Icuwb 2013 keynote uwb past and future - lansford - final

UWB – How did we get here, and where are we going?

Jim Lansford, Ph.D. Fellow, Global Standards and Adjunct Professor, University of Colorado at Boulder 18 September 2013

2

• Why am I here?

• The Gartner Hype Cycle

• What were the problems we thought we were solving?

• What Happened?

• The FCC R&O of February 14, 2002

• The Standards Fiasco

• The Post-standards fiasco

• High speed UWB Enters the Trough of Disillusionment

• Recent Efforts to Revive it

• Where do we go from here?

Agenda

3

• CTO of Alereon (2003-2010)

− UWB Chip company

− Spinout of Time Domain

− Raised over $80M

• Co-Chair of 802.15.3a

− Created to develop a high speed UWB standard

− Disbanded in January, 2006 (more on that later)

• Served on numerous committees in WiMedia

− Currently board member from CSR (!)

− Chair, Technical Steering Committee

− Member, Certification Review Board

− Chair, Mobile Applications Study Group

− Vice-Chair, Video Study Group

• Lots of regulatory meetings about UWB

• Lots of customer meetings about UWB

Why am I here?

4

First - The Gartner Hype Cycle

5

• Three usage models drove the need for high speed P2P WPAN:

− Wireless Docking

− Video streaming from mobile to remote display

• Wi-Fi Display (“Miracast”) is an example

− Content sync (“sideloading”) of photos, video, music, etc.

• There are other use cases for UWB that have been cited:

− Precision location

− Sensors

− Imaging

• The high speed use cases stirred up the most hype

What were/are the use cases for high speed WPAN?

6

Why high speed P2P WPAN? (1)

• If traditional 2.4/5GHz WLAN bands are used for high speed WPAN, these usage models, especially video streaming, will overload existing WLAN spatial capacity

HD video streaming today

4k/UHD and 3D in the future

Sync

These applications require high bandwidth and

energy efficiency

Android “Party Mode”, AllShare, GroupCast,

ShareShot

7

• Wireless Docking allows a portable/mobile device to treat a group of devices as a cluster

Why high speed P2P WPAN? (2)

Wireless

Docking

Host

8

• These use cases are still needed in the market

−The Wi-Fi Alliance is working on them for use with 802.11ac/ad

• UWB failed to get into the market to solve these problems

−Was it a solution looking for a problem?

• How did UWB miss the market? Was it:

−Regulatory?

−Standards?

−Market makers?

−Market launch?

YES

What happened?

9

• Before the February 14, 2002 R&O from the FCC:

• Lots of hype from Time Domain and Xtreme Spectrum about the magical properties of UWB

− “Gigabits per second at kilometers of distance using microwatts of power”

− “Would replace cellular and Wi-Fi”

− And so on…it was going to be a multi-billion dollar market

• Millions of dollars were spent on regulatory lobbying

• Battle lines were drawn for a religious war:

− Impulse radio (IR)

− Direct Sequence UWB (DS-UWB)

− Multiband OFDM (MB-OFDM)

− Continuous pulsed UWB (C-UWB)

History 101 – The beginning of the hype 1/5

10

• On February 14, 2002, the FCC issued a Report and Order in response to NPRM 98-153

• Became part of the Part 15 rules for unlicensed devices: 15.501- 15.523

• This was very controversial in the US and the rest of the world

− “Spectral Underlay”

History 101 – The FCC R&O 2/5

IEE

E d

ocu

men

t 802.1

5-0

2/1

33r1

is a

go

od

su

mm

ary

11

• IEEE 802.15.3a was created in November 2001 to develop a high speed UWB PHY and MAC

• There were two main camps:

− DS-UWB (Xtreme Spectrum/Freescale)

− MB-OFDM (WiMedia – Intel, Broadcom, CSR, startups, et al)

• Under IEEE rules, a proposed draft must be approved by 75% of the voting members

• After four years of warfare with no compromise in sight, 802.15.3a was disbanded in January 2006

History 101 – The Standards Fiasco 3/5

12

• Freescale killed the DS-UWB program shortly after the IEEE 802.15.3a group disbanded

• The MB-OFDM group continued on as WiMedia

− Created ECMA standards for PHY & MAC (ECMA-368), and MAC-PHY interface (ECMA-369)

• PHY & MAC: http://www.ecma-international.org/publications/standards/Ecma-368.htm

• MAC-PHY Interface: http://www.ecma-international.org/publications/standards/Ecma-369.htm

− Created a certification program

− Worked with USB-IF to create the Certified Wireless USB program

• Initial products performed poorly

− All initial products were WUSB

− Even though the PHY could operate at up to 480Mbps, initial WUSB devices could only deliver 50-100Mbps

• Intel pulled out, followed by most other large companies

• Focus on the use cases shifted to 802.11ac/ad

History 101 – The Post-Standards Fiasco 4/5

13

• After the withdrawal of market makers from WiMedia,

the high speed UWB world entered the “Trough of

Disillusionment” − Aka the “Slough of Despond”

− Without market makers, WiMedia and WUSB momentum rapidly

evaporated

• After spending approximately $500m in venture capital,

the startups began fading away, as the market failed to

develop − Dead: WiQuest, Tzero, Artimi/Staccato, Wisair, Pulse-Link,

Radiospire

− Still alive: Alereon, Decawave

− WiMedia is still (technically) alive

• We waged a fight to the death over a “zero billion dollar

market”

• For high speed UWB to exit the Trough of

Disillusionment, a forcing function is needed

History 101 – The Trough of Disillusionment 5/5

14

• Discussion in IEEE and elsewhere about using 6-10GHz UWB for P2P connectivity using an 802.11 MAC – “802.11ac+”

− See references to IEEE 802.11 documents 11/743, 12/96, 12/375, 12/653, 12/929, 12/1143

− 6-10GHz would be used for short range, high speed peer to peer communications

• Wi-Fi Direct, Wi-Fi Display (Miracast), Wireless docking

− Can share the same antenna and much of the RF circuitry with 802.11ac

− Propagation is similar to 5GHz

− Allows higher power 5-6GHz channels to be used for longer range infrastructure connections

− Additional Tx power (~10dB) in the 6-10GHz band can enable the critical use cases

• Need >1Gbps at approx 3 meters NLOS using SISO

• Existing -41.3dBm/MHz rules don’t quite get us there

What has happened since?

35m

m

6mm

3-10GHz

antenna

15

US Feb-02

EU Dec-06

Japan Sep-2006

Canada March-2009

3.168 GHz 10.560 6.336 4.752 7.920 9.504 7.392 8.976

Korea Sep-2006

China

Jan-09

6-10GHz

= not allowed = allowed with DAA = allowed

The UWB industry did a LOT of heavy lifting – we

now have spectrum that isn’t being used.

16

• This is a major application for UWB – it is still needed in the market

• Being filled by 802.11ac and 802.11ad

Use cases revisited….

Wireless

Docking

Host

17

• Use case: docking in a “cube farm”

• Needs very high spatial capacity

• Short range is an advantage

− As long as link margin is adequate for NLOS shadowing

• Should not interfere with AP (infrastructure)

High density P2P deployments

This illustrates what these regulations were targeted to promote:

Short range, very high speed communications…this use case is

even more important now than it was in 2002

18

The market need still exists

• UWB is still a great technology for mobile platforms

− Fills market need for very high speed PAN/cable replacement

− High speed - video streaming and synchronization

− Very high spatial capacity

− Very low energy per bit – optimizes battery life

− High precision location

− Mobile platforms would only want to use UWB above 6GHz

• 3-5GHz bands will be needed for other applications

• Many companies have proven the technology

− UWB technology can work well

− Needs additional transmit power above 6GHz to compensate for path losses and provide additional range at higher speeds (10dB ideal)

UWB meets market need for high BW WPAN for mobile…and

avoids interference between P2P and AP networks

19

• Need 10dB additional power in the 6-10GHz bands

− Only way to convince market makers to get back into the game

− Will require more regulatory lobbying

• Need to use the 802.11ac MAC with a UWB PHY

− Puts UWB in the wildly successful Wi-Fi ecosystem

− Proven certification and testing methodology

− About 1.2 billion Wi-Fi chipsets will be shipped in 2013 (ABI Research)

• PHY needs to be an extension of 802.11ac

• UWB will likely remain a niche technology otherwise

− Sensors and positioning

• Questions? Comments?

Where do we go from here?

20

• IEEE 802.11 presentations:

− https://mentor.ieee.org/802.11/dcn/11/11-11-0743-00-0wng-6-9ghz-extensions-to-802-11.ppt

− https://mentor.ieee.org/802.11/dcn/12/11-12-0096-00-0wng-6-10ghz-extensions-to-802-11ac-part2.ppt

− https://mentor.ieee.org/802.11/dcn/12/11-12-0375-00-0wng-6-10ghz-extensions-to-802-11ac-part3.ppt

− https://mentor.ieee.org/802.11/dcn/12/11-12-0653-00-0wng-compatibility-of-6-10ghz-extensions-with-the-802-11ac-phy.ppt

− https://mentor.ieee.org/802.11/dcn/12/11-12-0929-00-0wng-expansion-of-802-11ac-to-6-10ghz.ppt

− https://mentor.ieee.org/802.11/dcn/12/11-12-1143-00-0wng-6-10ghz-rate-range-and-link-budget.ppt

References