1. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 1 Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [Spectral Sculpting and a Future-Ready UWB] Date Submitted: [15 Sept, 2004] Source: [David G. Leeper] Company [Intel ] [Claudio Da Silva (Intel), Evan Green (Intel), Hirohisa Yamaguchi, (Texas Instruments)] Address [5000 W Chandler Blvd, Chandler, AZ 85226] Voice:[480-552-4574 ], FAX: [], E-Mail: [david.g.leeper@intel.com] Re: [This submission is in response to the IEEE P802.15 Alternate PHY Call for Proposal (doc. 02/372r8) that was issued on January 17, 2003.] Abstract:[This document describes software-controllable spectral shaping capabilities that are available in OFDM-based UWB signals.] Purpose:[To show how MB-OFDM-based signals might adapt to local regulations, remedy future interference situations, and prepare UWB for the opportunities of Cognitive Radio.] Notice:This document has been prepared to assist the IEEE P802.15. 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 P802.15.

2. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 2 Abstract  An essential feature of a long-lived UWB PHY is the capability it offers to control its radiated spectrum shape.  This capability deserves more consideration than we’ve given it to date.  Spectrum-shaping for interference control is easier and more effective with MB-OFDM than with DS-UWB.

3. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 3 Key Assumptions 1.The first requirement of a UWB PHY is that it not cause harmful interference to other services, either now or in the future (Part 15). 2.UWB’s multi-GHz span will surely include future-service spectrum allocations and usage models that we cannot predict today. 3.Physical separations may be insufficient to avoid interfering with some future services, even under today’s FCC Part 15 limits. 4.Spectrum regulations for UWB will likely vary from place to place around the world. Spectrum mask compromises may be required. 5.R&O-foreshadowed UWB power increases may be granted only if we can demonstrate compelling means for interference control.

4. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 4 Under the preceding assumptions … An essential feature of a long-lived UWB PHY is the capability it offers to control its radiated spectrum shape. 1.Such capability must be practical in terms of complexity, power consumption, and cost. 2.Such shaping should be controllable via software, in near-real-time, to respond to local regulations and time-varying local conditions. 3.Such shaping should be as “surgical” as possible in order minimize the degradation to UWB performance. 4.A UWB PHY without this capability runs the risk of: a.Being short-lived because of unforeseeable interference to future services. b.Missing out on new regulatory initiatives, in particular, Cognitive Radio.

5. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 5 MB-OFDM Coarse Spectrum Control Comes from Multibanding  Whole bands can be dropped to accommodate local situations.  This is coarse control, but it may be useful in some cases.  Remainder of this presentation is about finer-grain control.

6. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 6 Reminder MB-OFDM Symbol = Summation of N Orthogonal Tones * http://www.iec.org/online/tutorials/ofdm/ T secs * amplitude frequency

7. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 7 Basic Spectral-Sculpting Capability  128-tone OFDM transmitter* “zeros out” 6 adjacent tones near detected service  122 out of 128 tones still available  Notch depth ~15 dB  All-digital operation -- no analog filtering required  Advanced techniques allow deeper notches Example Suppose a narrow-band service near 4 GHz needs protection. *6-bit DAC ~15 dB Notch

8. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 8 Windowed Symbols Yield Deeper Notches IFFT Block X Windowing Shape Digital Multiply CnCn DAC Root Cosine (0.5) Window ~ 23 dB Notch 6-bit DAC, 6 tones zeroed

9. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 9 Notching via 2-Tap FIR Filter  Freescale has proposed simple 2-Tap FIR filter for notching  Tunability?  Decidedly “non-surgical”  Useful spectrum suppressed  Multi-notch feasible?  In contrast, OFDM notching is:  Simpler to implement (all digital)  Easily tunable in software  Less UWB spectrum suppression  Rx FFT enables spectrum scan  Multi-notch capable – enables on-the-fly spectral sculpting  Better fit to future cognitive designs +  125ps FIR Notch “Plow” OFDM Notch “Scalpel” :

10. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 10 7.25-MHz notch (RAS band) generated by 6 tones MB-OFDM DC tone (0) Generated notch depth > 45 dB* Relative Power Spectrum Density (dB) Relative Frequency (MHz) Res bandwidth=3 MHz Averaging time = 72 usec Quantization effects (now under study) may reduce notch depth - 41.25dBm/MHz Achieving Deeper Notches via Active Interference Cancellation (AIC) Hirohisa Yamaguchi (Texas Instruments) *

11. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 11 7.25-MHz notch (RAS band) generated by 8 tones MB-OFDM DC tone (0) Generated notch depth > 70 dB* Relative Power Spectrum Density (dB) Relative Frequency (MHz) Res bandwidth=3 MHz Averaging time = 72 usec Quantization effects (now under study) may reduce notch depth - 41.25dBm/MHz Achieving Deeper Notches via AIC (cont’d) Hirohisa Yamaguchi (Texas Instruments) *

12. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 12 Measured Notch – Intel Labs  17 tones nulled per notch  Analyzer settings:  3 dB / div  Ctr Freq = 3432 MHz  Span = 528 MHz  Res BW = 1 MHz  Vid BW = 100 kHz Notch Depth > 20 dB Evan Green (Intel)

13. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 13 Cognitive Radio Concept FCC NPRM adopted Dec 17, 2003, ET-03-108* [Cognitive radio technologies] include, among other things, the ability of devices to determine their location, sense spectrum use by neighboring devices, change frequency, adjust output power, and even alter transmission parameters and characteristics. A cognitive radio (CR) is a radio that can change its transmitter parameters based on interaction with the environment in which it operates. This interaction may involve active negotiation or communications with other spectrum users and/or passive sensing and decision making within the radio. (*Paragraph 1) (*Paragraph 10)

14. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 14 Example of a One-Time, Just-in-Time, Sculpted Spectrum for a Future Cognitive Radio  Custom-sculpted spectrum for local rules / situations  MB-OFDM receiver FFT allows spectrum scan for other services  Boost power where policy engine & spectrum-scan allow  One-time, just-in-time, spectrum self-regulation  In this example: ~280 MHz out of 512 MHz active  Higher power => higher speeds and/or ranges plausible (Notches obtained via tone-zeroing only)

15. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 15 Summary  Precision shaping of UWB signal spectra will help us:  Offer new degrees of freedom in negotiating regulations  Adapt to local regulations & interference situations “on the fly”  Mitigate future interference problems we cannot foresee today  Encourage, and then exploit, more liberal UWB power limits  OFDM-based UWB offers spectrum notching and shaping capabilities superior to those of pulse-based UWB.  Spectral-shaping capability deserves more weight than we’ve given it to date in TG3a discussions.

16. Submission doc.: IEEE 802.15-04/0425r0 September 2004 Slide 16 Recommended Call To Action  DS-UWB and MB-OFDM camps each choose one scenario in which protection for other nearby services is required.  Victim services may be passive or active, current or future hypothetical  Victim services may be fixed (known) or transient (discovery needed)  Each camp will assess their solutions for both scenarios on the following bases:  Protection afforded to victim services  Additional complexity, power consumption, cost to provide protection  Performance impact to UWB system with protection active