Bug 2640

Summary: Use default gains of 0 dB
Product: ns-3 Reporter: sebastien.deronne
Component: wifiAssignee: sebastien.deronne
Status: RESOLVED FIXED    
Severity: enhancement CC: ns-bugs, tomh
Priority: P5    
Version: unspecified   
Hardware: All   
OS: All   
Attachments: Set default gains to 0 dB

Description sebastien.deronne 2017-01-28 12:43:34 EST
Created attachment 2766 [details]
Set default gains to 0 dB

Default TX and RX gains are defined to 1 dB for no real reason.
I suggest to set the default value to 0 dB.

3 tests are failing though:
    devices-mesh-dot11s-regression
    devices-mesh-flame-regression
    routing-aodv-regression

There are two proposed solutions:
- either we rescan pcap files;
- or we force gains to be 1 dB in those tests.
Comment 1 Tom Henderson 2017-01-28 14:06:33 EST
> 
> There are two proposed solutions:
> - either we rescan pcap files;
> - or we force gains to be 1 dB in those tests.


IMO, I support the change, but request also that CHANGES.html call out this change in model behavior.

As for pcap files, they should go away anyway in favor of behavioral tests, so to make progress on this now, I would support a local change to the test to preserve 1 dB (and mark that this configuration should go away in future revision to the test); actually I would suggest to leave this bug open to remove the 1 dB from the tests.
Comment 2 sebastien.deronne 2017-01-28 14:32:32 EST
(In reply to Tom Henderson from comment #1)
> > 
> > There are two proposed solutions:
> > - either we rescan pcap files;
> > - or we force gains to be 1 dB in those tests.
> 
> 
> IMO, I support the change, but request also that CHANGES.html call out this
> change in model behavior.
> 
> As for pcap files, they should go away anyway in favor of behavioral tests,
> so to make progress on this now, I would support a local change to the test
> to preserve 1 dB (and mark that this configuration should go away in future
> revision to the test); actually I would suggest to leave this bug open to
> remove the 1 dB from the tests.

Is it not better to open a thread per module to get rid of those pcap-based tests?
Comment 3 Tom Henderson 2017-01-28 17:39:21 EST
(In reply to sebastien.deronne from comment #2)
> (In reply to Tom Henderson from comment #1)
> > > 
> > > There are two proposed solutions:
> > > - either we rescan pcap files;
> > > - or we force gains to be 1 dB in those tests.
> > 
> > 
> > IMO, I support the change, but request also that CHANGES.html call out this
> > change in model behavior.
> > 
> > As for pcap files, they should go away anyway in favor of behavioral tests,
> > so to make progress on this now, I would support a local change to the test
> > to preserve 1 dB (and mark that this configuration should go away in future
> > revision to the test); actually I would suggest to leave this bug open to
> > remove the 1 dB from the tests.
> 
> Is it not better to open a thread per module to get rid of those pcap-based
> tests?

It does not make a difference to me; feel free to do so.
Comment 4 sebastien.deronne 2017-01-29 15:07:56 EST
OK, I will update CHANGES.html and update tests scripts.
Comment 5 sebastien.deronne 2017-01-29 16:02:43 EST
changeset 12589:5608f5d3d604