making a data series with 2 or 3 discrete pulse, rest are zero

Hello, I am new to matlab. I want to prepare a data series of say 1024, with 2 or 3 sharp pulses at discrete positions. I want rest points to be zero (preferable) or very small values, so that I can really identify the pulses. Later, if I want to add a normally distributed random noise, how I can do that. thanks a lot in advance. regards. asim

Answer 1

[Fangjun Jiang]

Series=zeros(1024,1);
RandPos=[10,13,50];
PulseHeight=2;
Series(RandPos)=PulseHeight;
Noise=rand(1024,1);
Signal=Series+Noise;
plot(Signal);

[Asim Biswas]

Thank you so much. its great.

I was just wondering if it is possible to create the pulse as harmonics. For example, i want to present the series as the harmonics in sine wave. The pulses will be there but in the form of harmonics.

thanks again.

regrds.

asim

[Chaowei Chen]

in addition to Jiang's code,

impulse_response=[1 2 3 2 1];

waveform = conv(Series,impulse_Response);

plot(waveform)

[Asim Biswas]

@Jinag, thanks.

@Chen, thank you so much. It better. I was wondering of creating the series as harmonics as a function of sine or cosine wave. Can I do this? thanks.

[Fangjun Jiang]

@Asim, then you need to specify the frequency, magnitude, also the time steps for your 1024 data point.

[Asim Biswas]

Thanks Jiang. I appreciate it.

lets say, 0.1 sec interval from 0.1 to 10.24 sec. I want a resonable frequency and say unit magnitude. However, I want to see the pulse for a very short period but high magnitude. It should look like a peak in the middle of no where.

I wonder if I can make the question clear.

thanks a lot.

regards.

asim

[Fangjun Jiang]

All right,the point here is to show you the method, not give you the complete solution. From the above example, I hope you start to understand how easy MATLAB can enable you to do this. To define time, t=0:0.1:10.24. To generate a unit sine wave with frequency f, f=2; Data=sin(2*pi*t); You can use the same method to generate the pulse signal with high frequency and high magnitude. Then to add them together at a certain time spot, you can use for example: Data(51:60)=Data(51:60)+PulseSignal(1:10). Note that your sample time is 0.1sec, which means 10Hz, you can't really represent any signal with frequency high than 5Hz