Latest Advances in 802.11 and Wi-Fi Wednesday, June 20, 2007 Penn Club, New York Craig Mathias, Principal, Farpoint Group.

Latest Advances in 802.11 and Wi-Fi

Wednesday, June 20, 2007Penn Club, New York

Craig Mathias, Principal, Farpoint Group

© 2007 Gerson Lehrman Group Inc., All Rights Reserved

Council Member Biography

Craig Mathias is the Principal of Farpoint Group, a Massachusetts-based advisory company specializing in wireless networking and mobile computing. Farpoint Group works with both manufacturers and end-users in technology assessment, strategy development, product specification and design, product marketing, program management, education and training, and the integration of new technologies into new and existing business operations, across a broad range of markets and applications. Previously, Mr. Mathias was the Director of Corporate Development, at Stardent Computer (formerly Stellar Computer, and now Advanced Visual Systems), where he was responsible for new product specification and corporate strategy. He was Director, Marketing, at Stellar Computer, developer of the first graphics supercomputer. Craig was also a member of the founding management team at GRiD Systems Corp., developer of the first laptop computer, where he was responsible for the development and marketing of all communications, networking, and server products. Mr. Mathias is an experienced industry and technologyanalyst, and serves on the Advisory Boards of industry conferences. Mr. Mathias has authored articles on mobile and wireless topics, and is a columnist for Computerworld, SearchMobileComputing.com, and Unstrung.com.

© 2007 Gerson Lehrman Group Inc., All Rights Reserved

Table of Contents

► How WLANs fits in the wireless landscape

► Review of the 802.11 standard and Wi-Fi specification

► Technology and market adoption trends of 802.11n

► The debate over WLAN architecture

© 2007 Gerson Lehrman Group Inc., All Rights Reserved

About GLG Institute

GLG Institute (GLGiSM) is a professional organization focused on educating business and investment professionals through in-person meetings. It is designed to revolutionize the professional education market by putting the power of programming into the hands of the GLG community.

GLGi hosts hundreds of Seminars worldwide each year.

GLGi clients receive two seats to all Seminars in all Practice Areas.

GLGi’s website enables clients to: ► Propose Seminar topics, agenda items and locations ► View and RSVP to scheduled and proposed Seminars ► Receive a daily briefing with new posts on your favorite tickers, subject

areas and from trusted Council Members ► Share Seminar details with colleagues or friends

© 2007 Gerson Lehrman Group Inc., All Rights Reserved

Gerson Lehrman Group Contacts

John AronsohnVice PresidentGerson Lehrman Group850 Third Avenue, 9th FloorNew York, NY [email protected]

Christine RuaneSenior Product ManagerGerson Lehrman Group850 Third Avenue, 9th FloorNew York, NY 10022212-984-8505 [email protected]

© 2007 Gerson Lehrman Group Inc., All Rights Reserved

IMPORTANT GLG INSTITUTE DISCLAIMER – By making contact with this/these Council Members and participating in this event, you specifically acknowledge, understand and agree that you must not seek out material non-public or confidential information from Council Members. You understand and agree that the information and material provided by Council Members is provided for your own insight and educational purposes and may not be redistributed or displayed in any form without the prior written consent of Gerson Lehrman Group. You agree to keep the material provided by Council Members for this event and the business information of Gerson Lehrman Group, including information about Council Members, confidential until such information becomes known to the public generally and except to the extent that disclosure may be required by law, regulation or legal process. You must respect any agreements they may have and understand the Council Members may be constrained by obligations or agreements in their ability to consult on certain topics and answer certain questions. Please note that Council Members do not provide investment advice, nor do they provide professional opinions. Council Members who are lawyers do not provide legal advice and no attorney-client relationship is established from their participation in this project.

You acknowledge and agree that Gerson Lehrman Group does not screen and is not responsible for the content of materials produced by Council Members. You understand and agree that you will not hold Council Members or Gerson Lehrman Group liable for the accuracy or completeness of the information provided to you by the Council Members. You acknowledge and agree that Gerson Lehrman Group shall have no liability whatsoever arising from your attendance at the event or the actions or omissions of Council Members including, but not limited to claims by third parties relating to the actions or omissions of Council Members, and you agree to release Gerson Lehrman Group from any and all claims for lost profits and liabilities that result from your participation in this event or the information provided by Council Members, regardless of whether or not such liability arises is based in tort, contract, strict liability or otherwise. You acknowledge and agree that Gerson Lehrman Group shall not be liable for any incidental, consequential, punitive or special damages, or any other indirect damages, even if advised of the possibility of such damages arising from your attendance at the event or use of the information provided at this event.

Why Wireless?

• Mobile people with mobile computers need mobile networks• Mobility = operational portability, not simply nomadic

relocateability• And only wireless can provide mobility

• Enabled by fundamental advances in technology• Radio, power, manufacturing, software, …

• Continuing high level of innovation across the board• Primary goal: Reduce, if not eliminate, the

behavioral and performance differences between wireline and wireless• Mobile broadband – triple play• And based on IP

Where Wireless Wins:Three Primary Scenarios

Where wire can’t be installed (indoor and outdoor)

Where wireless is cheaper(consider life-cycle costs)

Where mobilityis a factor – clients and infrastructure

If at least one of these conditions is not satisfied, then use wire!

Wireless Replaces Wire?• Increasingly, yes!• Broadband implies not just more throughput, but also

support for time-bounded traffic• Throughput is a function of bandwidth (directly related to spectrum)• Time-boundedness is a function of capacity• Making wireless behave like wire

• Replacement?• Wireless LANs – absolutely• Wireless MANs and WANs – we’re getting there• Improvements in mobile devices (subscriber units) and operating

models

• Mobile enterprises have competitive and productivity advantages

• But radio can be, um, difficult…

Convergence…

Consumer Demand

Standards

Technology

RegulatoryProducts/Services

Smart engineers;VLSI (design and

mfg tools)

Spectrum andother public

policy

Standards bodiesTrade associations

Business consumersConsumer consumers

Source: Farpoint Group

WPANs WLANs WMANs WWANs

UWB802.15.3c

Metro-ScaleWi-Fi

Mobile WiMAX802.20

IrDABluetooth

802.11/Wi-Fi

ProprietaryWiMAX

CDMA2000GSM/UMTS

SensorNetworks

RFID

SynchronizationSerial/USBHeadsets

Location/Tracking

BridgingAccess

DistributionBypass

Backhaul

VoiceData

Location/TrackingTelemetry

VoiceData

MessagingMedia

802.15.4,4aZigbee

SensingTelemetryMonitoring

Control

Near-FieldCommuni-

cations

<10M1-4 Mbps

<100M20-100+ Mbps

KM<75 Mbps

KM100 Kbps-14.4 Mbps

<10M<500 Mbps

<100M<54 Mbps

KM<75 Mbps

The Big Picture…

The Most Important Wireless Technology of All: MIMO• Multiple-Input, Multiple-Output

• What’s being “inputted” and “outputted” is the radio channel itself• More than a multiple-antenna technology – significant

signal processing• More than simple antenna diversity and maximal-ratio combining• Spatial multiplexing and beamforming

• Depends upon multipath!• Initial application in WLANs – 802.11n

• Drop-in replacement (PC Card, Mini-PCI, ExpressCard, MiniCard)

• Eventual application in cellular and metro (e.g. WiMAX Wave 2, UMB, LTE)

• MISO and SIMO also possible• Broadly influential – WiMAX and cellular• Not related to ultra-wideband (UWB)

Transmitter Receiver

MIMOSignalProcessing(DSP+RF)

MIMOSignalProcessing(RF+DSP)

Multipath

Reflecting Object

Source: Farpoint Group

MIMO Example – 2T, 3R (2x3)

Introduction to Wireless LANs (WLANs)

• Identical in function to a wired LAN• Began (1985-1991) as a low-throughput

data radio for telemetry and data collection applications• Good tolerance for latency• Good range

• Broad horizontal market has materialized with the 802.11 standard• Augmented (to say the least!) by Wi-Fi

• Growing towards ubiquity – no end in sight• The default LAN access vehicle

• Clients are free; infrastructure is cheap

WLANs – Constituencies and Core Applications

The Residence

The EnterprisePublic Spaces

SOHO/Small Business

Source: Farpoint Group

Benefits of Standards

• “Warm, fuzzy feeling” – complex technology becomes accessible to the technically unsophisticated

• Lower costs – thanks to VLSI, manufacturing economies of scale, and competition

• Interoperability – sometimes• Required vs. optional specs• Differing interpretations• “Standards Plus”• Trade associations (e.g., Wi-Fi Alliance, WiMAX

Forum, WiMedia Alliance, 3GPP, 3GPP2) make up the difference

The IEEE 802.11 Standard

• Work began in 1991, initial 1-2 Mbps standard in 1997• .11b/.11a in 1999• .11g in 2003

• Specifies single MAC/multiple PHYs• Compatible with 802.2 logical link control (LLC) and

802.1 upper-layer standards• Extensibility

• MAC includes power management, security, time-bounded performance, fragmentation – but airlink (Layer 2) only

• Unified WLAN technology and obsoleted other efforts (WLIF, HomeRF, HIPERLAN)

• Many required and optional features; complex• Does not specify interoperability testing or compatibility

802.11 and the Protocol Stack

Application

Presentation

Session

Transport

Network

DataLink

Physical

NetworkOperatingSystem(NOS)

802.11

Logical Link Control(LLC) - 802.2

Medium Access Control (MAC) –Power Management, Security, etc.

FH, DS, IR, CCK (b), OFDM (a),OFDM (etc.) (g)

Source: Farpoint Group

IEEE 802.11 – Current Activities

Task Group

Type Description Status

Original Both Original standard (97) – 1 and 2 Mbps, 2.4 GHz., FH, DS, and IR (IR never implemented)

Complete

a PHY Up to 54 Mbps in the 5+ GHz. band Complete

b PHY Up to 11 Mbps in the 2.4 GHz. band Complete

c MAC MAC-layer Bridging Complete

d PHY Additional regulatory domains for .11b Complete

e MAC Quality-of-Service (QoS) improvements Complete

F MAC Inter-Access-Point Protocol (IAPP) (Recommended Practice)

Complete

g PHY Up to 54 Mbps in the 2.4 GHz. band Complete

h MAC Dynamic Frequency Selection/Transmit Power Control (DFS/TPC) for .11a

Complete

i MAC Security enhancements Complete

j PHY 4.9-5 GHz. regulatory issues for Japan Complete

k MAC Radio Resource Measurement Working

Source: IEEE

IEEE 802.11 – Current Activities (cont’d)

Task Group

Type Description Status

m Both Maintenance and technical corrections to the original standard

Working

n Both Performance >= 100 Mbps Working

p Both Vehicular speeds (to 200 KPH, to 1000M, 5.85-5.925 GHz.). AKA Dedicated Short Range Communications (DSRC)/Wireless Access in Vehicular Environment (WAVE)

Working

r MAC “Fast Roaming” – reducing latency during handoffs

Working

s MAC Meshes (infrastructure and Client) Working

T MAC Performance Prediction - Benchmarking Working

u MAC Interworking with External Networks (handoff) Working

v MAC Network Management Working

w MAC Management Frame (and other control functions) Security

Working

y PHY Operation in the 3650-3700 MHz. band (contention-based protocols)

Working

Source: IEEE

802.11 Study Groups and Standing Committees

Group Type Description Status

DLS ? Direct Link Setup (station-to-station; extend DLS in .11e to work with non-DLS APs)

Working

QSE ? Quality of Service Enhancements Working

VHT ? Very High Throughput (1 Gbps) Working

VTS ? Video Transport Streams Working

WNG ? Wireless LAN Next Generation Standing Committee (ETSI BRAN, MMAC)

Working

Publicity N/A Publicity Standing Committee Working

Source: IEEE

The Wi-Fi Alliance

• A trade association, not a standards body• But there’s often a fine line…

• Originally the Wireless Ethernet Compatibility Alliance (WECA)

• Primary objective is to certify interoperability (but not 802.11 compliance) for members• And provide warm, fuzzy feeling for consumers

• New initiatives:• Wi-Fi Zone (hot spots)• WPA2 (802.11i) – includes AES• WMM (Wi-Fi MultiMedia) – 802.11e; optional• Joint project with CTIA on hard handoff• Simple Config• 802.11n certification• VoFi testing

WLANs – Key Trends and Opportunities

MIMO

DenseDeployments

UnifiedArchitecturesRFSM/SA

Managementand Assurance

Scalability

VoFi/Convergence

Location/Tracking

IEEE 802.11nTo 300+ Mbps

StructuredDistributionSystems, Capacity,Wireline Unification

Wired andWireless

ConfigurationPoliciesMonitoringControl

Signal StrengthTime Difference of Arrival (TDOA)

RF SpectrumManagement/SpectrumAssurance

GrowthChange

Voice overIP over Wi-Fi(and video)

Evolvingand

FutureWLANs

Security

Source: Farpoint Group

802.11n• The next-generation WLAN PHY (and MAC)• Draft 2.0 standard approved, Final in 1Q09

• Over 12,000 comments received on draft 1!• Over 3,000 comments on Draft 2! (Draft 3 underway)

• Some “pre-n”/”Draft n” products available now• Some will likely be software/firmware upgradeable• Excellent performance – throughput and range – “better

g than g”• Interim Wi-Fi Alliance spec – 2Q07

• “Real” products in 2H2007• We buy Wi-Fi, not 802.11

• Gradual phase-out of 2.4 GHz. (b/g) products in enterprises• Use “triple-mode” (a/b/g) clients in the interim• .11n will work best in the 5 GHz. bands

Key Features of 802.11n

• More than 500 pages of additions and changes to the 802.11 standard itself (Draft 2.0)

• 6.5 - 600 Mbps raw throughput• Based on MIMO/OFDM

• 1x1 to 4x4, with 2x3 most common initially• Backwards compatible to .11a/g• 20/40 MHz. channels

• More spectrum = more throughput• Channel bonding used in some .11g/a products• Possible interference in 2.5 GHz. band• Only one 40-MHz. channel at 2.4 GHz.

• 400 and 800 ns. Guard Intervals• PHY-layer rate adaptation – 77 modulation and coding schemes

(MCSes) indicies• To 600 Mbps using 40 MHz. channel, 4x4 MIMO, 5/6 coding rate, 108

data subcarriers, 2160 data bits per symbol• Backwards compatibility and protection

• But there is no magic here…• Keep in mind – there’s a big difference between the standard and

specific implementations…

802.11n Usage Models

• Greenfield vs. mixed mode (legacy 802.11 a/g and 802.11n)• Channel assignments will be critical• 802.11n at 5 GHz.

• The effects on the wired corporate LAN• Two words: Gigabit Ethernet

• The promise of wireless triple-play networks• Throughput, capacity, coverage, reliability

• Integration with future WWAN technologies - WiMAX and cellular• Convergence is essential to the future of WWANs• WWAN for coverage, WLAN for capacity

Will Wireless Replace Wire?

• Convenience• Cost-effectiveness• Guest access• Location and tracking• Mobility – all applications• Limited spectrum• Possible interference

• Bandwidth multiplication• Interconnect• Backhaul• Servers• Highest possible

performance (Gig-E)• Stationary equipment

(printers, scanners, storage, etc.)

Commonality in higher-level services

802.11n Market and User Expectations

• All major WLAN chip vendors are producing .11n chipsets

• Residential products very soon, enterprise products appearing now

• .11n clients as standard issue in notebooks• .11n clients in handsets by 2009• Replacement of .11b/g/a by 2009• Phenomenal growth and acceptance

• No remaining barriers for the market or users• The WLAN of choice for some time to come

Architectural Evolution

• The traditional AP does not make sense in the enterprise• Too much state; too much management

• A totally “thin” AP may not be optimal, either…• All traffic must go through the switch/controller

• The ideal solution may be to locate management and control planes in a server or appliance, and implement the data plane in the AP• Some layer-2 security in the AP as well

• Distinct wireless switches will disappear as a category• Unified architectures are coming

Unified (W)LANs - Background

Enhancer/Completer(“Gateway”)

ManagementAppliance (optional)

RFProbes

Wireless LANSwitch (Layer 3)

“Thin”AccessPoints

IP

Centralized(“Switched”)

CiscoWorks (WLSE Express)

SecurityMobilityManagement

Airespace (Cisco)ArubaMeruSiemensTrapeze

Source: Farpoint Group

The Rise of the Unified Network

Access Points

WLAN Controllers

Location Appliances

WLAN Management

Switches

Routers

LAN Management

WiredWireless

Source: Farpoint Group

Convergence• Really two opportunities

• FMC – integrating wireline and wireless• PBX integration at a minimum• Likely to reduce landlines to just backhaul/interconnect

• MMC (Mobile/Mobile) – Integrating cellular and Wi-Fi (and possibly others, like UWB, Bluetooth, WiMAX, …)

• Multiple PHYs – no single radio can do it all• Hard handoff

• Possible scenarios:• Wireline carriers take the lead – MVNOs• Wireless carriers see their destiny – core-network technology

• IMS, UMA

• The enterprise does it a la IP telephony (client and gateway, management software)

• Critical to the future of wireless operators• Locking in the enterprise• Displacing incumbent wireline providers

Enterprise-Centric Convergence

• A “make vs. buy” decision• Capital equipment – that’s more cost-effective than ever

• Keeps the enterprise in control• Carrier-independent• Security• Management, monitoring, and control• Application deployment strategy• Enables simultaneous voice and data

• Available now – will be the most important option this year

• Issues• Client devices• Client software• The business case

The Desktop of the Future…

• Looks like the desktop of the 1920s!• An AC outlet, and nothing else!• Lots of charging stands/batteries required…

• Automatic handoff between wide-area voice (CDMA/GSM) and VoFi• And data services as well

• Fixed/Mobile and Mobile/Mobile convergence

• An opportunity for cellular operators to:• Enhance their capacity in high-traffic areas

without more expensive spectrum• Displace wireline carriers in the enterprise

VoFi – Everywhere, All the Time

PublicSpace

Enterprise

WWAN

WLAN

Source: Farpoint Group

Two-way handoffof both voice anddata connections…

Gateway/IP Centrex

The Future of WLANs

• Established technology, products, and markets – here to stay• The LAN itself isn’t going anywhere…• No real challenge from WPANs, WMANs, or WWANs

• Much higher throughput – to perhaps 500+ Mbps• And eventually 1 Gbps

• Major role for voice• Position/location tracking as standard feature• Great integration into both wired infrastructure and

enterprise management• Becomes, for many if not most, the default network

connection – essentially everywhere!