What are the different methods or techniques with which we could evaluate the performance of channel estimation algorithms in wireless communications systems, especially in Massive MIMO OFDM systems?
The channel estimation algorithms are evaluated by calculating the the bit error rate versus the signal to noise ratio Pe vs. Eb/N0.
In order to carry out this calculation you need to build the system in form of functional block diagram. Then expressing the signal symbols at the output of the transmitter, inserting the channel transmission matrix according to the channel model. Then model the reliever and assuming at its input specific noise power.
Having the S/N to ratio at the receiver one can calculates the bit error rate.
Then one changes the estimation method and recalculate the bit error rate and assess the results.As an example the zero forcing method, minimum mean square error and the maxim liklihood methods.
There are many papers treating this problem using matlab.
Best wishes
I describe three different methods in Chapter 3 of my book Massive MIMO Networks, which can be downloaded from massivemimobook.com
The estimation quality is measured in terms of the estimation quality, such as the mean-squared error (MSE).
However, as indicated by Abdelhalim abdelnaby Zekry, it is not the MSE that is the final performance metric in a wireless communication system but the actual communication performance. The achievable data rate is normally the best performance metric for modern system where data blocks of several kbit are transmitted using channel coding.
Thank you Muhammed, Abdelhalim and Emil for your responses.
If we go for BER analysis, it involves the performance of detection algorithm also. So, if I have a channel estimation algorithm and want to evaluate its performance and publish the work, is it enough to show the MSE performance?
Dear Jazir
The mean square error can be taken as an indicator to measure the the accuarcy of the channel estimation algorithm as hinted by Emile Bjoernson. It can be used as a comparison criterion for the algorithms. The algorithm with smallest mean square error will be the best.
Best wishes.
Answering your question about:
If we go for BER analysis, it involves the performance of detection algorithm also. So, if I have a channel estimation algorithm and want to evaluate its performance and publish the work, is it enough to show the MSE performance?
I suggest you to use both the MSE and the BER, given that the estimation quality is only given by the MSE but you don't need to demodulate your received symbols to get this metric. If your estimation algorithm present lower MSE than other, it doesn't mean it is going to produce lower BER, because for the BER you need to use a specific MIMO receiver, it can be the Maximum Ratio Combiner or the MMSE receiver. The achievable BER depends on the receiver and the estimation technique.
By using the Kalman Filtering (KF) as an estimation technique with the Maximum Ratio Combiner you can achieve BER performance comparable to that of the MMSE receiver with Least Squares estimation, but if you use the KF estimation with the MMSE receiver, the noise is going to affect the estimation quality of the KF making it impractical and the BER increases.
So your estimation algorithm can reduce the effect of noise and/or interference but you should use a specific MIMO receiver to compare the achievable BER. Jazir Shahul Hameed
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning