What makes a signal analog or digital?
Let's say I have a square wave signal. Is this digital or analog? Or neither?
It depends on what I do with this signal. Am I sampling this signal at brief intervals of time and then taking decisions based on comparing the value of the signal with a threshold? Then the signal is digital.
Am I driving a stepper motor continuously with the square wave signal? Then it is analog, since I am neither sampling nor quantizing the signal.
Regarding the question on how the RF waves are transmitted, there is no difference in how they are transmitted (be it analog or digital communication). They are transmitted as Electromagnetic waves. What matters is are they encoded and decoded or not. Is the information transmitted as it is without sampling and quantization? Then it is analog communication, else it is digital communication. For digital communication, you have to ensure that the signal is sampled at or above the Nyquist sampling rate for no loss of information (read the Sampling Theorem for more details on that). The continuous time signal is converted to a digital signal at the receiver side, while the digital signal is converted to a continuous time signal at the transmitter side in digital communication. This conversion process is done using modulation at the transmitter side and demodulation at the receiver side.
Regarding the concern of noise, there are clever tricks used to ensure the signal can be transmitted and received even in the presence of noise. The digital signal is compressed, and redundant bits are added to enable error correction. There is the Shannon limit, that tells us the maximum bandwidth of a channel at a particular noise level. Shannon is the Godfather of Information Theory, which is a beautiful field of study in Electrical Engineering that covers topics like information compression and error correction and detection. Noise can also be filtered using digital filters, this is studied in digital signal processing, another beautiful field of study in Electrical engineering that covers topics like signal reconstruction and filtering.
One might also ask questions like Analog is Digital? to which some people (especially those from an electrical engineering background who have not studied quantum physics - as it is evident on Electrical Engineering Stack Exchange and a comment on this question by a user) might erroneously conclude that the digital signal is in fact analog. However, there is no proof that the world is analog and not digital, nor is there proof the other way round. We never know if the signals in the real world are actually continuous, or if it is discrete and appears to be continuous. Does a quantum state take on a continuous set of probabilities or is it just restricted to 0 and 1 (the probability wavefunction breaks down into a discrete value when measured)? Does light travel as a particle or a wave? All we know are theories, and theories can't (atleast so far they haven't) be proven to be an accurate description of reality itself. But using these theories we can do a lot of useful work and build useful inventions.