A Blog dedicated to Declutter 3GPP specifications

Wednesday, November 25, 2020

NR-U Physical Frame structure (Candidates)


NR-Unlicensed is still in the study phase with barely any work done on the TS. It is expected that by 2022 (so Release 18) we may see some specification for the 6GHz band in Europe. As of now, only potential candidates for NR-U implementation can be guessed. 

Physical layer aspects

NR-U supports both Type-A and Type-B mapping already supported in NR.

Initial active DL/UL BWP is approximately 20MHz for 5GHz band, though the final value will be quantized to number of PRBs. Initial active DL/UL BWP is approximately 20MHz for 6GHz band if similar channelization as 5GHz band is used for 6GHz band.

Frame structure

Single and multiple DL to UL and UL to DL switching points within a shared gNB COT is identified to be beneficial and can be supported.

For NR-U DL operation, it is identified that being able to operate all DL signal/channels with the same numerology for a carrier and at least for intra-band CA on serving cells on unlicensed bands has at least the following benefits (at least for standalone operation)

-    Lower implementation complexity (e.g., a single FFT, no switching gaps)

-    Lower specification impact

-    No need for gaps for measurements on frequencies with a configured serving cell in unlicensed bands

For NR-U UL operation, it is identified that being able to operate all UL signal/channels (except PRACH) with the same numerology for a carrier and at least for intra-band CA on serving cells on unlicensed bands has at least the following benefits:

-    Lower implementation complexity (e.g., a single FFT, no switching gaps)

-    Lower specification impact

-    Common interlace structure

-    No need for gaps for transmission of SRS on a configured serving cell in unlicensed bands

For unlicensed PCell, the UE assumes single SSB numerology per band.

It has been identified to be beneficial for the NR-U design to not require the gNB to change a pre-determined TBS for a PDSCH transmission depending on the LBT outcome, at least when the PDSCH is transmitted at the beginning of the gNB's COT.

The following options have been identified as possible candidates for PDSCH transmission in the partial slot at least for the first PDSCH(s) transmitted in the DL transmission burst. The options are not mutually exclusive.

-    Option 1: PDSCH(s) as in Rel-15 NR

-    Option 2: Punctured PDSCH depending on LBT outcome

-    Option 3: PDSCH mapping type B with durations other than 2/4/7 symbols

-    Option 4: PDSCH across slot boundary

In addition to the functionalities provided by DCI format 2_0 in Rel-15 NR, indication of the COT structure in the time domain has been identified as being beneficial.

It has been identified to be beneficial for the NR-U design to not require the UE to change a granted TBS for a PUSCH transmission depending on the LBT outcome.

The following options have been identified as possible candidate at least for the first PUSCH(s) transmitted in the UL transmission burst.

-    Option 1: PUSCH(s) as in Rel-15 NR

-    Option 2: Multiple starting positions in one or multiple slot(s) are allowed for PUSCH(s) scheduled by a single UL grant (i.e., not a configured grant) and one of the multiple PUSCH starting positions can be decided depending on LBT outcome.

For above options, the ending position of the PUSCH is fixed as indicated by the UL grant.

 

It has been identified that FBE operation for the scenario where it is guaranteed that LBE nodes are absent on a long term basis (e.g., by level of regulation) and FBE gNBs are synchronized can achieve the following: Ability to use frequency reuse factor 1; Lower complexity for channel access due to lack of necessity to perform random backoff.

It is noted that this does not imply that LBE does not have benefits in similar scenarios although there are differences between the two modes of operation. It is also noted that FBE may also have some disadvantages compared to other modes of operation such as LBE, e.g., a fixed overhead for idle time during a frame.

For wideband operation for both DL and UL,

-    Bandwidth larger than 20 MHz can be supported with multiple serving cells.

-    NR-U should support that a serving cell can be configured with bandwidth larger than 20 MHz.

For DL operation, the following options for BWP-based operation within a carrier with bandwidth larger than 20 MHz can be considered.

-    Option 1a: Multiple BWPs configured, multiple BWPs activated, transmission of PDSCH on one or more BWPs

-    Option 1b: Multiple BWPs configured, multiple BWPs activated, transmission of PDSCH on single BWP

-    Option 2: Multiple BWPs can be configured, single BWP activated, gNB transmits PDSCH on a single BWP if CCA is successful at gNB for the whole BWP

-    Option 3: Multiple BWPs can be configured, single BWP activated, gNB transmits PDSCH on parts or whole of single BWP where CCA is successful at gNB

For UL operation, the following options for BWP-based operation within a carrier with bandwidth larger than 20 MHz can be considered.

-    Option 1a: Multiple BWPs configured, multiple BWPs activated, transmission of PUSCH on one or more BWPs

-    Option 1b: Multiple BWPs configured, multiple BWPs activated, transmission of PUSCH on single BWP

-    Option 2: Multiple BWPs can be configured, single BWP activated, UE transmits PUSCH on a single BWP if CCA is successful at UE for the whole BWP

-    Option 3: Multiple BWPs can be configured, single BWP activated, UE transmits PUSCH on parts or whole of single BWP where CCA is successful at UE

CCA is declared to be successful or not in multiples of 20 MHz.


Related:

  1. NR-U Inactive and Idle procedures (Candidates)
  2. NR-U Control plane (Candidates)

  3. NR-U Layer 2 (Candidates)

  4. NR-U Physical layer channel designs(Candidates)

  5. NR-U Channel Access Schemes (Candidates)

NR-U Channel Access Schemes (Candidates)


 Note: NR-Unlicensed is still in the study phase with barely any work done on the TS. It is expected that by 2022 (so Release 18) we may see some specification for the 6GHz band in Europe. As of now, only potential candidates for NR-U implementation can be guessed. 

The channel access candidate schemes for NR-based access for unlicensed spectrum can be classified into the following categories:

-    Category 1: Immediate transmission after a short switching gap

 -     This is used for a transmitter to immediately transmit after a switching gap inside a COT.

-     The switching gap from reception to transmission is to accommodate the transceiver turnaround time and is no longer than 16 µs.

-    Category 2: LBT without random back-off

-     The duration of time that the channel is sensed to be idle before the transmitting entity transmits is deterministic.

-    Category 3: LBT with random back-off with a contention window of fixed size

-     The LBT procedure has the following procedure as one of its components. The transmitting entity draws a random number N within a contention window. The size of the contention window is specified by the minimum and maximum value of N. The size of the contention window is fixed. The random number N is used in the LBT procedure to determine the duration of time that the channel is sensed to be idle before the transmitting entity transmits on the channel.

-    Category 4: LBT with random back-off with a contention window of variable size

-     The LBT procedure has the following as one of its components. The transmitting entity draws a random number N within a contention window. The size of contention window is specified by the minimum and maximum value of N. The transmitting entity can vary the size of the contention window when drawing the random number N. The random number N is used in the LBT procedure to determine the duration of time that the channel is sensed to be idle before the transmitting entity transmits on the channel.

For different transmissions in a COT and different channels/signals to be transmitted, different categories of channel access schemes can be used.

Related:

  1. NR-U Inactive and Idle procedures (Candidates)
  2. NR-U Control plane (Candidates)

  3. NR-U Layer 2 (Candidates)

  4. NR-U Physical layer channel designs(Candidates)

  5. NR-U Physical Frame structure (Candidates)

Sunday, November 22, 2020

NR-U Inactive and Idle procedures (Candidates)


Note: NR-Unlicensed is still in the study phase with barely any work done on the TS. It is expected that by 2022 (so Release 18) we may see some specification for the 6GHz band in Europe. As of now, only potential candidates for NR-U implementation can be guessed. 

Inactive and Idle procedures (Candidates)

For Inactive and Idle mode procedures, Rel-15 NR design is considered as the baseline and as such, NR licensed measurement framework (cell and beam quality derivation for RSRP, RSRQ, and SINR, filtering and combining multiple beams) is also used as the baseline.

The UE measurements in Idle/Inactive mode will mostly assume recurring transmissions of SSB/PBCH and RMSI but possibly with reduced opportunities due to LBT.

In unlicensed bands, multiple PLMNs can use the same carriers without any coordination. Therefore, the best cell found by a UE on a frequency may not belong the registered PLMN. In this case, the UE can be enabled to camp on a non-best cell on a carrier if the best cell does not belong to the registered PLMN (or E-PLMN), where the non-best cell would still be the best cell of the registered PLMN.

For paging, it may be beneficial to introduce more opportunities per DRX cycle for the UE to receive the page. The additional locations can be provided in time domain by configuring an extended paging occasion (i.e. a paging window) or configuring multiple paging occasions to a UE. In any specified solution(s) based on additional paging opportunities, the UE power consumption will also be taken into account, so it would be beneficial that the paging occasions are transmitted in close time to or overlap with the reference signals.

Related Posts

Monday, November 16, 2020

Mobility enhancement in E-UTRAN Release -16


 Dual Active Protocol Stack (DAPS) handover

DAPS Handover is a handover procedure that maintains the source eNB connection after reception of RRC message for handover and until releasing the source cell after successful random access to the target eNB.

-    If DAPS handover is configured, the UE continues the downlink user data reception from the source eNB until releasing the source cell and continues the uplink user data transmission to the source eNB until successful random access procedure to the target eNB.

-    Upon reception of the handover command, the UE:

-     Creates a MAC entity for target cell;

-     Establishes the RLC entity and an associated DTCH logical channel for target cell for each DRB configured with DAPS;

-     For the DRB(s) configured with DAPS, reconfigures the PDCP entity to configure DAPS with separate security and ROHC functions for source and target and associates them with the RLC entities configured for source and target respectively;

-     Retains rest of the source link configurations until release of the source.

-     When DAPS handover fails, the UE falls back to source cell configuration, resumes the connection with source cell, and reports the DAPS handover failure via the source without triggering RRC connection re-establishment if the source link is still available; Otherwise, RRC re-establishment is performed;

Conditional Handover (CHO)

A Conditional Handover (CHO) is defined as a handover that is executed by the UE when one or more handover execution conditions are met.

-    UE maintains connection with source eNB after receiving CHO configuration, and starts evaluating the CHO execution condition(s) for the CHO candidate cell(s) and executes the HO command once the execution condition(s) are met for a CHO candidate cell.

-    To improve the robustness, the network can provide the up to 8 candidate cell configuration(s) associated with execution condition (s) to UE. If at least one CHO candidate cell satisfies the corresponding CHO execution condition, the UE detaches from the source eNB, applies the stored corresponding configuration for that candidate cell and synchronises to that candidate cell. UE stops evaluating the execution condition(s) for other candidate cells once the handover is triggered.

-    The UE accesses to the target eNB and completes the handover procedure by sending RRCConnectionReconfigurationComplete message to target eNB. The UE releases stored CHO configurations after successful completion of RRC handover procedure.

-    When initial CHO execution attempt fails or HO fails, if network configured the UE to try CHO after HO/CHO failure and the UE performs cell selection to a CHO candidate cell, the UE attempts CHO execution to that cell; Otherwise, RRC re-establishment is performed.

DL MIMO efficiency enhancements for LTE


MIMO is an effective technique to improve spectral efficiency and increase overall network capacity. SRS can be utilized to improve DL MIMO performance, especially for massive MIMO in TDD. In release-16 SRS capacity and coverage are enhanced by introducing more than one SRS symbol in a UL normal subframe and introducing virtual cell ID for SRS.

More than one symbol for SRS in a UL normal subframe

With the introduction of more than one SRS symbol in a UL normal subframe, the SRS capacity and coverage can be increased. These additional SRS symbol(s) are referred to as trigger type 2 SRS.

1 to 13 symbols of the first 13 symbols of a UL normal subframe can be configured to a UE for aperiodically triggered SRS transmission. Intra-subframe repetition, frequency hopping and antenna switching of the additional SRS symbols can be supported. A guard period of one SC-FDMA symbol can be configured for frequency hopping and antenna switching.

The number of repetitions can be configured from the set {1, 2, 3, 4, 6, 7, 8, 9, 12, 13}. If a UE has more receive antennas than transmit chains (e.g. 1T2R), the UE can be configured to transmit the additional SRS with antenna switching. And a UE can additionally be configured with frequency hopping for the additional SRS. The number of antennas (pairs) to switch is:

-    2 for 1T2R, or the number of pairs is configured as 2 for 2T4R

-    3 if the number of pairs is configured as 3 for 2T4R

-    4 for 1T4R

Both legacy SRS and additional SRS can be configured to the same UE, and transmission of legacy SRS and additional SRS symbol(s) in the same subframe for the UE is supported.

For sequence generation, per-symbol group hopping and sequence hopping are supported.

Independent open loop and close loop power control is supported for additional SRS, and DCI formats 3/3A/3B are used for close loop power control.

UEs not configured with SPUCCH/SPUSCH are not expected to be triggered with additional SRS in the same subframe as PUSCH/PUCCH in the same serving cell. UEs configured with SPUCCH/SPUSCH drop the SRS transmission in the symbols colliding with SPUCCH/SPUSCH in the same serving cell.

The additional SRS transmission in the symbols colliding with PUSCH/PUCCH of another serving cell in the same TAG, the same band and with the same CP, is dropped.

The additional SRS transmission in the symbols colliding with PUSCH/PUCCH of another serving cell in the different TAG is dropped if the total transmission power exceeds the PCMAX in any overlapping portion.

Virtual cell ID

Virtual cell ID within the range from 0 to 503 can be configured for SRS to increase SRS capacity.

The virtual cell ID can be configured to only additional SRS symbol(s) or both legacy and additional SRS symbol(s). If virtual cell ID is not configured, the physical cell ID is used.

Optimisations of UE radio capability signalling


UE Radio Capability ID

UE Capability Segmentation

Background

With the increase in the size of UE radio capabilities driven by additional supported bands, the size of the UE Radio Capabilities will significantly grow from Rel-15 onwards, therefore an efficient approach to signal UE Radio Capability information was needed.

UE Radio Capability ID

The system optimisations for the 5GS (documented in TS 23.501) and for the EPS (documented in TS 23.401), that apply to both NR and E-UTRA, but not NB-IoT, consisting of using UE Radio Capability IDs as an alternative to signaling the UE Radio Capabilities container in system procedures:

-  between the UE and the CN (over Uu)

- between the CN and the RAN (impacting N2/S1 interfaces)

-  within the RAN in e.g. the handover procedures (impacting Xn/X2/S1/N2 interfaces)

-  within the CN.

The UE Radio Capability ID format is defined in TS 23.003. The UE Radio Capability ID is signaled by the UE in NAS as specified in TS 24.501 for the 5GS and as specified in TS 24.301 for the EPS. Two possible options for the assignment of UE Radio Capability ID exist:

-  Manufacturer-assigned: The UE Radio Capability ID may be assigned by the UE manufacturer in which case it includes a UE manufacturer identification (i.e. a Vendor ID). In this case, the UE Radio Capability ID uniquely identifies a set of UE radio capabilities for a UE by this manufacturer in any network.

-  Network-assigned: If a manufacturer-assigned UE Radio Capability ID is not used by the UE or the serving network, or it is not recognised by the serving network’s UE Capability Management Function (UCMF), the UCMF may allocate UE Radio Capability IDs for the UE corresponding to each different set of UE radio capabilities which the network may receive from the UE at different times. In this case, the UE Radio Capability IDs which the UE receives are applicable to the serving network and uniquely identify the corresponding sets of UE radio capabilities in this network. The network-assigned UE Radio Capability ID includes a Version ID in its format. The value of the Version ID is the one configured in the UCMF, at the time when the UE Radio Capability ID value is assigned. The Version ID value makes it possible to detect whether a UE Radio Capability ID is current or outdated.

 

UE Radio Capability IDs and the mapping to the corresponding UE radio capabilities are stored in a new function called the UE Capability Management Function (UCMF) in the CN. The UCMF is used for:

- storage of dictionary entries corresponding to either Network-assigned or Manufacturer-assigned UE Radio Capability IDs.

- assigning Network-assigned UE Radio Capability ID values.

-  provisioning of Manufacturer-assigned UE Radio Capability ID entries in the UCMF performed from an AF that interacts with the UCMF either directly or via the NEF/SCEF (or via Network Management).

    

UCMF architecture and related reference points in 5GS (left) and EPS (right)
UCMF architecture and related reference points in 5GS (left) and EPS (right)

System procedures are defined for 5GS in TS 23.502 and for EPS in TS 23.401.

UE Capability Segmentation

The RAN work item [10] calls for specification of a segmentation mechanism, so that in cases of excessively large UE capability signalling (e.g. capability information messages exceeding the maximum size of a PDCP SDU), the capability can be segmented into multiple RRC messages.  Segmentation applies to both NR and E-UTRA.

Segmentation is performed in the RRC protocol layer, with a separate RRC PDU for each segment.  The UE encodes the capability information message, then divides the encoded message into segments such that the size of each segment does not exceed the maximum size of a PDCP SDU (8188 octets in E-UTRA, 9000 octets in NR); the RAN node (eNB or gNB) receives the segments and reassembles them to reconstruct the original capability information message.  Segmentation is applied only in case the size of the encoded capability information message exceeds the maximum size of a PDCP SDU.  The signalling formats support up to 16 segments for a single capability information message.

Network Slicing


 Release 16 Network Slicing addresses two major limitations of Release 15 in 5GC:

(1) Enhancement of interworking between EPC and 5GC when UE moves from EPC to 5GC, the target serving AMF may not be able to serve all the PDU sessions that the UE intends to move to the 5GC.  More specifically, the following aspects needs to be addressed:

-              Selecting an AMF based on the slices associated to the active PDU connections that serve the UE in the EPC in Connected mode and during the Idle mode mobility

-              Selecting an appropriate serving V-SMF based on the slices associated to the active PDN connections that serve the UE in the EPC in Connected Mode

 (2) Support for Network Slice Specific Authentication and Authorization (NSSAA)

-              Enable the support for separate authentication and authorization per Network Slice.  The trigger of NSSAA in the 5GC is based on UE subscription information from UDM and also operator’s policy.  However, the UE shall indicate its support for NSSAA to its serving 5GC.

-              The AMF performs the role of the EAP Authenticator and communicates with the AAA-S via the AUSF. The AUSF undertakes any AAA protocol interworking with the AAA protocol supported by the AAA-S.

Enhancement of Interworking Between EPC and 5GC

During the mobility from EPS to 5GS, in case of CM-IDLE state, the PGW-C+SMF sends PDU Session IDs and related S-NSSAIs to AMF in Registration procedure. The AMF derives S-NSSAI values for the Serving PLMN and determines whether the current AMF is appropriate to serve the UE. If not, the AMF reallocation may need to be triggered. For each PDU Session the AMF determines whether the V-SMF need to be reselected based on the associated S-NSSAI value for the Serving PLMN. If the V-SMF need be reallocated, the AMF trigger the V-SMF reallocation.

In case of CM-CONNECTED state, during handover preparation phase the PGW-C+SMF sends PDU Session IDs and related S-NSSAIs to AMF. Based on the received S-NSSAIs values, the target AMF derives the S-NSSAI values for the Serving PLMN, the target AMF reselects a final target AMF if necessary and forwards the handover request to the final target AMF. When the Handover procedure completes successfully, the UE proceeds with the Registration procedure. For each PDU Session based on the associated derived S-NSSAI values, if the V-SMF need be reallocated, the final target AMF triggers the V-SMF reallocation. The final target AMF sends the S-NSSAI value for the Serving PLMN to V-SMF to update the SM context. The V-SMF updates NG RAN with the S-NSSAI value for the Serving PLMN via N2 SM message.

 

Network Slice-Specific Authentication and Authorization (NSSAA)

In Release-16, based on UE’s 5GMM Core Network Capability and subscription information, the serving AMF will trigger Network Slice-Specific Authentication and Authorization for the S-NSSAIs of the HPLMN. If a UE does not support this feature but requests these S-NSSAIs that are subject to Network Slice-Specific Authentication and Authorization, these S-NSSAIs will be rejected by the PLMN.

If a UE supports this feature and requests these S-NSSAIs, which are subject to Network Slice-Specific Authentication and Authorization, the UE shall leverage the corresponding credentials for these S-NSSAIs for the Network Slice-Specific Authentication and Authorization. As for how to these credentials in the UE are not specified.

To perform the Network Slice-Specific Authentication and Authorization for an S-NSSAI, the AMF invokes an EAP- based Network Slice-Specific authorization procedure for the S-NSSAI.

This procedure can be invoked for a supporting UE by an AMF at any time, e.g. when:

a.            The UE registers with the AMF and one of the S-NSSAIs of the HPLMN which maps to an S-NSSAI in the Requested NSSAI is requiring Network Slice-Specific Authentication and Authorization; or

b.            The Network Slice-Specific AAA Server triggers a UE re-authentication and re-authorization for an S-NSSAI; or

c.            The AMF, based on operator policy or a subscription change, decides to initiate the Network Slice-Specific Authentication and Authorization procedure for a certain S-NSSAI which was previously authorized.

Based on the outcome of the Network Slice-Specific Authentication and Authorization, the Allowed NSSAI for each Access Type will be updated accordingly.   It is network policies to decide for which Access Type to be used if both Access Types are subject for the Network Slice-Specific Authentication and Authorization. However, if the Network Slice-Specific Authentication and Authorization fails for all S-NSSAIs in the Allowed NSSAI, the AMF shall execute the Network-initiated Deregistration procedure with the appropriate rejection cause value for each Rejected S-NSSAI.

After a successful or unsuccessful UE Network Slice-Specific Authentication and Authorization, the UE context in the AMF shall retain the authentication and authorization status for the UE for the related specific S-NSSAI of the HPLMN while the UE remains RM-REGISTERED in the PLMN, so that the AMF is not required to execute a Network Slice-Specific Authentication and Authorization for a UE at every Periodic Registration Update or Mobility Registration procedure with the PLMN.

A Network Slice-Specific AAA server may revoke the authorization or challenge the authentication and authorization of a UE at any time. When authorization is revoked for an S-NSSAI that is in the current Allowed NSSAI for an Access Type, the AMF shall provide a new Allowed NSSAI to the UE and trigger the release of all PDU sessions associated with the S-NSSAI, for this Access Type.

The AMF provides the GPSI of the UE related to the S-NSSAI to the AAA Server to allow the AAA server to initiate the Network Slice-Specific Authentication and Authorization, or the Authorization revocation procedure, where the UE current AMF needs to be identified by the system, so the UE authorization status can be challenged or revoked.

The Network Slice-Specific Authentication and Authorization requires that the UE Primary Authentication and Authorization of the SUPI has successfully completed. If the SUPI authorization is revoked, then also the Network Slice-Specific authorization is revoked.