What is the difference between "channel modeling" and "channel estimation"? The difference can be summarized in many aspects, e.g., objective of tasks.
So, the comparison among channel modeling, channel sounding, channel estimation, and channel prediction is wanted.
Channel modeling refers to the construction of a geometric and/or statistical model of a communication channel in a way that is realistic for a particular propagation scenario.
Channel estimation refers to the estimation of channel parameters. It can either be the complex parameter between a pair of antennas at a given subcarrier. This can always be done. If you know that the channel can be modeled in a particular way, with a reduced number of channel parameters, then you can instead try to estimate those channel parameters.
You can read about channel modeling for LOS and NLOS systems in Chapters 4 and 5 of my new book, where we also describe channel estimation methods:
https://www.nowpublishers.com/article/BookDetails/9781638283140
Thanks Prof. Emil Björnson Emil Björnson.
In some channel modeling references, I have also noticed that key parameters such as path loss-related coefficients are estimated or Ray-tracing procedures. So, I wonder if both are estimating channel parameters, but with different goals?
There are different types of channel models. Ray tracing tries to predict the exact propagation in a specific environment by simulating how signal rays bounce on objects and reach the receiver. Many other channel models are stochastic, so they try to generate "typical" channel in a particular environment.
Regardless of what model is used, you can generate a "true" channel parameter.
The goal of channel estimation is to determine the parameter value by sending signals between the transmitter and receiver.
A channel model serves as a mathematical depiction of how a communication channel impacts the propagation of wireless signals. To eliminate the noise and distortion caused by the channel from the received signal, it is necessary to determine the properties of the channel. The identification of channel characteristics is known as Channel Estimation.
I can answer this question using an example in wireless communications. Fading, based on Jak's model, can be generated by an autoregressive model (h[n]=ah[n-1]+e[n]. So using channel samples (h[0],...h[N-1]) we can estimate channel parameters ('a' and driving noise variance) this is called channel modelling. Now consider the received signal as y[n]=h[n]+w[n], we can estimate h[n] (using Kalman filter), this is called channel estimation. To summarize, we have the following problem
h[n]=ah[n-1]+e[n] (1)
y[n]=h[n]+w[n]; (2)
(1) estimating model parameters is channel modelling, and (2) estimating h[n] is the channel estimation problem.
- Badges
- Science topic
- Similar topics
- Books