Interference modeling in simulink

Hello everyone,

I am seeking your help with interference modeling in Simulink. I found two helpful examples on the MathWorks website: Adjacent and Co-Channel Interference and Interference Modeling. Both examples show that the primary transmitter and the interferer use the same sampling rate before being combined in the Matrix Concatenate block.

However, my case is a bit different, as my primary signal and the interferer have a significant difference in bandwidth. My primary signal has a bandwidth of 600 Hz with DBPSK modulation, while my interferer signal has a bandwidth of approximately 250 kHz.

From my understanding, the primary signal must be upsampled to match the highest sampling rate of the signals being combined, which is about 417 times. I have tried using the Rate Transition block, Upsample, and FIR Interpolator blocks in Simulink for this upsampling.

Unfortunately, the output spectrum now shows significant imaging/aliasing, and the receiver cannot decode the signal anymore (the BER has gone to 100%).

Is this the correct approach for interference modeling? Are there alternative methods I should consider?

Best regards,

Fumihiko Sato

Answer 1

[Kothuri]

Hi Fumihiko Sato,

In interference modelling in Simulink in which the primary signal and the interferer signal are having different bandwidths, you can try the following steps:

• Before upsampling, apply a low-pass filter to the primary signal to reduce spectral content near the Nyquist frequency. This can help minimize imaging artifacts after upsampling.
• You can use a bandpass filter around the signal's bandwidth (600 Hz) after upsampling to remove unwanted spectral components.
• Upsampling by a factor as high as 417 can introduce significant artifacts and is computationally intensive. You can try using multistage interpolation in which you upsample in smaller stages (e.g., by factors of 4, 8, etc., until you reach the required rate). This helps reduce aliasing at each stage, as each upsampling stage can have an anti-aliasing filter.
• Simulink offers Multirate Filters, which can be configured to apply gradual interpolation and filtering across multiple stages.
• You can use the “Sample-rate match” block
• You can use MATLAB’s “resample” function outside Simulink for preprocessing the signals. The “resample” function in MATLAB applies an FIR anti-aliasing filter automatically and may yield better results. Once resampled, you can import the signal back into Simulink for interference modelling.

You can refer the below documentation links for more info on “resample” function and “sample-rate match” block

https://www.mathworks.com/help/signal/ref/resample.html?&s_tid=srchtitle_support_results_1_resample

https://www.mathworks.com/help/comm/ref/sampleratematch.html?&s_tid=srchtitle_support_results_3_interference%20modelling%20with%20different%20sample%20rates

You can refer the below example for more info

https://www.mathworks.com/help/5g/ug/modeling-and-testing-an-nr-rf-receiver-with-lte-interference.html