Meaning: channel unknown and known at the transmiiter

What does "channel unknown" and "known at the transmitter" mean?

In books about signal processing, it has always mentioned it if capacity or transmission (estimation) is discussed. I can't understand the difference.

If I simulate the transmission, the channel is given to me (the channel is known at the Transmitter). I can model it using the standard function in Matlab or as a random function. In estimation algorithms, like ZF, it is also used as a channel ( inverse of channel matrix) for an estimate. What is the difference? How can channel be unknown?

edit 1.

It is unclear what difference is between CSI unknown and known. If I simulate transmission system, I need a channel. If H is given as a matrix or via rayleigh-function, it means CSI is known, isn't'? The Capacity is computed using a channel matrix (in its equation). I cant understand then why there are two difference equation: CSI known and unknown to the transmitter.

• I'm not quite sure what to explain here. Your channel is random. You can either have knowledge of the channel, or you don't. That's it. Feb 4 '20 at 10:00
• what is ZF? Can you please expand it Feb 4 '20 at 10:06
• In general, the answer in @MarcusMüller's comment is what people imply in talking about known/unknown channel. If it is still not clear, you can quote several paragraphs that show your point and we will discuss them. Feb 4 '20 at 10:38
• @jomegaA I'd bet it's "Zero Forcing" Feb 4 '20 at 10:39
• @jomegaA ZF is zero-forcing Feb 8 '20 at 8:58

You need to distinguish between transmitter (TX) and receiver (RX) here. For the receiver it is generally quite easy to gain knowledge of the channel: have the transmitter send a few known symbols, then the receiver can estimate the channel coefficients (e.g., via Least Squares or MMSE). The receiver needs this channel estimate for equalization. For example, the ZF receive strategy you mentioned essentially means you divide out the channel coefficient, i.e., if you receive $$y = h \cdot s + w$$ where $$h$$ is the channel, $$s$$ your unkown information symbol and $$w$$ the noise, the simplest you can do to estimate $$s$$ is divide out $$h$$ (use $$y/h$$ as an estimate), for which you need to know it.
• I have just read an implementation in Matlab: the channel is defined as Matlab function rayleigh, data as a random vector. It was written in a description " the channel state information is given", what does it mean? Does it mean the channel $H$ is given ( defined as rayleigh channel via Matlab function), isn't? if yes, then what it means if "channel state information is not given"? Feb 8 '20 at 8:58
• Yeah I guess given and not given refers to the channel matrix $H$. Feb 9 '20 at 20:46