Good question, we have to get a little technical here.
When Claude Shannon wrote his communication theory, which became the basis of all the modern communication technologies, he created a distinction between two ideas called signal and noise. Signal was the useful information and noise was the useless information. The useless information was always inserted into every signal due to the medium or environment in which the message was sent. For example, if you are sending digital information (1s and 0s) some 1 may be flipped to 0 and vice versa, and that would effectively entail miscommunication. So, communication theory was essentially about how to alleviate the effects of noise.
One simple way to overcome the problem was noise was to repeat the information. For example, if you send the same message three times, and noise distorts the message once, you still have two good copies of the message. But this also meant that I have to use three times the bandwidth to send the message three times. So, the information capacity of a noisy channel was less than the bit rate of the channel. If you could send 100 bits per second, by repeating the message three times, the effective information rate would be reduced to 33.33 bits per second.
This insight then led to two directions of research. First, sending messages three times (or even more times if the channel noise increases) is not very efficient. You want to maximize the use of bandwidth by sending the least amount of redundant (repeated) information. This gave rise to coding theory in which the goal was to find an encoding scheme by which we can send the least amount of redundant information (thereby maximizing the channel capacity). Today lots of such codes are known such as Convolutional and Block codes, which don’t send the raw message, but a coded message. A hardware chip is used to encode and decode the message. Each encoder and decoder is designed to overcome the effects of noise. In short, you can take the encoded message, randomly flip some bits, and then decode and you will still retrieve the original message. The goal of information theory is to pack more and more information in fewer bits.
Second, Shannon showed that this ability to pack more information in fewer bits has a limit. Every time you have a noisy channel, you must waste some bits for redundant information although there is a minimum on the amount of redundancy you have to add. Shannon did not say how to achieve that minimum but his theorem allowed people to compute the minimum. As technology progresses, people are trying to get closer and closer to this minimum amount of redundancy. This minimum is called Shannon’s Limit. With 4G wireless, for example, we are very close to that minimum, which means to send more data you need to get more bandwidth rather than packing more and more information into the same channel by improving coding.
A side effect of these two problems is that your information rate is always lesser than the bit rate because some bandwidth is wasted in packing redundant information. Now comes the question of modulation. The ‘carrier’ wave being modulated represents the bit rate. The ‘information’ wave superimposed on that modulated wave represents the information rate. Since information rate < bit rate, therefore, the carrier wave must be of a much higher frequency than the information wave. Remember that the root cause is noise, over which we have no control.
Now, when we talk about information, there is no noise; it is only a signal. Some signal we don’t have control over, so we call it ‘noise’. But it is not noise; it’s just something that nature does to destroy the information we are sending. So, we have to speak about two kinds of effects — (1) some information is being created by our mind as we encode it in transmission, and (2) some information is being destroyed in the process of communicating it to others. Coding is the process by which we maintain the intended information by creating redundancy. Therefore, when we speak about time, we have to speak in terms of creating, destroying and maintaining.
Then there are two more effects of hiding and revelation, or unmanifest and manifest when we take quantum effects into account. So, the signal we are sending is manifest as a possibility, this possibility is then encoded into a symbol, the symbol is then transmitted, partly destroyed in the process, and then recovered and understood. Eventually, it becomes unmanifest or impossible. This idea is described in some Vedic texts by saying that Sakti has five functions — creating, maintaining, destroying, hiding, and revealing. Hope this was useful in understanding.