianw? sh-1'-&?£_¢L£ T6 FKIENJS 1 - ma ALREADY G01’ A mo xmxmr. z THINK * » 1 mu mum us: mm. FRAME RE1,_A*~: DROP" 156* MN w£& 55%-r£R {hi ‘ME. 5EU"£l~lG 9.06:-'~. I5 I GUING "F0 BE A. SALE. H£§{LQ‘Q'! arm N23 DAYA imam we j " 5? RTS
  14. Business Network of Today I Dilbert’s mom gets it I

    For success, copy what businesses do I Frame relay drop to Ethernet LANs is wildly popular for internet access ~ not a dialup solution I THE NETWORK MODEL FOR THE HOUSE OF TOMORROW WILL BE THE SAME AS FOR THE BUSINESS OF TODAY
  15. Business of Today I TECHNOLOGIES FOR THE HOME WON’T SUCCEED

    UNTIL THEY’VE BEEN MA TURED BY BUSINESSES - AND MADE A LOT CHEAPER I Look at other consumer/tech items - FAX, PCs, Cell Phones, Pagers I Cost and maturity are most often achieved through evolution & refinement, not revolution I’ Continuum of markets between office & home - schools, hotel, apartments, universities, cafes, arcades, cars...
  16. From Business to Home I Today, many office users already

    have IP Dialtone - they’re connected all the time I What will happen as IP Dialtone makes its way to the house of tomorrow?
  17. Home Usage I BANDWIDTH@HOME > BANDWIDTH@ WORK I Need it

    - for effective telecommuting I Want it - for the fun stuff on the Internet ~ 3D VR Real time gaming - Video on Demand - Catalog Shopping ~ Effective Gov’t Access I Get it - because others want to get it to you I Already - more MIPS/PC @ home
  18. Services I Raw bandwidth and simple connectivity are not enough

    to satisfy the fickle consumer, or even today’s business I CONNECTIVITY IS BECOMING A COMMODITY, REAL DEMAND IS TURNING TO BETTER AND MORE DIFFERENTIATED SERVICES
  19. Differentiated Services I bandwidth - access bandwidth control, end-end bandwidth

    guarantees for some paths / offices/ transactions I latency - guarantees of good response for interactive gaming or voice applications I priority - first class, business, coach style I privacy - encryption & anti-snooping, VPNs, extranets I reliability - redundant provision, rebates for downtime I billing plans - cell phone options I Content related services - being close to the data
  20. Why Services? I Without well defined, consistent & measurable levels

    of service, customers will become frustrated & lose interest I Providers need differentiated sen/ices to manage revenue, differentiate customers, not be stuck in commodity business
  21. Current Differentiated ISPs I ISPs already starting to differentiate.. .

    I Toyota Digital Cruise ~ Connects all Toyota suppliers, dealers, factories - Target efficiencies in the supply & delivery functions - Worldwide, gateways to the rest of the Internet I Sprint DRUMS - Production collaboration tools for Hollywood ~ Studio or field connections I Freese-Notis Weather Net ~ Gather & Distribute weather information - High bandwidth, near real time, big agro customers
  22. Corporations as Service Providers I IF YOU RUN THE CORPORATE

    NETWORK, YOU’RE A SERVICE PROVIDER TOO I If you’re a monopoly, just ignore the users - but these days, the CEO is a user too I 2 ways to better service ~ socialist -_ add resources everywhere, hope users stop complaining, charge everyone ~ capitalist - control the resources so that users who pay for more resources get them; users who pay less get less I ln a small company, or at home, we’re one happy family - but as we grow, we need different rooms, good fences, control in the network
  23. Driving Forcesfor IP Dialtone I Customer Interest I Content Providers

    I Internet as a new, efficient medium I Advertising I Competition
  24. Customer Interest I Business and consumer interest levels are extremely

    high I Beyond technical revolution to social revolution I New forms of social interaction - email, surfing, chat rooms, MUDs, MOOs I We may not understand why, but why ask why?
  25. Content Providers I If you can imagine it, its on

    the Internet I Can’t begin to index the indices I Now anyone can publish, apparently there’s some basic human urge to do so I But businesses become content providers because
  26. Internet as a new, efficient medium - ii , -i

    v_ _. ., . ,. _, , _ . ., . ii . vii. , , ,. . . ,, i. i.-. ., Ti... Y W I-Tin“ in-I |,/ I, -»i|= , .. , ~.|»/.4‘ 1.," I-ii ii e . _, __ . , I The Internet is a wonderfully efficient medium which is better at a lot of things than other media I Email replacing paper mail I Web replacing publishing I Push services replacing news I Voice over IP replacing phone calls I Video over IP replacing broadcasting I Even CyberCash replacing money l l l
  27. Advertising I All this stuff will require a lot of

    new infrastructure. Who will pay? There’s someone willing and eager to pay for the bandwidth... ADVERTISERS I Advertising already pays for - newspapers, magazines, TV, radio, junk mail I Telephone? Over 1/2 of phone calls are now to toll- free, business funded numbers I Internet is great at both mass media and direct marketing - they know who we are and what we like
  28. Advertising .1 . . . .> 1i it I I/"1-i-»~i»i~~_'g'

    .. it - ~i.-"- I»-It I ~ - ._. ~ » lim ii v It'‘Vi » mi .. K ~ I.i»@»|-liiliiii ii . n-i I...‘/Inn >A iii‘ iv.-d h% mi-T3“‘hi.’ I ti“iihlmhi iniii I Every web page has 3 or 4 banner ads already I Ads are becoming interactive ~ HP ping-pong banner ad I $300 million in 1996 on Web ads ~ Growing fast, like everything else on Internet I Advertisers compete for your attention
  29. Competition I Effective competition exists or is coming ~ between

    content providers - between Internet Sen/ice Providers ' between equipment & software providers - between basic infrastructure providers I Everyone wants a piece of the action I Everyone with real estate, right-of-way, or radio spectrum wants a piece of the Internet
  30. New Infrastructure I The result of these driving forces is

    incredible activity and growth in all the infrastructures which can deliver digital data. I Bringing us IP Dialtone is what they will all have in common. I As we speak, they’re digging up our streets and launching rockets to get to our houses. I YOU WILL BE CONNECTED. RESISTANCE IS FUTILE.
  31. Telephone Infrastructure I telephone companies are looking for more uses

    for their wires in place - 700 Million I phone companies have best assets for IP Dialtone I but regulations & business models hold them back I xDSL, ADSL - 100x faster than modems I requires all new packet switched backbone
  32. Cable Infrastructure H iii“. iii.-"3~‘=~I|~@ I-ii. H __ Ilv‘i‘i‘i‘Vl*"l.HI\ Ia‘

    - '~||"»‘ ",7 V i , . ts, ,_ . N .. , (‘wig .i , i , , _ t __ , . - .». . » , H I Cable TV goes to 70M homes in US . I Cable modems deliver up to 3OMbps in 6Mhz I Eventually, up to 2.5Gbps on cable I Shared around neighborhood I Upstream more limited, sometimes not possible
  33. Ground based Wireless Iv»--..| i» -.‘.-, . >\ 77 .

    4>‘i - . , , T-Iniv mm»i.=-ill“-»i ‘I-ii ills-»» -IiiitriVii.. "i'li‘r~~~ _ ~ , ., ».i . U . . , , I Intense, but low key, activity I From low to very high frequencies I Did you know some Coke machines speak IP? ' CDPD - IP packets on cellular telephone b/w I ATT “Angel” wireless telephone service ~ Voice today, ISDN speed Internet tomorrow I VBI on existing TV channels - 1OOKbps I “Local Multipoint Distribution System” 28Ghz band - 1Gbps reusable over metro area
  34. Satellite based Wireless V. .. ...I- .-_ -. _‘ _

    ,, _, , - .. -. ._ ‘V _ _ v_ . » , ,, .|, , \ , . . 1 »| I»,-i I-"inI»|i|ii-1»-III/~»_' ijiln-liln Ianmii I Digital Satellite Services (DSS) DirecPC, DSS-PC ~ high latency to GEO; good for multicast services - about 1Gbps today I TeleDesic - 840 LEOs “lnternet in the Sky” ~ sounds better than “l P in the sky” - low orbit means low latency, many Gbps I Many other GEO, LEO systems planned
  35. Wavelength Division Multiplexing i “- Iii.‘-iii"iii-~I‘ ‘Hi-‘.1.»Ii»~|i‘~|-iii H»-In-if‘ ii“mi-|'~.|I". -l-i-

    "n vi “I. . . ,. , ,..i. . ., ., , ,, .. , _ . .i.. i . >\ __ . , , . _ ,, ., ,1 i . . . I n The profusion of high bandwidth access technologies will require new backbone bandwidth to support them. I WDM - Wavelength Division Multiplexing - cures bandwidth bottlenecks in network backbones without the need to lay more and more fiber in the ground n Check out a company called CIENA
  36. Infrastructures I The profusion of L1, physical layer, technologies is

    good - many different niches in the physical world I Curiously, most deliver great downstream bw, but not upstream I Like Paris last century, fashionable to improve water systems in every administration - but no one ever fixed the sewers!
  37. What’s common? I These new infrastructures all seek to deliver

    the same sen/ices, so they have more common properties than unique I Consumers don’t care how the bits get there, as long as they get a service they understand - IP Dialtone. n Network operators and equipment vendors want to take advantage of new infrastructure technologies with minimal investment and disruption to their established products and services - layer 1 is a after all only a small slice of the Internet pie
  38. Protocol Technologies I Above the_phys_icaI layer, onI%3 protocol technologies WI”

    mat er - IP, thernet, and ATM IP Switching is the simplest and most effective way to tie them together I IP - The Language of Computing ' common whether between 2 or 2000000 computers I Ethernet - The_ Voice of the Computer - simple, fast, cheap, universal - Gig Ether like fast talker on old FedEx commercial ~ RJ11 of phone = RJ45 of Ethernet - Problems - privacy, cocktail parties, distance ~ The Home network will be Ethernet (wireless?)
  39. ATM ' ~' V » \ Q. ~»~~ii Wig '

    , ~—- we ~ = ’- ‘i' .~.=~i<& ~»i~-i Ii“n ‘i.''&-~“*i ea j / ilk~ In» IIlia‘, vi" Em.‘/i e » -.»-“ii, /i‘Q-i IH |'“.|"i-la I “I-i~m|.,i'&Ii.E'|iii'Q'|-E"-i/1“ n - Ii.‘-Iiamli-» I“ itIf -- i Th Q“ i3 I4 ‘g I ATM - The Radio of the Internet I ATM now in backbones of the largest lSPs I ATM is useful as a layer 2 switching and multiplexing technology, not as the grand alternate universe of protocols invented by the various standards bodies I We all enjoy radio everyday, few of us understand radio stations - its the job of the sen/ice providers I ATM invented by service providers for service providers - it provides differentiated services *easily*
  40. IP Switching Inf l|- _, t. , _ _, ,,

    ., - t , ,, V>I , G ‘V , '/ ,, ‘» ,, . , ,.,,i»,,»ii_~i~,|t.-ii."., 1|", M 1 mm in-M Tl“rm-||"~|'n““_~V.‘/,1 EMI“ m H m n IP Switching - the decision maker of the Internet, like the director in a radio station I Unlike routers, IP Switching harnesses efficient packet forwarding of ATM and Ethernet switches; Leverages what ATM does best - control & what Ethernet does best - share u Traditional IP routers makes routing & sen/ice decisions, but do it one packet at a time
  41. IP Switching I As different services proliferate, the decisions get

    harder - priority, security, speed limits/guarantees I IP Switching identifies flows of packets, packets which all receive the same treatment, allows cost of complex decisions to be spread over many packets I Flows are what the users experience, and what IP Switching controls I In essence, lP Switching efficiently provides the different sen/ices that the users expect, be they listening to CB Radio, AM News, or FM Stereo
  42. Tomorrow ’s Internet I Given the power of the driving

    forces, the competing infrastructures, and the unifying technologies, what will the sen/ice rich Internet of Tomorrow be like? I The Dial-Up paradigm will definitely be gone, with IP Dialtone in its place I You’ll be connected all the time I You use of the Internet will be immediate, casual, and dependable I Many new applications will be possible.
  43. Voice on the Internet I Today it's Internet over the

    voice network. Tomorrow it’s voice over the Internet. I Voice over IP is not a new thing, studied for 20 years I Recently become tolerable because of bandwidth in the Internet and horsepower in the PC I Easily 5x more bandwidth efficient than phone network I Won’t get far without some level of guaranteed sen/ice in the network - possible now, but barely deployed
  44. Video over IP I Video over IP, coming soon to

    a screen near you I Video conferencing, video on demand, video broadcasting are on the way - postage stamp PC windows don’t do it for me I Not yet quite tolerable in most networks - need more bandwidth, more horsepower I Nevertheless, lots of video sources on the Internet - WebCams for traic, beach conditions, etc. I Next century’s gramophone will have a new definition - quality videoconferencing with grandmom! _.____l
  45. IP Multicast I Multicast allows simultaneous delivery of information to

    multiple interested recipients - A fundamental tool for conservation of bandwidth n Not widely deployed - poorly supported by routers n Especially needed for real time, multimedia events I Tomorrow - tune into your favorite sport in the Olympics via the Internet
  46. The Secure Internet I As Firewalls and encryption become widely

    deployed, the Internet will become far more secure than the telephone network. n This fact annoys government agencies who are trying their best to slow things down so they don’t lose wiretapping capabilities. I CheckPoint is the leading vendor in this industiy. n Checkpoint’s latest products now available, fully integrated with lpsilon IP Switching. l 4e l l l
  47. Developing Countries I Each country/geography will have infrastructure that fits

    it ~ US is ahead, but for how long? - Urban will be always be better than rural u Malaysia MultiMedia Super Corridor - $2B, 200 sq mi “bandwidth park” I Singapore - fiber to every public housing project, gas stations becoming Internet Cafes I Cost to build Internet from scratch less than to build voice network from scratch - build Internet, run voice ontop
  48. The Changing ISP I This is the year that the

    Internet comes of age - it must leave its birthplace in the phone network and strike out on its own I To succeed in the Internet of tomorrow, carriers must become ISPs, and ISPs must become carriers I once dial-up is gone, there’s no longer a separate network with its own sen/ice peculiarities, between a user and their ISPs I Today’s ISPs rent bandwidth, sell connectivity; tomorrow’s ISPs build infrastructure and sell services
  49. Summary I The New Internet will be as critical and

    influential as we move forward as the phone system has been in this century. I IP Dialtone, with its network services, will enable the casual yet dependable use of the Internet by anyone, anywhere I The Internet is the most powerful tool yet for giving everyone the power of information and communication. I We are only at the beginning of the change this will have on society, both good and bad.
  50. Obstacles n There are obstacles in the path of the

    New Internet: I Regulators must stop punishing phone companies for their former monopoly. n Phone companies must stop looking to regulators to solve technology problems. 1 Standards bodies must focus on refinement, not invention. The Internet is moving too fast to be reinvented. I Users must learn to insist on Internet sen/ices that make sense, both for themselves and the sen/ice providers.
  51. Close I The way must be cleared for users to

    have IP Dialtone - the assurance of Internet services that we can love, trust, and depend on. n The Global Village is here I IP Dialtone puts your home in the global village