# Simulation at Baseband or Passband

A naive question I guess. Would it make a difference in the results when a system is simulated at Baseband or at Passband. If I have a plot of a BER vs SNR curve that has been plotted at Baseband, will the same change if I perform the simulation at Passband with a carrier frequency of 2.4GHz? AFAIK, it should not make a difference, but I would like to know if what I am thinking is correct.

Another question is that if I have two signals from two transmitters and add these signals at the receiver, does it make a difference if these two signals are passband or baseband. Do the results vary? Since Passband simulation consumes a lot of time, would it be correct if I simulate the behavior at Baseband and assume that the result would be almost the same at the Passband?

• Lab Exercise 1: Bandpass signals & White Gaussian Noise Simulation %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 1. Band-pass signal simulation %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Sampling period and frequency Ts=0.001; fs=1/Ts; % Duration of signal t=[-511:512]*Ts; % Carrier frequency fc=200; % Bandwidth B=100; % Baseband message signal m=sinc(B/pit); % Bandpass signal s=m.*cos(2*pifc*t); % Plot time-domain signal figure, plot(t,s),title('Sample bandpass signal') axis tight % Power spectrum using FFT FFTsize = 1024; spec = fftshift(abs(fft(s,FFTsize)).^2); freq = [-fs/2: fs/FFTsize: fs/2]; freq – dilara Mar 3 '17 at 8:26

## 1 Answer

The only thing I could see, with the information you gave, is that the way you get from passband to baseband and vice-versa could affect your performances. It should not, but it could.

Let's say you have to down sample to get to baseband, and up sample to do the contrary and you use filters to do so. If your filters have variations of 1dB in their passband and a cutoff frequency that cuts some of your baseband signal, then your performances could be deteriorated. In an ideal situation it should not, but some considerations (restriction on the number of coefficients of your filters, for example) may imply some performance degradation.

• Let me rephrase my question: consider that I have simulated a communication system at baseband and would implement the same on hardware but using a carrier at 2.4GHz(Passband conversion). Would it make sense to compare the simulated results at baseband and the results from the experiment at passband? – smyslov Jun 8 '15 at 7:13
• Ok, so I guess you would like to compare a simulated baseband signal with a real RF passband signal, in terms of performance. Well, you could totally do so and it would make sense. How much depends on what you simulated as degradations of your signal. Did you only implement degradations from the RF channel (multi paths, AWGN, doppler etc etc) or did you also add some imperfections in your transmission system (like some frequency and/or time offsets, non-linearities of your power amplifier etc etc) ? – Loufylouf Jun 9 '15 at 22:39
• Exactly. I have just implemented degradations from the RF channel and haven't added any imperfections to the transmission system yet. I just wanted to find out if it is really necessary to simulate the system at Passband frequency which consumes a lot of time. – smyslov Jun 11 '15 at 7:05
• You can simulate the imperfections of your transmission system at baseband (simulating some frequency and time offset is pretty easy), that will give you a better idea of the performances you will obtain. Once done, it's not really necessary to test this at passband, but you should test your system in real-life after that. Note that reality is hell and simulation is heaven, so you might have some great differences of performance (coming from a lot of things, like improper RF channel simulation, frequency offsets etc etc). – Loufylouf Jun 11 '15 at 11:41