upgrading to ns-3.39 (having issue with m_channel)

[George]

Hello,

I am having issues with upgrading to ns3.39 due to modification I did (https://gitlab.com/GeorgeFrangulea/ns-3-dev/-/commit/65e669180e514e30e3435e9b0d158fe51ba9c9ca

This modification allows me to set WiFi with Uniform Plannar Array, which I need in order to set WiFi with the same 3GPP spectrum settings as the NR module. 

Now, I get the following error:
ns-3-dev/src/wifi/model/spectrum-wifi-phy.cc:493:5: error: 'm_channel' was not declared in this scope; did you mean 'Channel'?
  493 |     m_channel->StartTx(txParams);
      |     ^~~~~~~~~
      |     Channel


Due to the fact that m_channel changed from 

•     Ptr<SpectrumChannel> m_channel; //!< SpectrumChannel that this SpectrumWifiPhy is connected to

to 

•         WifiScanParams::Channel m_channel; ///< The channel the management frame was received on  

Can anyone please give me some indication on how I could account for this change? 

Thank you. 

best regards,

George

[George]

According to this commit (https://gitlab.com/nsnam/ns-3-dev/-/commit/5c44a7ea014ad64f95d22176543445740ebc3cb3), as the name suggests the StartTx was moved to Spectrum Phy Interface. 

So I just did the modification to pass the Ptr<Object> instead of Ptr<AntennaModel> and it compiled ok. 

I hope this is not just surface level and this change will not create weird behaviour in WiFi. If anyone thinks otherwise, please let me know. 

[Tom Henderson]

Sorry, I am not following what you are asking. Are you talking about a
potential modification to ns-3-dev to change some type?

> So I just did the modification to pass the Ptr<Object> instead of
> Ptr<AntennaModel> and it compiled ok.
> I hope this is not just surface level and this change will not create
> weird behaviour in WiFi. If anyone thinks otherwise, please let me know.
>
> On Tuesday, 1 August 2023 at 01:37:57 UTC-4 George wrote:
>
> Hello,
>
> I am having issues with upgrading to ns3.39 due to modification I
> did

> (https://gitlab.com/GeorgeFrangulea/ns-3-dev/-/commit/65e669180e514e30e3435e9b0d158fe51ba9c9ca <https://gitlab.com/GeorgeFrangulea/ns-3-dev/-/commit/65e669180e514e30e3435e9b0d158fe51ba9c9ca>

>
> This modification allows me to set WiFi with Uniform Plannar Array,
> which I need in order to set WiFi with the same 3GPP spectrum
> settings as the NR module.
>
>
> Now, I get the following error:
> ns-3-dev/src/wifi/model/spectrum-wifi-phy.cc:493:5: error:
> 'm_channel' was not declared in this scope; did you mean 'Channel'?
>   493 |     m_channel->StartTx(txParams);
>       |     ^~~~~~~~~
>       |     Channel
>
>
> Due to the fact that m_channel changed from
>

> *     Ptr<SpectrumChannel> m_channel; //!< SpectrumChannel that

> this SpectrumWifiPhy is connected to
>
>
> to
>

> *         WifiScanParams::Channel m_channel; ///< The channel the

> management frame was received on
>
>
> Can anyone please give me some indication on how I could account for
> this change?
>
> Thank you.
>
> best regards,
> George
>

> --
> Posting to this group should follow these guidelines
> https://www.nsnam.org/wiki/Ns-3-users-guidelines-for-posting
> <https://www.nsnam.org/wiki/Ns-3-users-guidelines-for-posting>
> ---
> You received this message because you are subscribed to the Google
> Groups "ns-3-users" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ns-3-users+...@googlegroups.com
> <mailto:ns-3-users+...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/ns-3-users/45d3565f-cb89-41a1-b5a0-6007b5352341n%40googlegroups.com <https://groups.google.com/d/msgid/ns-3-users/45d3565f-cb89-41a1-b5a0-6007b5352341n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[George]

Hi Tom, 

Yes, you're right, let me give more context maybe this could actually have a more permanent fix so we can also keep NR-U updated to latest ns-3-dev and NR. 

5G-LENA NR-U module still uses ns-3.35 and 5g-lena-v1.2.y. Before I started working on it I got this explanation (regarding wifi-setup.cc):
(March 2022)  In the next release of ns-3 dev, this function will not exist anymore: 

• spectrumPropagation->AddDevice (netDevice, antennaArray).  
  Where, const Ptr<ThreeGppSpectrumPropagationLossModel> & spectrumPropagation

So this code will need to be upgraded. E.g. one will need need to do to set to WifiSpectrumPhy an antenna object that is of type Ptr<UniformPlanarArray>.

I upgraded the code myself so I could work with the latest ns-3-dev and NR. However, since the changes I did can not go into ns-3-dev, this issue prevents anyone to further any work using NR-U since important bugs that were fixed (in the latest ns-3-dev and NR) are not available in these early versions. 

All in all, I always have to do small changes to allow the use of Ptr<Object> instead of Ptr<AntennaModel> in order to use the latest ns-3-dev with NR-U.

• spectrum-signal-parameters
• spectrum-wifi-phy
• wifi-spectrum-phy-interface (Due to the latest change mentioned in my previous message) 

Anyway, thanks for your interest. Not sure how much interest is for NR-U atm, but if this is fixed, we could have one more module working with the latest ns-3-dev while also fixing major bugs that prevent people from running the current simulation for NR-U. 

Best regards,

George

[George]

I have come across this old MR (https://gitlab.com/nsnam/ns-3-dev/-/merge_requests/686) which explains better the need of having Ptr<Object> also seems to also have happened around the time NR-U stopped being updated to the latest NR and ns-3-dev.

[Tom Henderson]

OK, so can you clarify, once you made this change, are things working for you again, or do you still have issues?

If you have issues, can you put some code on a branch somewhere for someone to look at more closely?

- Tom

To unsubscribe from this group and stop receiving emails from it, send an email to ns-3-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ns-3-users/978a4d69-6d3d-48e5-8319-5cde146b1ce1n%40googlegroups.com.

[George]

Hi Tom, 

It does build with the modifications I did, but the code that worked with ns-3.38 does not work anymore. I posted a description of the issue I encountered now and also some code snippets and traces from gdb. 

I consider that it may be my misunderstanding of the proper WiFi spectrum setup. If that is not the case and has no easy solution, I will create a branch so someone could have a look. 

Thanks.