We have found an interesting scenario we would like you to share with the SGSN and HSS experts .
Below Description of scenario
When the user moves from X SGSN to MME (3G to LTE network), X SGSN is never informed of this RAT TYPE change, and it keeps a false entry for this user in the SGSN despite it's not attached to the 3G network anymore. However, the HLR knows the UE is not attached anymore to the 3G network, because the SGSN number and address cannot be seen anymore from the HLR GUI after the RAT TYPE change is completed. We think this might be one of the main reasons why we see more users attached to the 3G network in X SGSN than those we see in the HLR.
We think this is happening because the HSS-FE removes the SGSN number and address itself from the HSS when it receives the ULR from MME. However, it cannot send a CANCEL LOCATION to X SGSN because S6D interface is not available .
In order to make myself clear, you can find more comments below about the scenario.
1) The UE is attached to the 3G network, that is, it’s attached to SGSN
2) It moves from SGSN to our MME (from 3G to LTE). The MME sends the LOCATION UPDATE via S6A interface towards the HSS.
3) However, the CANCEL LOCATION is never received by our SGSN because the S6D interface between the HSS-FE and the SGSN is not implemented in our network. Therefore, the SGSN does not know that this user moved from 3G to LTE, and keeps the record in the SGSN. On the other hand, the SGSN number and address are removed from the HLR, and we think this is done by the HSS-FE itself.?
4) The SGSN keeps the record while the HLR don't, which creates a difference in the number of subscriber attached to the 3G network between the SGSN and the HLR. The SGSN sees more subscribers attached to the 3G than those seen by the HLR.
Could you please confirm that you have now a clear picture of what the issue is?
Could you please provide your recommendations in order to fix this issue?
Is't S6d interface is mandatory for this scenario?
If you have any question, please, let me know.