I'm not sure about the '1-bit' term. This part does not make any sense for me.

Regarding compressed/compressive sensing: do you already know the wikipedia page   https://en.wikipedia.org/wiki/Compressed_sensing  ?

Hope this helps - a bit

Did you check out this page. I don't have experience with 1-bit CS. From their definition, it captures the sign of the conventional CS measurement. That is

Conventional CS: y = Phi * x ,

1-bit CS: y = sign(Phi * x);  

 As results, its measurements are binary number  which can represent using 1-bit.

http://dsp.rice.edu/1bitCS/

thank you for your answers, 

@ Thuong: I am asking about this one, the sign means + or - right? and how can we consider analog signals in this case, 1-bit compressive sensing can be applied to sample and recover analog ignals for spectrum sensing? 

Thanks,

@Fatima In theory, CS sampling matrix follow Gaussian random i.i.d, so that negative measurement is possible. In practice, CS sampling matrix are all positive ( mostly they consider 1/0 value). Therefore, instead of comparing to 0 in common sense of sign, you can compare it with a value L. 

IMO, you better read the introduction paper on 1-bit CS from that team.   http://boufounos.com/Publications/Boufounos_Baraniuk_CISS08.pdf

@Thuong: thank yu very much, 

i am going to read the 

1-bit compressive sensing has to do with quantization of the compressed measurements. If the quantization is so coarse that each of the measurement is quantized to one of only two codewords (e.g. >=T or

@Raziel: thank yu very much, 

Recovering sparse signal from the sign of the linear measurement

@Bo: Thank you,

@BO, Raziel, Thuong: what are the requirements of the sampling matrix that can work under the 1-bit compressive sensing?

thank you.

@Fatima Salahdine : The sampling matrix is the same as the one for ordinary compressive sensing

@Bo and Valerio: thank you for the great replies.