beyond third generation telecommunications architectures

1

Beyond Third GenerationBeyond Third GenerationTelecommunications Architectures:Telecommunications Architectures:

The Convergence of InternetThe Convergence of InternetTechnology and Cellular TelephonyTechnology and Cellular Telephony

Prof. Randy H. Katz

EECS DepartmentUniversity of California, Berkeley

Berkeley, CA [email protected]

http://www.cs.Berkeley.edu/~randy

2

Presentation OutlinePresentation Outline

• Comparison of Telecomm & Data Comm Industries• Voice-centric versus Data-centric Viewpoint

• Internet versus Telephone Technology• Implications Beyond the Third Generation

3




4

Global Telecommunications TradeGlobal Telecommunications Trade

0

20

40

60

80

100

120

1990 1991 1992 1993 1994 1995 1996

OtherNe t Se ttlementsEquipment

Source: ITU (The Economist, 8 March 1997)

$ Billions Total Revenue1996: $867 Billion2000: $1270 Billion

5

Cellular Subscriber Growth in USCellular Subscriber Growth in US

85 87 89 91 93 95

0

10000

20000

30000

40000

50000

44 million

Source: CTIA Web Page

Cellular telephonyremains the hottest andfastest growing sector of the telecomm market

Nordic Countries: 10 mobilephones being added for

each wireline phone!

6

Growth in Cell Sites in USGrowth in Cell Sites in US

85 87 89 91 93 95

0

5000

10000

15000

20000

25000

30000

35000

30000

Source: CTIA Web Page

25% growthin one yearrepresents

major marketfor infrastructure

vendors

7

Mobile Telephone and Internet UsersMobile Telephone and Internet Users

0

100

200

300

400

500

600

1993 1994 1995 1996 1997 1998 1999 2000 2001

Source: Ericsson Radio Systems, Inc.

Mobile TelephoneUsers

Internet Users

Millions

Year

8

World’s Cellular SubscribersWorld’s Cellular Subscribers

0

100

200

300

400

500

600

1993 1994 1995 1996 1997 1998 1999 2000 2001

Millions

Year

Digital

Analo g

Source: Ericsson Radio Systems, Inc.

Will providea ubiquitousinfrastructurefor wirelessdata as well

as voice

9

Note: surveys consistently underestimategrowth in cellular subscriber base

1994 1996 1998 2000 2002 20040

102030405060708090

100

1994 1996 1998 2000 2002 2004

PCSCellular+PCS

Predicted Continued Growth ofPredicted Continued Growth ofWireless TelephonyWireless Telephony

Millions ofSubscribers

Source: Yankee Group Forecast, 1996

Slow growthin USA in PCS

digital svrcs

10

Cellular Phone Growth:Cellular Phone Growth:An International PhenomenomAn International Phenomenom

1990 1995 20000

5

10

15

20

25

30

35

40

45

1990 1995 2000

EuropeUnites StatesJapan

% of main linesthat are mobilephones

Source: Economist, 4 May 1996

By Year 2000:– One in three telephones will be mobile– Mobility becomes a lifestyle

11

Telecomms Service Revenue GrowthTelecomms Service Revenue Growth

100

200

300

400

500

600

700

800

900

90 91 92 93 94 95 96

MobileInternational

Source: ITU (the Economist, 13 Sept 97)

Total Revenue 1996: $670 Billion

20% Mobile10% Int’l60% Local10% Other

12

Services Most Often RequestedServices Most Often Requested

• Call Forwarding 37%

• Paging 33%• Internet/E-Mail 24%

• Traffic/Weather 15%• Conference Calling 13%• News 3%

DataApplications

Source: CTIA Web PagePeter D. Hart Research Associates, March 1997

After basic wireless telephony service

13

Global Markets for Portable ComputersGlobal Markets for Portable Computers

90 91 92 93 94 95 96 97 98 99

2000

0

10

20

30

40

50

60

90 91 92 93 94 95 96 97 98 99

2000

$Bil

Source: Arthur D. Little, Industry Estimatesin the Economist Magazine, 14 May 94

Year

Most rapid growing sector of the PC marketNokia: 20-30% of revenue from data services

14

Data Giants vs. Telecomm GiantsData Giants vs. Telecomm Giants

• CISCO– Market Cap: $46 Billion– Annual Sales: $6 Billion

• 3COM/US Robotics– Market Cap: $16.1 Billion– Annual Sales: $5.5 Billion

• Ascend/Cascade– Market Cap: $7.7 Billion– Annual Sales: $1.3 Billion

• Newbridge– Market Cap: $7.4 Billion– Annual Sales: $0.9 Billion

• About $29 Billion inrevenue; growing to $72Billion by 2000

• AT&T– Market Cap: $57 Billion– Annual Sales: $52 Billion

• LUCENT– Market Cap: $48 Billion– Annual Sales: $24 Billion

• GTE– Market Cap: $42.3 Billion– Annual Sales: $21.7 Billion

• Bell Atlantic– Market Cap: $33.3 Billion– Annual Sales: $13.3 Billion

• About $860 Billion inTOTAL revenue; growingto $1272 Billion by 2000

15




16

Voice-Centered vs.Voice-Centered vs.Data-Centered?Data-Centered?

• The dramatic rise of the Internet and theWorld Wide Web– Reuters

» 82 million PCs on Internet today» Growing to 286 million by Year 2001

– Internet Demographic Survey (CommerceNet)» 55 million Internet Users in US & Canada» 100 million worldwide

– Telephone Subscribers» 1 billion worldwide» 180 million mobile phone users worldwide

17


• More than 50% of telecomm traffic in BayArea is already data, not voice– Internet FIND/SVP Survey (Nov/Dec 95)

» 25% of Internet users make fewer long distance calls» 32% watch less television

• Telephone switching infrastructure has beenbrought to its knees– Design metrics based on voice quality, not data

throughput– Design for short duration voice conversations,

not long duration computer sessions

18


Bell Atlantic

US West

Pacific Bell

SBC

Average Peak-HourLocal Circuit Use

Average ResidentialCall Length

All Circuits Circuits to ISPs All Circuits Circuits to ISPs

5.0

5.0

6.7

6.7

4-5

2-4

3.8

na

43-47

45.0

31.7

52.3

17.7

14

20.8

na

Source: FCC (The Economist, 13 Sept 97)

Network Usage, Minutes

19

Will Trend Towards Data-Will Trend Towards Data-Centric Accelerate?Centric Accelerate?

• Sir William Preece, Chief of the British PostalSystem, 1876:

“The Americans may have need of the telephone, butwe do not. We have plenty of messenger boys.”

• Analogy with the Post Office– Telephones have largely replaced the personal letter– Posts used mainly for business-oriented

correspondence: documents (Fed-Express), bills(direct debit/e-commerce), advertising (WWW),delivery of merchandise (UPS)

– Email reduces number of telephone conversations– What will be the effect of the Internet/WWW?

20

Will Trend Towards Data-Will Trend Towards Data-Centered Accelerate?Centered Accelerate?

• Many calls to obtain information are already beingreplaced by the WWW– E.g., Ordering Books– E.g., Ordering CDs– E.g., Fyte Trax Service– E.g., Package Tracking– E.g., Booking Flights– and so on

21

Efficiencies of InteractionEfficiencies of Interaction

Source: McKinsey (The Economist, 13 Sept 97)

Search: Finding a High-Rate Certificate of DepositTelephoneWWWWWW w/ agent

Co-ordination: Reordering an Inventory ItemMailE-mailEDI

Monitoring: Updating an Equity PortfolioNewspaperWWWWWW w/ agent

25.0

10.0

1.0

Min

-60%

-90%

3.7

1.6

0.3

-57%

-81%

5.1

1.8

0.5

-65%

-72%

22




23

What is the Internet?What is the Internet?

“Internet” refers to the global information system that -- (i) islogically linked together by a globally unique addressspace based on the Internet Protocol (IP) or itssubsequent extensions/follow-ons; (ii) is able to supportcommunications using the Transmission ControlProtocol/Internet Protocol (TCP/IP) Suite or itssubsequent extensions/follow-ons, and/or other IP-compatible protocols; and (iii) provides, uses or makesaccessible, either publicly or privately, high level serviceslayered on the communications and related infrastructuredescribed herein.

Federal Networking Council Resolution, 24 October 1995

24

Today’s Internet TechnologyToday’s Internet Technology

• Strengths– Intelligence at the end

points; No state in thenetwork;

– Highly decentralized control– Enables operation over very

heterogeneous collection ofaccess technologies; fewassumptions about thenetwork necessary

– Achieves robustcommunications throughpacket switching & store-and-forward routing

– Depends on cooperativeforwarding of packets

• Weaknesses– No differential service– No control mechanisms for

managing bottleneck links– Store-and-forward routing

introduces variable delay inend-to-end performance

– Decentralized controlmakes introduction of newprotocols/functions difficultsince all end nodes must beupgraded

– Lack of truly trustedinfrastructure leads tosecurity problems

25

Today’s Telephone NetworkToday’s Telephone Network

• Strengths– Requires no end-point

intelligence; supportsheterogeneous end devices

– Provides excellentperformance for voice

– End-to-end performanceguarantees achievedthrough well-definedsignaling layer to switchingfunction

– True utility functionalitythrough sophisticated andhierarchically arrangedswitches controlled byservice providers

• Weaknesses– Achieves performance by

overallocating resources– 3.4 KHz audio voice band

signal converted to 64 kbpsdigital representation

– Switching designdetermined by statistics ofcall traffic

– Difficult to add new servicesto the so-called “IntelligentNetwork” due to complexfeature interaction

– Expensive approach torobustness

26

ATM: The Grand Convergence?ATM: The Grand Convergence?

• Strengths– Virtual circuits with call set-

up to manage scarceresources and achieve QoSguarantees

– Fixed/small size “cells” toenable fast switching

– Sophisticated statisticalmultiplexing mechanisms tomake possible variety oftraffic models

– Integrated services

• Weaknesses– Connection-orientation has

some problems with latencyand robust operation; everycell must follow same path inorder

– ATM unlikely to be a universalend-to-end technology,especially for data traffic inlocal area

– Quaranteed performance end-to-end in heterogeneousenvironments is lost

27

Next Generation InternetNext Generation Internet

• “Integrated Services Packet Network (ISPN)”• Ubiquitous support for multipoint-to-multipoint

multicast communications• Built-in support for mobility and mobile route

optimization

• Resource allocation mechanisms based on RSVPsignaling– Performance promises rather than guarantees– Receivers initiate signaling; nice scaling properties– Soft state in the network allows robust recovery to

failure; protocol works around link and switch failures

28


• Microprocessor performance/software algorithmsnow sufficient for real-time encode/decode ofvideo and audio– Traditional telephony hardware operates at 64 kbps for

PCM coding– Mbone software audio coding at many rates

» 36 kbps ADPCM» 17 kbps GSM» 9 kbps LPC

Adequate video at 28.8 to 128 kbps» Scalable codecs» Layered video

29


• Real Time Protocol (RTP)– Application Level Framing– End nodes adapt audio/video streaming rates to what

the network can support

• Easy integration of new services like proxies– Hardware/software is not specialized– Easy to integrate distributed applications

• Solve performance problems by adding morebandwidth

30

Internet TelephonyInternet Telephony

Local Call Local CallInternet

SF to Frankfurt via Internet Service: $0.28 per min via AT&T Long Distance: $1.25 per min

Analog Voice toPacket Data

Packet Data toAnalog Voice

Source: G-Cubed

Gateway Gateway

Why so Cheap?

Less expensive infrastructureCircumvents government-backed monopoliesExisting long distance tariffs far exceed costsWTO worldwide deregulation

31

Internet TelephonyInternet Telephony

• Quality Issues: High Latencies, Dropped Packets– Solutions

» Deployment of private networks» Faster and scalable hardware reduces gateway latency» RSVP + H.323 + Reconstruction of lost packets + Better voice

coding at 8 kbps» VoIP: Voice over Internet Protocol Forum

• Integration of circuit-switched local infrastructurewith packet-switched wide-area infrastructure– Wide-area b/w is a commodity, not true for local access

» 1996-2000: 5X increase in Atlantic/Pacific cable capacity– Many leading telecomms already doing this

» Internet FAX services» Cheap way for RBOCs to get into long distance service

32

U.S. Internet Telephony MarketU.S. Internet Telephony Market

1998 1999 2000 2001 2002 2003 20040

0.5

1

1.5

2

2.5

3

1998 1999 2000 2001 2002 2003 2004

SpentSavings

Source: Forrester Research (The Economist, 13 Sept 97)

NewRevenues

CostSavings

4% of U.S. Telephony Revenue

33




34

Third GenerationThird GenerationTelecommunications ArchitecturesTelecommunications Architectures

High-tier

Low-tier

Satellite

High Mobility Low MobilityWide Area

Regional Area

Local Area

• FPLMTS/UMTS/IMT-2000– Universal multimedia information access with mobility

spanning residences, businesses, public/pedestrian,mobile/vehicular, national, and global regions

– Low end: 512 kbps; High end: 155 mbps??

35

“Mobile BroadbandSystems”

WirelessLocal AreaNetworks

Wireless Access TechnologiesWireless Access Technologies

Mbps

0.01

0.1

1

10

100Wired

Cellular

Cordless

“Universal MobileTelecomms Systems”

(UMTS)

60 GHz100 m range

Office orRoom

Building

Indoors

Stationary Walking

Outdoors

Vehicle

36

Beyond Third GenerationBeyond Third Generation

• 1st Generation was analog cellular– e.g., AMPS

• 2nd Generation is digital cellular– e.g., IS-54 (TDMA), IS-95 (CDMA), GSM (TDMA)

• 3rd Generation– Still being defined . . .– Will embrace multiple radio access technologies

suitable for local, wide-area, satellite, etc. and able toachieve higher bandwidth than existing airlinks

– Likely to be based on GSM wireline infrastructure

• 4th Generation– ???

37

EIRAUCHLR

Mobile Telecomm NetworkMobile Telecomm Network

BTS

MSCBSC

BSC

VLR

MSC

TSC

MS

Interexchan geNetwork (IXC)

Local ExchNetwork

Local LoopRBOCs

TransitSwitching

Center

AT&TMCI

SprintWorldCom

. . .

38

EIRAUCHLR

Data Over Analog CellularData Over Analog Cellular

MSCBSC

VLR


Local ExchNet (LEC)

TSC

Digital -- Analog (modem) --Digital (IXC) -- Analog (local loop) --

Digital (modem)Circuit Switched

39

EIRAUCHLR

Data Over Digital CellularData Over Digital Cellular

MSCBSC

VLR


Local ExchNet (LEC)

TSC

IWFDigital -- Analog (modem) --

Digital (IXC) -- Analog (local loop) --Digital (modem)Circuit Switched

9.6 kbps to100+ kbps

40

Internet

EIRAUCHLR

Internet ArchitectureInternet Architecture

MSCBSC

VLR

Local ExchNet (LEC)

GW

Digital -- Analog -- DigitalPacket Switched

Local Circuit Switched

ISDN: digital tothe home

ISP

41

Internet

EIRAUCHLR


MSCBSC

VLR

CorporateIntranetwork

GW

Fully Digital/Packet Switched

GW

42

NextGeneration

Internet


BSC

EIRAUCHLRVLR

CorporateIntranetwork

GW GW

SmartCommunicator

MSC

ProxyServers

BS

Digital CellularLink Layer

WLAN/WPBXLink Layer

HAFA

43

Summary and ConclusionsSummary and Conclusions• Accelerating importance of data access in cellular

telecommunications infrastructures: data willdominate voice applications

• Internet’s emerging capabilities for real-timetraffic, multipoint communications, broadcast-based information dissemination

• Secret of Internet’s success– Intelligence at the end points– Simple and inexpensive infrastructure components– Gain performance through bandwidth

• Economies of scale favor the Internet– Cause for a dramatic re-thinking of telecomms

infrastructure beyond the 3rd Generation

beyond third generation telecommunications architectures

Documents