Welborn, FreescaleSlide 1

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Project: IEEE 802.15 Working Group for Wireless Personal Area Networks (WPANs)Project: IEEE 802.15 Working Group for Wireless Personal Area Networks (WPANs)

Submission Title: [Recent Progress for UWB Rules]

Date Submitted: [September 2005]

Source: [Matt Welborn] Company [Freescale Semiconductor, Inc]Address [8133 Leesburg Pike Vienna, VA USA]

Voice:[703-269-3000], E-Mail:[matt.welborn @freescale.com]

Re: []

Abstract: [Report on regulatory progress.]

Purpose: [Information for TG3a members to help them make a more informed decision.]

Notice: This document has been prepared to assist the IEEE 802.15. It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein.Release: The contributor acknowledges and accepts that this contribution becomes the property of IEEE and may be made publicly available by 802.15.

Welborn, FreescaleSlide 2

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Overview

• Briefly describes Detect-And-Avoid (DAA) and the detection and regulatory issues surrounding it that need further study

• This set of slides does not address the “Avoid” piece– Rules must specify emission limits with and without DAA

Detections– The emission limits and specific avoidance mechanisms are

beyond the scope of this set of slides• DS-UWB advocates support for world-wide rules and

next steps

Welborn, FreescaleSlide 3

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Regulatory Position

• Various administrations are moving forward with draft UWB proposals

– Worldwide activities underscore the tremendous interest in and value in worldwide commercial UWB availability

• Goal: deliver safe, non-interfering, commercial silicon to customers around the world

– It appears that a global footprint in the 6 to 10 GHz band is possible

– DS-UWB advocates are committed to protecting future 4G cellular and WiMax systems that may operate within the 3.1 to 5 GHz band

• We are eager to participate with regulators in Europe and Asia to attempt to:

– Develop demonstrably affective mitigation technology

– Develop compliance tests that ensure acceptable operation

Welborn, FreescaleSlide 4

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Next Steps• Help all parties reach consensus for a world-wide UWB band

• Target immediate rules in 5.8 to 10.6 GHz bands providing – Low risk for revisions to accommodate future mobile services

– Acceptable performance for key UWB applications

– No technology risk, provide low implementation cost/power

– Acceptable TTM coinciding with UWB products & standards

• Possible additional 6 dB in the high-band– Helps to match performance in 3-5 GHz low-band

– Nokia, Ericsson, Freescale have already called for consensus

• Urge to issue rules quickly to provide an environment for UWB industry to get established

• Support with white papers and measurement procedures and results as needed

Welborn, FreescaleSlide 5

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

DAA Technology Assumptions

• Receivers do not transmit – Hence they cannot be directly detected

• DAA assumes that a receiver in the UWB unit can detect an incumbent signal and use this information to avoid transmitting on the incumbent’s frequency

– Two potential cases:• victim is a receiver (DAA detects the incoming signal)

• victim is a transceiver (DAA detects the transmitter emissions)

• DAA assumes that regulatorily useful performance metrics exist

– Probability of missed detections, false alarms, needed shut-downs etc. are at levels good enough to allow rules to be written that both protect incumbents and preserve the value of UWB

Welborn, FreescaleSlide 6

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

DAA Problems Are Outside Of Designer’s Control

• The DAA receiver in the UWB device is likely to miss a signal useful to the victim receiver because, in most cases:1. The DAA receiver cannot be made as sensitive as the victim

receiver

2. The field strength at the DAA receiver can be less than it is at the victim

3. Local RFI can cause an unacceptable level of false alarms and/or missed detections

4. The DAA device and victim have no information on each others timing

5. Unknown frequency offsets means that detecting an up link does not tell you how to avoid the downlink• A Difficulty For Transceivers

Welborn, FreescaleSlide 7

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

(1) Target DAA Receiver Sensitivity is not as Sensitive as The Victim Receiver

• The DAA antenna has less gain than victim antenna– UWB devices have very tiny antennas that fit in cell-phones, a USB memory stick, etc.– Victim receivers have full dipoles, arrays, and large dish antennas

• Lower-cost lower-performance and less-optimized LNA, receiver and detector.– Most microwave systems have expensive and exotic ultra low-noise-amplifier (LNA)– The LNA in the DAA receiver is built on a lowest cost CMOS process and integrated with

other circuitry that results in significantly more noise.– Rather than a highly optimized and sophisticated signal processing and detection

circuits used in a microwave link, the DAA receiver is again an ultra low-power, low-cost, small-size sub-optimum implementation.

– Processing gain must be high enough to compensate for lower antenna gain (20-50 dB), lower field strength (10-40 dB), and higher effective noise figure with CMOS plus RFI (~10-20 dB) --- which adds up to 40 to 110 dB for the given estimates.

– Integration does not solve all the problems or have enough gain• Incoherent integration may take too long/provide limited performance• Coherent integration takes too much sophistication• Both can produce a false alarm “Detect” event on local RFI

Welborn, FreescaleSlide 8

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

DAA and Directional Antennas

high-gain, large collection area receive dish with low-noise amplifier (reliably receives incoming signal)

DAA with small square-inch antenna and low-cost receiver (misses incoming signal)

weak signal from distant transmitter

Example: DAA unable to protect fixed microwave receiver

Welborn, FreescaleSlide 9

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

(2) Low Field Strength At DAA—Yet High At The Victim Receiver

• Multipath, shadowing, and blocking affect DAA and the victim receiver differentially – the path loss to each is different

• Directional antenna used by the victim system transmitter can give the victim receiver a stronger signal, even if the path loss were identical

Welborn, FreescaleSlide 10

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

(3) Local RFICan Render DAA Ineffective

• The DAA device is part of an RFI-generating product– Digital noise in a DVD player, TV, PC, USB-memory stick, etc. is significant– Cell-phone sensitivity is set by self-induced RFI, not the noise figure

• A DAA receiver may not differentiate between RFI and a victim signal– Both are modulated RF

• Probability of missed detections can be too high– If the threshold is set so that RFI does not cause false alarms, Then the DAA

device will likely miss the victim signal– The faint signal being received by a victim receiver can be masked by RFI

• Probability of false alarms can be too high– If the threshold is set so that the victim signal can be detected, then the false

alarm rate can be too high– Due to the broadband nature of RFI, the entire band could be blocked– long-term or frequent un-needed UWB performance reductions are not a

success

• The net result is that RFI can render DAA ineffective– even with other issues solved

Welborn, FreescaleSlide 11

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

(4) No Cooperative Timing— DAA Device And Victim Have No Timing Information For Each

Other

• Timing rules are critical, yet they must work for current and future systems

– If it detects something, How long must DAA listen before enabling transmissions again?

– How long can a packet be in order for the victim to recover if it needs to receive a transmission that began just after the DAA device started transmitting?

– How often must DAA stop to listen again?

• DAA is not a closed system—unlike, for example, CSMA– All relevant parameters are not known in advance and carefully set

up to allow the UWB and victim systems operate effectively together.

Welborn, FreescaleSlide 12

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

(5) Unknown Frequency Offsets—A Problem For Transceivers

• Theory: Monitoring transmitter emissions solves the receiver protection problem

– BUT… The DAA is detecting the wrong thing—when the transceiver is transmitting, not when it is receiving a transmission from elsewhere.

• Two transceiver cases:– Frequency division duplex:

• Detecting the transmit frequency does not indicate what receive frequency to protect

– Frequency-offsets vary with service, frequency, and country– Offset is unknown given a generic “future proof” DAA.

• Differential path-loss, timing, and sensitivity issues remain– Time division duplex:

• Timing rules are critical, same issues as previous slide• Does the DAA device only get to transmit when the victim is

transmitting? If so, how low of a duty cycle is acceptable?

Welborn, FreescaleSlide 13

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Status So Far—Attempts At DAA Have Not Succeeded

• DFS  (dynamic frequency selection) was introduced to allow sharing of the bands 5.25-5.35 GHz and 5.470-5.725 GHz used by primary radar and satellite services

• DFS was first proposed, in Europe (just like DAA), as a means for RLANs to share frequency bands in 1997

• CEPT adopted a decision in 1999 that called it out as a requirement (29 November 1999 (ERC/DEC/(99)23)).

• Further study lead to Recommendations from the World Radio Conference WRC-2003, and new regulatory decisions in many countries calling for DFS

• Today, years after the 11a standard, DFS has cost many man years of effort by both administrations and companies resulting in:

– No commercial deployments or even demonstrations of workable systems– No development of a means of verification or testing that it is satisfactory to the

incumbents and administrations. – No economic return, yet hope that somewhere in the future someone will figure

out how to make it work.

• Luckily for 802.11a, channels 5.15-5.25 GHz are available without DFS worldwide allowing deployment to start, but the use of the additional channels is still pending proof of DFS as a detect-and-avoid system.

Welborn, FreescaleSlide 14

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

DAA—Not Ready For Regulatory Use

• The prerequisites for effective regulations do not exist– no performance metrics

(e.g. probability of missed detections, false alarms, needed shut-downs)

– no criteria for adequate performance• criteria must address both protecting incumbents and preserving the

value of UWB– long-term or frequent un-needed UWB performance reductions are not a

success– A base station shutting down UWB in a 10 mile radius, yet with only a few

operational receivers, is not a success.

– no proven test methods(i.e., procedures that enforce regulatory performance specifications)

– No demonstrable solutions to validate any of the above• trials must be complete and include the real-life

statistics of all problem classes

Welborn, FreescaleSlide 15

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Conclusions• DAA requires significant more study as it may never work• DAA has not been demonstrated and appears that it cannot work as

advertised due to the difficulty of simultaneously solving these classes of problems:

– The DAA receiver is less sensitive than victim receiver– The field strength at the DAA receiver can be less than it is at the victim– RFI local to the DAA device can render DAA ineffective – The DAA device and victim do not cooperate on timing, yet timing is

critical– Unknown frequency offsets—a problem for transceivers– Compliance test must be specific pass-fail test that allows product

shipments– Passing must both protect incumbents and preserve the value of UWB

• DAA is not ready for regulations:– no performance metrics & no criteria for adequate performance– no test methods– no demonstration trials on which to base regulations—no real-life

statistics• trials must be complete and include the reality of all problem

classes

Welborn, FreescaleSlide 16

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Regulatory Position

• Various administrations are moving forward with draft UWB proposals

– Worldwide activities underscore the tremendous interest in and value in worldwide commercial UWB availability

• Goal: deliver safe, non-interfering, commercial silicon to customers around the world

– It appears that a global footprint in the 6 to 10 GHz band is possible

– DS-UWB advocates are committed to protecting future 4G cellular and WiMax systems that may operate within the 3.1 to 5 GHz band

• We are eager to participate with regulators in Europe and Asia to attempt to:

– Develop demonstrably affective mitigation technology

– Develop compliance tests that ensure acceptable operation

Welborn, FreescaleSlide 17

doc.: IEEE 802.15-05/0583r0

Submission

September 2005

Next Steps• Help all parties reach consensus for a world-wide UWB band

• Target immediate rules in 5.8 to 10.6 GHz bands providing – Low risk for revisions to accommodate future mobile services

– Acceptable performance for key UWB applications

– No technology risk, provide low implementation cost/power

– Acceptable TTM coinciding with UWB products & standards

• Possible additional 6 dB in the high-band– Helps to match performance in 3-5 GHz low-band

– Nokia, Ericsson, Freescale have already called for consensus

• Urge to issue rules quickly to provide an environment for UWB industry to get established

• Support with white papers and measurement procedures and results as needed