Understanding the "Modeling and Testing an NR RF Receiver with LTE Interference" example

Hello everyone, I have a question about the 5G toolbox example: “Modeling and Testing an RF NR Receiver with LTE Interference”.

The example uses 2140 MHz (Bandwidth = 5 MHz) as the useful carrier frequency (NR) and 2150 MHz (Bandwidth = 3 MHz) as the interfering carrier frequency (LTE). In this case, there is no overlap between the NR and LTE signals.

If I move the interfering carrier too far away (eg, LTE carrier frequency = 2400 MHz) I get similar results when I would expect that moving the interfering carrier spectrally away would not have a significant effect.

Likewise, if I make a practically total overlap of the interfering carrier (for example, LTE carrier frequency = 2141 MHz), I understand that the overlap is total and the useful signal should suffer a high degradation, however I get EVM values very similar ​​to when it was totally remote (EVM about 2% in both cases, a very small difference in the last case).

I always use Interferer Gain = 1 (Waveform NR and LTE with default parameters).

Why this situation maintain similar results regardless of where the interfering carrier is located?

Thanks in advance!

Answer 1

[MJ Canavate Sanchez]

Hi Juan,

Let me start with your first question. If the interfering carrier is moved to 2400MHz, it will still be captured by the Bandpass Filter block and so the interferer will still have an impact in the EVM results (although less impact than the default configuration, as both signals are further apart). If you want to remove the impact of the interferer you can either:

1) Choose an interfering frequency which does not fall within the bandwidth of the Bandpass Filter so that it can be filtered out

2) Change the bandwidth limits of the Bandpass Filter

3) Set the 'Interferer gain' to 0

Regarding your second question, the reason why the EVM is the same in both cases, when the spacing is 10MHz and when the spacing is 1MHz, is that RF Blockset treats each carrier as a separate envelope so, if the carriers are really close to each other (they overlap), then you should probe both carriers at the output and then combine them. However, aggregating both signals at the output means that the thermal noise would be accounted twice, so better to combine them at the input and use a unique carrier/frequency in RF blockset. This way you will get the EVM values you expect.

I hope this helps!

[Juan Cattaneo]

Hi MJ!

Thank you very much for your answer, I really appreciate it very much and it was very useful for me to understand part of the processing.

I think that if I combine the signals at the input, before the RF Blockset and then work with a single carrier (NRFreqCarrier), I would be combining the baseband signals and that would be valid only in the case where both carriers are equal (NRFreqCarrier = LTEFreqCarrier).

So I think I could add to the example model a first RF Blockset to drive the baseband signals at their respective modulation frequencies and then use the Combine signal block to sum them up and convert the output signal of that first RF Blockset using NRFreqCarrier. Then I would use a second RF Blockset as figure in the example and only use one carrier (NRFreqCarrier).

However, viewing the signals at each point I see that I do not achieve the effect of combining them in RF (at the output I still have the baseband signal of NR). How is this proposal failing? Attach screenshot about this proposal (LTECarrierFreq = 2141 MHz / NRCarrierFreq = 2140 MHz, other params with default values).

On the other hand, when you indicate that it could combine the carriers at the output, in what specific place in the circuit would it be and how? That is, I understand that they should be combined within the RF Blockset from the example ...

Thank you very much in advance and sorry to have to move on ...

[MJ Canavate Sanchez]

Hi Juan!

You can combine both signals before sending them to the RF Blockset domain even when their carriers are not equal. One way you can do this (maybe the easiest way) is using the Signal Aggregator block. This block is available in the following example:

https://www.mathworks.com/help/comm/ug/interference-modeling-simulink.html

You can grab the block from that example and use it in your model. Note: do not forget to also copy the file 'signalAggregator_MaskInit' from that example folder to your model folder. Otherwise, the block will not work in your model.

I have sent you an attachment showing where the block should be placed in your model (right before the NR Input Buffer) and how you should set its parameters to have a spacing of 1MHz between both waveforms. Another thing you should do is selecting only one frequency at the Inport block in RF Blockset, e.g. NRCarrierFreq, instead of using both carriers (only one stream or "waveform" is entering RF Blockset). This way, you will transmit one carrier (centered at NRCarrierFreq) containing both wavevorms, LTE and NR.

If you want to combine both signals at the output (although I do not recommend this option due to the noise performance), you should use two more Outport blocks in RF Blockset to get the I and Q samples of the other carrier and then combine both carriers using a Signal Aggregator block after the Output buffer.

Basically, you need to combine both signals either at the input (recommended) or at the output.

Let me know if this works for you now!

[Juan Cattaneo]

Hi MJ!

Thank you very much for your detailed answer and your patience!

Your explanation is very clear and it was very useful to me.

I had just been analyzing models ("Interfence Modeling", "Interfence and Co-Channel Interference", among others) and had analyzed the 'Signal Aggregator' block internally (I just took the Frequency Shift block without the Interpolation FIR and the Check Token Attributes block) and achieved the expected results from EVM.

Your solution to combine both baseband signals at the input or the explanation at the output worked perfect for me!

Thank you very much MJ for your time!

[MJ Canavate Sanchez]

Glad to hear that! =)

[Yaser Abdallah]

Hi Juan, can you please share the simulink file with me that reflects the photo MR Sanchez suggested. thanks

Answer 2

[Yaser Abdallah]

Hello Mr Sanchez,

I have a question regarding this block diagram. I am trying to implement LTE RF receiver and study the Transmitter and receiver tests. does it work to use similar block diagram (5G NR RF receiver) for the LTE RF receiver?

I am implementing LTE RF receiver with another LTE interfering signal. thank you