Talk:Standing wave discrete Fourier transform

Have added a reference to the paper reviewing this research which was presented at the Engineering in Medicine and Biology Society, 1990. Proceedings of the Twelth Annual International Conference of the IEEE. The method of spectral analysis (SWDFT) for this application is used for the regulation of blood pressure during surgery, there are several (N) discrete Fourier analysis calulations vs a blocked FFT which has only 1. Maximal overlapping of spectral estimates combined with other indicators (uses Paresval's theorem to measure reliability of the estimator) maximizes the probability that of a reliable estimate of Patient-sentivity to the ansethesia, drug infusion delay, and the blood pressure baseline pressure of the patient is accurate during surgery. You don't want the baseline pressure estimate to go to "0".

Reference added: "Identification Of Patient-sensitivity, Baseline-pressure, And Infusion-delay Using Fast Spectral Estimates During Closed-loop Control Of Mean Arterial Pressure. Johnson, D.T.; Schneider, A.M.; Martin, J.F.; Quinn, M.L. University of California This paper appears in: Engineering in Medicine and Biology Society, 1990., Proceedings of the Twelfth Annual International Conference of the IEEE Issue Date: 1-4 Nov 1990 On page(s): 1878 - 1879 Print ISBN: 0-87942-559-8 Digital Object Identifier: 10.1109/IEMBS.1990.692063 Date of Current Version: 06 August 2002"

—Davidmeo (talk) 08:16, 21 April 2011


 * First, multiple papers by the same author do not establish notability. Does anyone else use this terminology?  Second, if this is your own work, see WP:COI.  Third, adding links to this everywhere is WP:SPAM.  Fourth, looking at the algorithm on the page it appears nearly identical to what is commonly known as the "sliding DFT" algorithm for the Short-time Fourier transform (except that your algorithm may differ by an overall phase shift), and hence I've changed your page to a redirect.  — Steven G. Johnson (talk) 15:41, 21 April 2011 (UTC)

I wanted to share this as it may be beneficial, however looks like it will be deleted - yes, I am the original author. This shouldn't link to short time Fourier transform though - The ability to tune heterodynes, use of heterdyne Low-Pass filter for spectral leakage and still identify FFT type redundancy, arbitrary quadrature modulator on the synthesis side for secure communications (if some "carriers" spectrum is attunated others get through), high number of spectral estimates (one per ADC sample) with no loss of resolution, no mention of using memory for harmonic weights to samples, etc make it useful. I don't believe one can figure out how to implement Figure 1 from any other Wiki article, try it in MATLAB (with some debug), you'll get N operations per Fourier transform on streaming data samples. Best regards, David

—Davidmeo (talk) 21:03, 21 April 2011 (UTC)


 * This article seems to clearly be on a non-notable subject. It should be deleted (but I am unfamiliar with the WP:AFD process, and the previous suggestion to delete the article seems to have been removed by the primary author of the article - who also seems to be an author of the subject matter). The originator of the proposal for deletion (Dicklyon at 05:26, 21 April 2011) said "Not notable. Does not appear in any book, and only one primary paper, as far as I can see'". I agree. I still agree even if more than one article has been written on the subject. This topic should not have an article until/unless the subject has been very widely recognized by third parties. —Mulligatawny (talk) 23:32, 21 April 2011 (UTC)


 * If the author will stop objecting, we can redirect it again and make it easy on everyone; if he finds a better place to redirect it, that's OK, too. We should add a section on sliding DFT to the DFT article, probably.  His implementation is slightly different from the usual, in a somewhat clever way, perhaps, instead it rotates the input, not the state, so avoids the potential amplitude creep that you sometimes see discussed in the sliding DFT.  But that means it gives a different transform, measuring phase differently (no big deal, and easily correctible if anyone cares).  But it's also just a standard frequency-shifting approach, a bunch of parallel filters, each implemented by shifting CF to DC, running a well-known efficient complex boxcar lowpass filter there, and not converting back up.  It's not clear to me why anyone would find such a filterbank useful, since you could for not much more cost implement a much better lowpass that implicitly gives a much nicer window.  It might be worth mentioning in a section on sliding DFT, but definitely not as an article, due to the WP:COI and lack of notability, and all the licensing language crap.  Dicklyon (talk) 00:56, 22 April 2011 (UTC)

I believe the redirect is incorrect - whether Wiki editors agree or not, the original article is covered by the Creative Commons attribution of license and copyright, specifically:
 * The SWDFT, SWFFT, SWDCT, variations and derivations may be used freely for benevolent scientific research, education, and 503(c)(3) non-profit public benefit purposes if implemented in software using a serial method (not implemented in parallel for example in an ASIC, ASSP, FPGA, DSP chip, hardware etc). Any other use including military or for-profit purposes requires a license from the author of “Estimation of Patient-Sensitivity, Baseline-Pressure, and Infusion-Delay using Fast Spectral Estimates for Closed-Loop Regulation of Mean Arterial Pressure” first published at the University of California, San Diego in 1990.

http://creativecommons.org/licenses/by-sa/3.0/ Attribution — You must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work).

—Davidmeo (talk) 02:21, 12 May 2011 (UTC)


 * David, I think you're mixing up a couple of things; first the redirect is to get readers to a relevant section of an article, since the topic is not notable enough to have its own article (see WP:GNG and thereabouts for more on notability). Second, there's this stuff about licensing.  You don't really need to say anything; any text you enter you are giving that license for.  As the use of the idea itself, wikipedia is not a good place to discuss that; it sounds like you're being an amateur lawyer here, but it's not clear you have any IP to license, and if you do, this is certainly not a good place to talk about it.  As for sources, as far as I can see your master's thesis is the only place this algorithm is published; master's theses are not always regarded as meeting the standards of WP:RS; especially when it's not easily availble for inspection and evaluation; do you have a copy online?  If the thesis is accepted as an RS, you can add material about your algorithm in the section that the redirect points to; but please leave out the licensing language.  Dicklyon (talk) 03:54, 12 May 2011 (UTC)


 * Also, I think you'll find that your SWDFT is the same as Julius Smith's Running-Sum Filter Bank (and ); other people may have other names for it, like "steady-flow DFT" in this old paper, though their algorithm is a bit different. Dicklyon (talk) 05:23, 12 May 2011 (UTC)