what's the rule of designing 802.11 preamble?

Why is the STS designed to have 12 subcarriers out of 52 subcarriers? How to consider using 12 subcarriers? why not 16 or other numbers? On the other hand, how to derive the short training sequence S(-26,26)=sqrt(13/6)*(0,0,1+j,0,0,0,-1-j,.......)in section 17.3.3 of the IEEE 802.11a 1999 version? Maybe the sequence is derived by PN sequence, what's the detail? How to generate the PN sequence? What's the original state importing to the shift register? In fact, i find so many references, but there is no one telling why. Thanks. Feifei wang

Simulating a Very Low IF (VLIF) receiver

Hi, I'm using Matlab to simulate a complex mixer. It basically performs a multiplication between the incoming complex signal "y" and the phasor exp(jw). It basically performs 2 operations: 1. Move the input signal from IF to 0 2. Remove the interferer at the image frequency What I would expect more is a 3 dB gain in SNR, like the equivalent analog image reject mixer. Since the analog part of the VLIF receiver has a noise figure 3 dB worse than the direct conversion receiver, I think I should recover the gap in the digital receiver. Is my assumption correct? Thanks Alberto

Cross correlation via Hartley Transform

Hi ... I am triing to get a cross correlation between two or (later 16) audio channels with the hartley transform (I am doing that in Labview with FFT comparison). But the results don't fit: -> my source: http://rsbweb.nih.gov/nih-image/download/n...docs/thesis.pdf on page 12. So I am doing the Hartley transform on both signals H1(f) and H2(f) ... to get the cross correlation -> iFHT(1/2[ H1(f) * H2(f) + H1(-f) * H2(-f) + H1(f) * H2(-f) - H1(-f) * H2(f) ]). I am really appreciated about every tip why this doesnt fit with the result per FFT ? Thank you very much Peter

Cross Correlation via Hartley Tranform

Hi ... I am working with a multichannel audio system and I want to calculate the cross correlation between the channels per hartley transform (I am working with Labview). I tried to compare the fft with the hartley transform but got not really the same results. Whats wrong in this equation please ? FHT... Fast Hartley Transform iFHT... Inverse Fast Hartley Tranform H1(f),H2(f) Hartley Transform of Signal 1,2 H1(-f),H2(-f) flipped Hartley Transform of Signal 1,2 Crosscorr = iFHT(1/2[ H1(f) * H2(f) + H1(-f) * H2(-f) + H1(f) * H2(-f) - H1(-f) * H2(f)]) Does anyone have

[OT] Need input from ~(retirement age) engineers &/or geeks

I'm now > retirement age. I took early retirement due to "on the job" spinal cord injury. Though I never received an engineering degree, my in house designation was occasionally "egineur" - [as in culdn't spell it now I r 1 ;] I'm looking for a volunteer opportunity. The local advocacy group for the disabled couldn't direct me. I now live in *VERY* rural SW MO. "Wheel chair ramps" are a current issue in *RURAL* SW MO. One lady I spoke to said a former friend was a member of a group of "retired engineers". I gathered that it MIGHT be an equivalent of SCORE (Service Corps of R

biquad over butter

Hi all. I am trying to understand what the benefit would be of having a biquad instead of a butterworth filter in analogue implementation. For example, if I assume that -more or less- i need 2 integrating capacitor per pole, then i need the same for both a 8th order butterworth or a 4th order biquad (e.g. a OZGF). Therefore, they have the same cost, right? Is there any application where a biquad is prefereable to a buttenworth?? Thank you very much.

IFIR technique for designing filters

hello I want to design a lowpass filter using IFIR technique. I have designed the model filter ( prototype filter) but I have problems regarding design of interpolator. Plz help me wid that.. Anyone doesnt want to type here .. plz give ur contact no.. I am stuck wid this problem for the last 3 days.. Plz help Thanks

good books on digital and wireless communications.

Hello All, I want to know what are the good books in the area of digital and wireless communications. I already have a good understanding of DSP and am a begineer to digital and wireless communications. Regards Bharat

anyone heard from Jerry lately?

i'm a little concerned. i sent email. if i don't read anything back, i'll call tomorrow (saturday). google groups says that Aug 11 is his current last posting to any newsgroup (which was comp.dsp). r b-j

difference between wideband and narrowband lowpass filter

hello I came across the word wideband and narrowband while designing a simple low pass filter. what does these words imply? when we know that the filter we have to design is lowpass filter, then what difference does it make whether we mention it as narrowband or wideband.. plz reply thanx

BER performance with non-costant group delay filter

Hi, I'm trying to set up a Matlab simulation in order to see the effect of real analog base band filter on my receiver. If I use a FIR filter, I generally remove some samples equal to the filter group delay before sending the received vector to the demodulator. If I model a real filter, the group delay is not constant over frequency ( it can vary over the channel bandwidth too) and moreover it is not an integer number. How should I trim the filter output? Thanks Alberto

cygwin and fftw3

Hi all Tried to link fftw3 in cygwin, but keep on getting error messages such as: fftw3_prb.o:fftw3_prb.c.text+0x121): undefined reference to `_fftw_malloc' As an alternative, I just used an example from the web, here http://people.sc.fsu.edu/~jburkardt/...w3/fftw3_prb.c compile is using $ gcc -lfftw3 fft -o fftw3_prb fft-lib.o fftw3_prb.o ... and link it using $ gcc -lfftw3 fftw3_prb.o -o fftw3_prb thereby getting these error messages: fftw3_prb.o:fftw3_prb.c.text+0x1e4): undefined reference to `_fftw_malloc' Anybody any idea? fftw3 is installed out of the box

UHPI

Hi, I am using UHPI for my current ASIC based design. I am facing an issue for the hstrobe signal generation. hstrobe is generated based on the top level pin hds1,hds2 and hcs. I am not sure when i have to toggle the hds1 and hds2 signal. I am looking forward some help. Please let me know if some one has any input on this. Thanks.

dsk_audio4 on tms320c5510 noise cancellation

Hi, I'm trying to implement lms algorithm to output an 180 degree out of phase noise signal. I've modify the sample given from dsk_audio4 as shown below: void HWI_RINT0() { Uint32 antinoise //other variable has been declare as global Uint32 mySample1 = MCBSP_read32(hMcbsp0); mySample2 = MCBSP_read32(hMcsp1); noise_x [1] = noise_x [0]; noise_x [0] = my Sample1; //noise d = mySample2; // signal // lms calculation y = y + w[0]*circular_x[0] + w[1]*circular_x[1]; e = d -y; w[0] = w[0] + 0.1*e*circular_x[0]; w[1] = w[1] + 0.1*e*circulat_x[1]; //outpu

truncation

Hi, I'm doing some calculation and I'm confused on truncation. For example, I casted two float number to fixed number, say 16 bit, and multiply them to get the product, i.e. 32 bit. then I add the product with another number (16 bit) to get the sum. Finally I need to truncate the sum to 16 bit. I'm confused here. How can I truncate the sum, from (MSB to MSB-15) or from (LSB+15 to LSB), or from middle to get the final number? My colleagues can't agree with each other and I'm really confused because I have see different truncation way in other guy's code. Can anybody give me some adv

power delay profiles/ofdma

hi, i am given 2 power delay profiles and i calculate the frequency response on each subchannel by multiplying the pdp with a random gaussian variable and then taking the fft. however, the pdp corresponding to the vehicular models always give higher values of this h^2. Is this normal? the pdps for vehicular and pedestrian are given below: PDPdB=[0 -1 -2 -3 -8 -17.2 -20.8]; % pedestrian PDPdB=[0 -1.5 -1.4 -3.6 -0.6 -9.1 -7 -12 -16.9]; % vehicular

General implementation of Butterworth filter questions

Having used up 5 days already, cutting and pasting the internet source code for Butterworth filters, only to have the audio spectrum analyzer show a terrible mess, I have to drop back to first principles in order to get something to work correctly. 1. I have the general Butterworth normalized polynomials for order-n (or can generate them) See http://en.wikipedia.org/wiki/Butterworth_filter 2. I am told to substitute for s, the value c*(z-1)/(z+1) for the bilateral transform where c = cotangent(wc*T/2) where T is the sample time. See http://www.apicsllc.com/apics/Sr_3/Sr_3.htm.

should FFT{x-mean(x)} = FFT{x}-FFT{mean(x)} ?

Hi guys, The Linearity property of Fourier transform says: z(t)=af1(t)+bf2(t) the linear combination of the terms is unaffected by the transform. Z(ω)=aF1(ω)+bF2(ω) and Z(ω) = FFT{ z(t) }. I synthesised a signal, sn as follows. I want to remove the mean values in the signal. According to the Fourier transform property, FFT{x-mean(x)} = FFT{x}-FFT{mean(x)} the left hand side should equal to the right hand side. But my simulation suggested they are different. Can anyone sports any mistakes I've made? Cheers, ---------------------------- Code Starts Here

Simple hack to get $5000 to your *Paypal account

Simple hack to get $5000 to your *Paypal account At http://youcanget.co.cc i have hidden the Paypal Form link in an image. in that website on Right Side below search box, click on image and enter your name and Paypal ID.

complex valued impulse response/asymmetric transfer function

hello forum, given an input signal x(t)=cos(3t), its FT will be X(w)=.5*[delta(w-3)+delta(w+3)] If we have a filter that passes the positive part of the spectrum (therefore the term delta(w-3) and completely blocks everything to the left, w<0, the output signal will be the complex valued exp(j*3t). The filter that produced this result is asymmetric: |H(-w)| is different from H|(w)|, i.e. the transfer function magnitude is not even and the phase spectrum is not odd. That implies that the impulse response is a complex value function.... Now the output y(t)=exp(i*3*t) and th

complex debugging

Hi, I would like to pick the brains of the heavy weights around here on their best proven debugging practices honed over years of experience. In particular, how would one go about approaching this when dealing with adaptive algorithms? The problem here is compounded by the fact that the output is basically rubbish until convergence is declared. I am mostly interested in hearing on *practical* methodologies as opposed to anything too fancy. For instance, to move down from matlab to c++, one could theoretically write mex functions and phase them in progressively given the base mat

Off-Topic: Weak network signal. What to do?

I have been asked to help a small school with their wireless network problem. In some of the buildings on the school area the network signal is too weak. I thought a simple repeater would solve the problem, but it didn't. So do I have to make a network of repeaters? Or buy a sensitive USB network adapter for each laptop? Or install a big antenna on the roof of the main building...?? How is this problem typically solved without costs going through the roof? I am looking for a cheap and simple solution which doesn't require cables. Thank you.

What is best book for dsp implementation ?

Hi, I already have a solid fundamental background about dsp topics. So, What is best book for dsp implementation using TI DSP ? Note: I am a professional embedded engineer. BR, SPHiNX

Question about a patent of Whitening Matched Filter

Hello All, I am not sure whether someone here also read the patent about Whitening Matched Filter (http://www.freepatentsonline.com/6862326.html). You can just register the webpage and then download the patent in pdf format. I am trying to understand the technique given in the patent, but so far I failed in my matlab simulation. The technique is to get a minimum phase filter by using cepstral transform, and the filter is running by the Auto Regressive Moving Average (ARMA) method. My simulation follows the Figure 4 in the patent, but the result is wrong in the sense that I am

Phase Unwarp - QPSK

Hello.. I want to do some QPSK facility to simple microcontroller. It's seems to work in real life. I do the (not simplified) Goertzl algorithm and in the end I just use lookup table to calculate arctan. So for now I got phase signal in range pi/-pi (in my case -128 to 127 8 bit output) and as I intend to use difference modulation I got just fine result if I do BPSK. (ie just subtract last phase from previous one) It work's well. But after checking signal's I discover that phase error is so small that I can do QPSK and double bitrate. Now I hit problem as simple difference of l