Rlc unrecoverable error
Download as DOCX, PDF0 likes458 views
An RLC unrecoverable error occurs when the maximum number of transmissions of a RESET PDU is reached for an RLC-AM entity. The UE will initiate a cell update procedure with a cause of "RLC unrecoverable error". The network can either support or not support call re-establishment based on timer values. When both CS and PS RABs are configured, an error in the PS domain can cause both connections to drop. To address this, 3GPP proposed using a SCRI message instead of cell update to report RLC errors, triggering RLC re-establishment without dropping both connections.
![RLC unrecoverable error
If the maximum number of tranmissions of a RESET PDU is equal to MaxRST-1,
then RLC unrecoverable error shall be indicated to upper layers. The concept of
RLC unrecoverable error is valid only for RLC-AM entities.
Cell Update Procedure
The UE will initiate a RRC Cell update procedure, when a RLC unrecoverable error
is detected. The cell update cause will be set to "RLC unrecoverable error".
Network behaviour on receiving cell update message with cause RLC unrecoverable error
There are two options available for the network
a) It can support call re-establishment
b) It cant support call re-establishment.
This is governed by the timer values T314(CS) and T315(PS) signalled in SIB 1. if they are set to 0,
then network doesnt support call re-establishment i.e RRC connection signalling release message will be
sent to UE and if these timers have a value other than 0, then the network will do the re-establishment via
the cell update confirm message.
Problems with multi-rab case
When CS + PS RAB is configured for UE and if there is a RLC unrecoverable error for PS domain, the resulting
procedure as described above ends up in the dropped call for both PS and CS domain. Even though the
problem is specific to PS domain, we end up in losing the voice call. This seems to give the user a bad
perception about the network.
Even if a network could support call re-establishment, there will be CS voice disruption in this case.
[more info needed]
Solution discussed in 3GPP HSPA RAN2 to overcome this problem
Instead of UE using cell update procedure to indicate the RLC unrecoverable error to NW, it was proposed
to use the SCRI(signalling connection release indication) RRC message with a different cause value.
In simplest terms:
1. Ue reports the RLC unrecoverable error via SCRI.
2. NW triggers RLC re-establishment via a Reconfiguration message.
3. UE calculates a new START value prior to the RLC re-establishment
4. UE sends back a response message with the new START value. NW and UE use the new START value for COUNT -
C initialisation upon RLC re-establishment.
UE Capability reporting](https://image.slidesharecdn.com/rlcunrecoverableerror-151009075111-lva1-app6892/85/Rlc-unrecoverable-error-1-320.jpg)
![UE will report the support for CU Less RLC error recovery in RRC connection setup complete and UE Capability information
message.
This is an optional capability.
Typical Network scenario for RLC error recovery with SCRI
1. NW sets up the PS RAB with rb id 'x' and if the UE and the NW supports CU Less RLC error recovery, the IE 'SCRI RLC
error report'
is included in the RRC Radio bearer setup message and set to true.
2. UE responds with RRC Radio Bearer complete to finish the RAB assignment procedure.
3. Later on, the UE realises that there is an RLC unrecoverable error. Also if UE supports and NW has requested for CU less
RLC error
recovery, then UE initiates the RRC SCRI message including new cause value "RLC unrecoverable error" and the RB 'x'
for which
the rlc unrecoverable error is detected.
4. NW receives the SCRI and initiates a reconfiguration message for RLC Re-establishment.
[more info needed for the calculation of START and COUNT-C values]](https://image.slidesharecdn.com/rlcunrecoverableerror-151009075111-lva1-app6892/85/Rlc-unrecoverable-error-2-320.jpg)
Rlc unrecoverable error
- 1. RLC unrecoverable error If the maximum number of tranmissions of a RESET PDU is equal to MaxRST-1, then RLC unrecoverable error shall be indicated to upper layers. The concept of RLC unrecoverable error is valid only for RLC-AM entities. Cell Update Procedure The UE will initiate a RRC Cell update procedure, when a RLC unrecoverable error is detected. The cell update cause will be set to "RLC unrecoverable error". Network behaviour on receiving cell update message with cause RLC unrecoverable error There are two options available for the network a) It can support call re-establishment b) It cant support call re-establishment. This is governed by the timer values T314(CS) and T315(PS) signalled in SIB 1. if they are set to 0, then network doesnt support call re-establishment i.e RRC connection signalling release message will be sent to UE and if these timers have a value other than 0, then the network will do the re-establishment via the cell update confirm message. Problems with multi-rab case When CS + PS RAB is configured for UE and if there is a RLC unrecoverable error for PS domain, the resulting procedure as described above ends up in the dropped call for both PS and CS domain. Even though the problem is specific to PS domain, we end up in losing the voice call. This seems to give the user a bad perception about the network. Even if a network could support call re-establishment, there will be CS voice disruption in this case. [more info needed] Solution discussed in 3GPP HSPA RAN2 to overcome this problem Instead of UE using cell update procedure to indicate the RLC unrecoverable error to NW, it was proposed to use the SCRI(signalling connection release indication) RRC message with a different cause value. In simplest terms: 1. Ue reports the RLC unrecoverable error via SCRI. 2. NW triggers RLC re-establishment via a Reconfiguration message. 3. UE calculates a new START value prior to the RLC re-establishment 4. UE sends back a response message with the new START value. NW and UE use the new START value for COUNT - C initialisation upon RLC re-establishment. UE Capability reporting
- 2. UE will report the support for CU Less RLC error recovery in RRC connection setup complete and UE Capability information message. This is an optional capability. Typical Network scenario for RLC error recovery with SCRI 1. NW sets up the PS RAB with rb id 'x' and if the UE and the NW supports CU Less RLC error recovery, the IE 'SCRI RLC error report' is included in the RRC Radio bearer setup message and set to true. 2. UE responds with RRC Radio Bearer complete to finish the RAB assignment procedure. 3. Later on, the UE realises that there is an RLC unrecoverable error. Also if UE supports and NW has requested for CU less RLC error recovery, then UE initiates the RRC SCRI message including new cause value "RLC unrecoverable error" and the RB 'x' for which the rlc unrecoverable error is detected. 4. NW receives the SCRI and initiates a reconfiguration message for RLC Re-establishment. [more info needed for the calculation of START and COUNT-C values]