It's the first time someone asked a question directly, so I appreciate it. The previous answer is also an attitude of communication, and I dare not say that I am absolutely right. Awesome people are not afraid of jokes at all. I also took the postgraduate entrance examination this year, so we can communicate with each other.

1, yes, the bit rate Rb after time division multiplexing has nothing to do with the duty cycle of symbols, but with the time width between bits.

2. Yes, when there is no correlation between symbols, the first zero bandwidth is the reciprocal of the pulse width.

3. 1? The problem is the minimum system bandwidth required without ISI. It can be understood that the system is an ideal low-pass band-limited system, which has nothing to do with the first zero-crossing bandwidth of the signal. It should be B? =? Rb/2, what is the highest frequency band utilization rate at this time? η=Rb/B=2？ Porter/Hertz (here is binary, Rs=Rb, 1 porter = 1bps).

I think I know what you mean. You should mean that since it is an equivalent low pass, all frequency bands of the signal can pass, and the first zero-crossing bandwidth of 2Rb is greater than the Nyquist bandwidth of Rb/2, right? But I understand it this way, because this is a digital signal, which can recover the original information from the sampling value at the sampling time, which is very different from the waveform distortion or complex envelope distortion required by analog signals; In addition, this can also be seen from the derivation of Nyquist criterion for ISI-free baseband transmission, as long as there is no error code in the sampled values. (Moreover, the so-called intersymbol interference-free transmission originally considered the intersymbol interference-free transmission with band limitation. )

3.2? First of all, I think "no ISI" and "minimum" mentioned in the question are the key points. You are right, it depends on the transmission mode. If it is a raised cosine roll-off system, the transmission bandwidth is different, but I don't say it here, I think it should be understood as an ideal low-pass system because it requires the least, which is an ideal situation. Frequency band utilization is independent of duty cycle. To understand how the band utilization rate is defined, assuming that the code rate is Rs and the system bandwidth is W, the band utilization rate is η = RS/W. If the roll-off coefficient is α(0≤α≤ 1), the highest band utilization rate (baseband) without ISI is η=2/( 1+α). Porter/Hertz, α=0 is equivalent to the ideal low pass of the system, which I think you know, is in the book.

3.3? Based on the above, I think you should understand the previous question. In fact, these three terms may all refer to the same thing. Personally, I think asking "minimum system bandwidth" and "minimum transmission bandwidth" should be the same question without ISI, and asking "system bandwidth" depends on the specific situation. What I actually ask more often is "What is the bandwidth of the signal".

I haven't learned much about the design of this frame, and I feel that it should be the reuse of low-order groups and high-order groups. You said f 1, 2f 1, 3f 1 (the relationship after sampling is still 1 times 2 times 3) can be regarded as basic groups, quadratic groups and cubic groups. Look at this icon and the contents of 6.3.4 of this webpage/JP2007/15/wanguochengo/text/6.3.htm &; Nbsp, see if it's what you want to know.