-We have seen in the previous episode that the band used by a signal depends on the shape of the physics symbols associated to a binary element or a group of binary elements. We will see in this episode that it also depends on the rate of the symbols we want to transmit. Here are two examples of NRZ signals transmitted with different symbol rates. Ten symbols sent per second on the left, twenty symbols sent per second on the right. The lower illustration represents the energy distribution of these two signals depending on the frequencies, their spectrum. We note that the signal with the highest symbol rate is the widest spectrally speaking. So it has the most important band used. It will require the most bandwidth in the propagation channel. This is the reason why, at the beginning of the week, I told you that if we wanted to constantly increase rates it would be costly in terms of bandwidth. Increasing the symbol rate in a signal spreads the energy in the frequency domain and thus increases the band used by this signal. We can see that for a NRZ signal here but whatever the shape of the physics symbol associated to binary elements or a group of binary elements, the band used by the resulting signal is always proportional to the rate of the symbols we want to transmit. A way to narrow the band used by a signal is to reduce the symbol rate. As we previously saw it, the symbol rate is inversely proportional to the number of bits we put in a symbol. In this example, two signals have the same bit rate but different symbol rates because in the right-hand signal, we coded two bits in a symbol. So we reduced the symbol rate and the band used by this signal for a same bit rate. Of course, there is a price to pay. We have the same bit rate and a lower band used because we coded several bits in a symbol. Thus we have a better spectral efficiency. The frequency spectrum will be better used. The price to pay will be in terms of energy. To get the same bit error rate, we will have to dedicate more energy to the transmission of these signals. It is always the same issue about the compromise to be found, the rectangle law, between energy, bit error rate, band used and the rate we want to transmit. We have seen this compromise by studying the modulator aspect of the transmitter, the way a signal can be formed to get the best possible compromise. You will get back to this issue during week 4 from the point of view of the channel coding which is the other key element of the transmitter part, and also from the receiver point of view.