Bug 3015 - RadioBearerStats are not correctly connected
RadioBearerStats are not correctly connected
Status: RESOLVED FIXED
Product: ns-3
Classification: Unclassified
Component: lte
ns-3-dev
All All
: P5 normal
Assigned To: Manuel Requena
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2018-11-16 10:35 UTC by Manuel Requena
Modified: 2019-05-10 05:31 UTC (History)
3 users (show)

See Also:


Attachments
Patch to RadioBearerStatsConnector and eNB/UE RRC (35.68 KB, patch)
2018-11-16 10:36 UTC, Manuel Requena
Details | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Manuel Requena 2018-11-16 10:35:15 UTC
Currently, there are some issues with the RadioBearerStats. A few TX/RX PDUs are missing and a few PDUs are duplicated in the traces files of PDCP and RLC.

The previous RadioBearerStatsConnector relays on the states of eNB RRC and UE RRC. But these states does not match how and when the bearers (PDCP and RLC entities) are created.

The attached patch fixes how the traces are connected in RRC and creates new trace sources in eNB/UE RRC to align with the creation of the SRB1 and DRB.
Comment 1 Manuel Requena 2018-11-16 10:36:47 UTC
Created attachment 3217 [details]
Patch to RadioBearerStatsConnector and eNB/UE RRC
Comment 2 Tom Henderson 2018-12-06 20:37:58 UTC
Please push when you are ready, and please update CHANGES.html (new API:  new trace sources), and if existing trace examples documented in the ns-3 LTE model library documentation are changed, please update those too.
Comment 3 Zoraze Ali 2019-05-07 14:54:31 UTC
Hi Manuel, 

It looks good to me. I just have one comment and a question.

Comment: 

Could you please remove the following dead code?

188	//   std::set<uint64_t> m_imsiSeenUe; //!< stores all UEs for which RLC and PDCP traces were connected

268	  std::set<uint64_t> m_imsiSeenEnb; //!< stores all eNBs for which RLC and PDCP traces were connected	189	//   std::set<uint64_t> m_imsiSeenEnb; //!< stores all eNBs for which RLC and PDCP traces were connected

Question:

I simulated lena-simple example, and in RLC stats (DL and UL) I notice that we lose 2 PDUs in the start of the simulation. Do you know why?

Kind regards,
Zoraze
Comment 4 Zoraze Ali 2019-05-10 05:31:14 UTC
pushed to changeset ed411fa9

https://gitlab.com/nsnam/ns-3-dev/commit/ed411fa9c033bcbcf8b2a44e75c01982848f0df1