How to recover tx data from nrPerfectChannelEstimate

From Deep Learning Data Synthesis for 5G Channel Estimation - MATLAB & Simulink - MathWorks 中国. we can get 612*14*1 nrPerfectChannelEstimate.But I want to know how to recover tx data from nrPerfectChannelEstimate.Tx bit is 612*14*4 after qammode(tx,16),get 612*14*1 complex.And then thougth channel get y.I think the nrPerfectChannelEstimate is h. y./h we can get the tx.after qamdemod get the tx data.But it can not recover the real data,i do not wahy

Answer 1

[Balavignesh]

Hi 重阳,

It is my understanding that you would like to recover transmitted data from the channel estimate. Recovering transmitted data from the channel estimate in a 5G NR (New Radio) communication system involves several steps, including modulation, channel estimation, equalization, and demodulation. The process you've outlined generally follows these steps but may miss some details crucial for accurately recovering the transmitted data.

At the tx side, binary data is modulated using some scheme like '16-QAM', resulting in a complex symbol for each group of bits (612*14*4). The modulated data is then transmitted through a wireless channel, which introduces various impairments like noise, fading, and interference, resulting in a received signal 'y'. 'nrPerfectChannelEstimate' represents the estimated channel response 'h', of size 612*14*1, obtained at the receiver. This estimation is crucial for correcting the channel impairments.

The received signal is then equalized and demodulated using 'qamdemod' to recover the transmitted bits.However, there might be some potential areas that might cause issues in data recovery.

• Noise and Imperfections: The actual channel includes noise and other imperfections not accounted for by simply dividing by the channel estimate. Consider using more sophisticated equalization techniques that account for noise, such as 'MMSE' (Minimum Mean Square Error) equalization.
• Mismatched Dimensions: Ensure that the dimensions of your transmitted symbols, received symbols, and channel estimates align correctly. Misalignment can lead to incorrect equalization and data recovery.
• Demodulation Settings: When demodulating with 'qamdemod', ensure that the settings (e.g., modulation order, normalization method) match those used during modulation. Any mismatch can lead to errors in the recovered data.

Kindly have a look at the following documentation links to have more information on:

• 'qamdemod' : https://in.mathworks.com/help/comm/ref/qamdemod.html#mw_7c6d4f2f-a364-4c2d-9c43-929dcb91d58a:~:text=M%20%E2%80%94%20Modulation%20order
• 'MMSE' equalization: https://www.mathworks.com/help/5g/ref/nrequalizemmse.html

Hope that helps!

Balavignesh

[重阳]

Thank you a lot.

I refer to DL-SCH and PDSCH Transmit and Receive Processing Chain - MATLAB & Simulink - MathWorks 中国 this site, and know how to recover the signal briefly. Thank you for your explanation, which is same to that site.

Yours

Chongyang