doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 1 Project: IEEE P Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [CSS Comment Resolution of Sponsor Ballot] Date Submitted: [Nov. 2006] Source: [(1) Kyung-Kuk Lee (2) Rainer Hach] Company [(1) Orthotron Co., Ltd., (2) Nanotron Technologies] Address [(1) 709 Kranz Techono, Sangdaewon-dong, Jungwon-gu, Sungnam-si, Kyungki-do, Korea , (2) Alt-Moabit 60, Berlin, Germany] Voice:[(1) , (2) ], FAX: [], [(1) (2) Re: [] Abstract:[] Purpose:[Submit Comment Resolution of SB] Notice:This document has been prepared to assist the IEEE P 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 P

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 2 20g Comment : 6.5a.5.1 Comment : 6.5a.4.6 Comment : 6.5a.4.6 QPSK Output phase sequence does not agree with mapper input sequence and Table 26c symbol map, i.e. (-1+j) is mapped to -pi/2 in the example and pi/2 in Table 26c Accept in principle The example will be reworked due to comment

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 3 Comment : 6.5a.2.2 bit

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 4 Comment : 6.5a a.2.5

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 5 Comment : 6.5a.5.1 The value of implies about 0.5 dB amplitude and 4 degree phase RMS accuracy. This seems too strict for low cost devices, sometimes measured through an integral : Additional Editorial 0.005

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 6 Floating-Point SpectrumFixed-Point (4-bit) Spectrum Properties of Quantized CSS Signal

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 7 Floating-Point WaveformFixed-Point (4-bit) Waveform (MSE = ) Properties of Quantized CSS Signal

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 8 Comment : 6.5a.4.3 Comment : 6.5a.2.6 The text says "The constant tau(m) is either added or subtracted", however eq. 1b shows (1-(-1)^n)tau(m) which means that the delay is eithen not added or added twice The use of QPSK differential encoding, although improves robustness to frequency offsets, incurs a significant sensitivity degradation due to "squaring loss" effect when used after a low-rate encoder The commenter states concern about sensitivity degradation of DQPSK. This statement is true for a very simple and straightforward combination of the demodulator and decoder. However it is envisioned that by known techniques a siginificant part of the sensitivity loss can be recovered with reasonable extra effort. Comment : 6.5a.5.1 Accept: Replace by is either „not added or added twice“ Proposed Reject.

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 9 Comment : Figure E28 and E29 Comment : 6.5a.4.3 In equation 1a, since chirp is described, the time should probably be squared. Proposed Reject The commenter obviously has strong experience and a very accurate feeling and expectation on the behaviour of b/g and in practice. The effect observed by the commenter might be caused by the assumptions made in Table E1 on the shapes of the transmit spectra and receive filters which at some point might be too conservative. Unfortunately the assumptions and conditions for these simulations have been subject of numerous discussions with TAG19 and thus cannot be changed easily. We apologize if some gap between simulation and reality remains.

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 10 Comment : 6.5a.5.1 Comment : 6.5a.5.1 Comment : 6.5a.5.1 Recommended Measurement Conditions: - Resolution BW: 100 KHz - Video BW: 1 KHz Need to add measurement conditions, as in 6.8a.11.2 incorrect inequality sign - does not tie up with fig 20g range is at least strangely specified and may be completely wrong. I think you are trying to say "between 11 and 22MHz"

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 11 Comment : 6.5a.5.1 Comment : 6.5a.4.3 Comment : page 55 Using |f| does not give a response normalised on the centre frequency. Applies several times in this para. Need to Re-alignment Bit 9 …Bit 11

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 12 Comment : page 49

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 13 Comment : 6.5a.2.2 bit

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 14 Comment : 6.5a.4.3 Comment : 6.5a.2.6 Comment : 6.5a &these time and frequency parameters are assumed to be derived from a reference crystal in a locked manner&'. Is this trying to say that ppm error in chirp sub-band centre frequencies, in chirp rate, and time gaps should all be equal? Feedback memory of 4 does not seem to be explained by the equation for DQPSK output - nowhere in the equation references the fourth sample back in memory The 2450MHz CSS physical layer violates the PAR. Comment : E6.5 typo "sytstems" Add the following reference to the first paragraph of 6.5a.2.6: For a detailed explanation of the index variables n and k, please see 6.5a.4.3 Accept in principle

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 15 Comment : 6.5a.5.4 Comment : 6.5a.5.4 Comment : 6.5a.2.2 non-overlapping upper

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 16 Comment : 6.5a.4.2 Comment : 6.5a.4.6 replace "more orthogonal" with "increase effective span of the vector space" Orthogonal is a absolute. replace with and rework the example based on this Table 26a shows that this choice generates all possible combinations of phase and is more illustrative than the example given where the subchirp coefficients degrade to only -1-j and -1+j Accept, Replace Input data: Demux, I: 0 0 1, Q: Serial to parallel, I: { 0 0 1}, Q{ 1 1 0} Bi-orthogonal Mapping (r=3/4), I : Q : P/S and QPSK symbol mapping, Mapper Input: (1-j), (-1-j), (1+j),(-1+j) QPSK Output Phase: -π/2, π, 0, π/2 D-QPSK Coding (Initial phase of 4 feedback memory for D-QPSK are all π/4) -π/4, -3π/4, π/4, 3π/4 Sub-chirp sequence of D-QCSK output: [exp(-jπ/4) * subchirp(k=1), exp(-j3/4 π) * subchirp(k=2), exp(jπ/4) * subchirp(k=3), exp(j3/4 π) * subchirp(k=4)] The time gaps are chosen to “give the four different sub-chirp sequences more orthogonallity”. by “to make the four sequences even closer to being orthogonal” Accept in principle, replace

doc.: IEEE a Submission Kyung-Kuk Lee, Rainer Hach Slide 17 Comment : All Comment : page 49 Comment : page 42 Please look over the use of italics and upright text within the equations of this draft. Same with Comment same with comment Need to upgrade