Uplink Data Switching and UDC in NR PDCP

Quick facts

Contents

1. Introduction to Uplink Data Switching and UDC
2. What This Page Covers
3. Where These Functions Sit in PDCP
4. Uplink Data Switching
5. Uplink Data Switching Example
6. UDC Protocol
7. Configuration of UDC
8. Configuration Examples
9. UDC Header
10. UDC Packet and Header Examples
11. Uplink Data Compression
12. Compression Flow Example
13. PDCP Control PDU for UDC Feedback
14. UDC Feedback Example
15. Pre-defined Dictionary Handling
16. Dictionary Example
17. UDC Buffer Reset Procedure
18. UDC Checksum Error Handling
19. Buffer Reset and Checksum Example
20. Annex B Packet Reference
21. What Should Stay Off This Page
22. References
23. FAQ
24. Related PDCP Pages

Introduction to Uplink Data Switching and UDC

These two topics sit together well because both are uplink-side PDCP behaviors that affect how already-known PDCP SDUs are treated before or while they move toward lower layers. Uplink data switching is the DAPS-bearer path-move procedure. UDC is the DRB-oriented compression procedure that adds a UDC header, manages a compression buffer, and exchanges feedback.

This is the page to open when the question is “what happened to pending uplink PDCP traffic?” rather than only “what does this header field mean?” For field-level decoding, continue into the PDU format page.

What This Page Covers

Where These Functions Sit in PDCP

Uplink data switching is a DAPS transmit-side procedure. UDC is a user-plane compression function. They belong together here because they both affect uplink processing before lower-layer submission, and both need to be read together with header compression, ciphering, and integrity ordering.

Uplink Data Switching

Uplink data switching is defined for DAPS bearers. The reader should separate the AM DRB and UM DRB cases, because the switching starting point is different. In both cases, pending uplink work is redirected toward the target-cell RLC entity while preserving the already associated COUNT values.

Uplink Data Switching Example

Example 1: AM DRB on a DAPS bearer

Initial condition
- UE has a DAPS bearer
- source-cell path is active
- some PDCP SDUs already have PDCP SN and COUNT
- delivery of some PDCP Data PDUs is not yet confirmed by the source-cell RLC entity

Trigger
- upper layers request uplink data switching

Illustrative PDCP behavior
- find the first PDCP SDU whose corresponding PDCP Data PDU is not yet confirmed by source-cell RLC
- retransmit or transmit all already-associated PDCP SDUs in ascending COUNT order toward the target-cell RLC entity
- apply header compression
- apply integrity protection and ciphering using the COUNT already associated with each PDCP SDU
- submit the resulting PDCP Data PDUs to lower layers

Example 2: UM DRB on a DAPS bearer

Initial condition
- UE has a DAPS bearer using UM DRB
- some PDCP SDUs have been processed by PDCP but not yet submitted to lower layers

Trigger
- upper layers request uplink data switching

Illustrative PDCP behavior
- transmit the pending PDCP SDUs in ascending COUNT order toward the target-cell RLC entity
- apply header compression
- apply integrity protection and ciphering using the COUNT already associated with each PDCP SDU
- submit the resulting PDCP Data PDUs to lower layers

UDC Protocol

UDC is the uplink data-compression topic on this page. It is based on DEFLATE with static Huffman coding and uses a small PDCP-facing packet model built around a 1-byte UDC header.

Configuration of UDC

UDC is configured for DRBs by upper layers. The main things to check are the configured bufferSize and whether a pre-defined dictionary is present.

Configuration Examples

Example 1: DRB with UDC enabled

bearerType: DRB
userPlaneData: true
uplinkDataCompression:
  protocol: UDC
  bufferSize: <configured by upper layers>
  preDefinedDictionary: none
result:
  - compression buffer is set to all zeros
  - UDC compression function is applied to PDCP SDUs from upper layers on this DRB

Example 2: DRB with UDC and pre-defined dictionary

bearerType: DRB
userPlaneData: true
uplinkDataCompression:
  protocol: UDC
  bufferSize: <configured by upper layers>
  preDefinedDictionary: <configured by upper layers>
result:
  - compression buffer is first reset to all zeros
  - configured dictionary is then prefixed into the compression buffer

Example 3: Editorial note

These are illustrative configuration patterns only. Keep actual ASN.1 and RRC field rendering on a separate configuration-reference page if needed.

UDC Header

The UDC header is the compact PDCP-side marker that tells the peer whether the payload is compressed, whether reset signaling is present, and which checksum information should be interpreted. It also shows where the UDC data starts relative to the PDCP header and optional SDAP header.

UDC Packet and Header Examples

Example 1: UDC packet with uncompressed payload

Illustrative UDC packet
[ UDC Header ][ original PDCP SDU payload ]

Header interpretation
- FU = 0
- payload is not compressed by UDC
- packet still belongs to one PDCP SDU

Example 2: UDC packet with compressed payload

Illustrative UDC packet
[ UDC Header ][ DEFLATE-compressed data block ]

Header interpretation
- FU = 1
- payload is compressed by UDC
- checksum and reset signaling may be relevant

Example 3: Header placement with SDAP

Illustrative PDCP Data PDU ordering
[ PDCP header ][ SDAP header if configured ][ UDC header ][ UDC data block ]

Uplink Data Compression

The main reading rule is straightforward: one UDC packet is associated with one PDCP SDU, and that UDC packet keeps the same PDCP SN and COUNT as the related PDCP SDU. UDC is not applied to the SDAP header or SDAP Control PDU.

Compression Flow Example

Example 1: UDC compression path

PDCP SDU received from upper layers
→ UDC compression function checks buffer and dictionary context
→ UDC header is added
→ UDC data block is generated
→ resulting UDC packet keeps the same PDCP SN and COUNT as the related PDCP SDU
→ PDCP Data PDU is submitted to lower layers

Example 2: Non-compressed UDC packet

PDCP SDU received from upper layers
→ UDC decides not to compress this packet
→ UDC header still indicates non-compressed packet
→ original payload is carried as the UDC data block

PDCP Control PDU for UDC Feedback

UDC feedback is carried in a PDCP Control PDU applicable to AM DRBs. The useful receive-side rule is simple: the PDCP entity delivers the feedback directly to the UDC protocol without performing deciphering or integrity verification.

UDC Feedback Example

Example feedback path

Step 1: A PDCP Control PDU carrying one UDC feedback item is received from lower layers.
Step 2: The receiving PDCP entity does not perform deciphering.
Step 3: The receiving PDCP entity does not perform integrity verification.
Step 4: The receiving PDCP entity delivers the UDC feedback directly to the UDC protocol.

Mini format example

Illustrative PDCP Control PDU
[ D/C ][ PDU Type ][ FE ][ R ][ R ][ R ]

Pre-defined Dictionary Handling

The dictionary view is about how the compression buffer is initialized and how that starting context evolves over time. The key things to check are which dictionary is configured, whether the dictionary fits inside the buffer, and how FIFO behavior eventually moves old dictionary content out.

Dictionary Example

Example 1: SIP/SDP dictionary case

Illustrative behavior
- UDC is configured with the standard SIP/SDP dictionary
- the dictionary is inserted into the tail of the compression buffer
- as more uncompressed data is processed, FIFO behavior eventually pushes the dictionary content out

Example 2: Dictionary larger than buffer

Illustrative behavior
- configured dictionary size > compression buffer size
- only the tail of the dictionary is inserted into the compression buffer

UDC Buffer Reset Procedure

Buffer reset is the controlled way to recover from compression and decompression buffer mismatch. The main point is that synchronization has to be restored explicitly, and after reset the first compressed PDU marks that new synchronized starting point.

UDC Checksum Error Handling

Checksum error handling is the main signal that the compression and decompression buffers have moved out of synchronization. That indication leads into the UDC buffer reset procedure.

Buffer Reset and Checksum Example

Initial condition
- compression buffer and decompression buffer become unsynchronized

Detection
- UDC checksum error notification indicates mismatch

Illustrative behavior
- UE triggers UDC buffer reset procedure
- compression buffer is reset to all zeros
- after reset, the FR field in the first compressed PDU is set to 1

Annex B Packet Reference

Annex B is the supporting detail for the UDC packet and header model. Use it for the UDC header, UDC data block, FU, FR, and checksum bit details. This page keeps Annex B as supporting detail rather than making it the main backbone.

What Should Stay Off This Page

References

• ETSI TS 138 323 V18.5.0, NR PDCP specification
• 3GPP TS 38.331 V18.5.1, NR RRC configuration context for DAPS and UDC-related PDCP settings

FAQ

What is uplink data switching in NR PDCP?

Uplink data switching is the DAPS-bearer PDCP procedure that moves pending uplink traffic from the source-cell path to the target-cell path.

Which bearers use uplink data switching?

Uplink data switching is defined for DAPS bearers.

Which bearers can use UDC?

UDC is configured for DRBs by upper layers in this PDCP context.

What is bufferSize in UDC configuration?

bufferSize is the upper-layer configured size of the UDC compression buffer context.

Where is the UDC header placed in a PDCP Data PDU?

The UDC header is placed after the SDAP header if SDAP is configured.

Does a UDC packet keep the same PDCP SN and COUNT as the original PDCP SDU?

Yes. A UDC packet keeps the same PDCP SN and COUNT as the related PDCP SDU.

How is UDC feedback carried in PDCP?

UDC feedback is carried in a PDCP Control PDU applicable to AM DRBs.

What happens when a UDC checksum error is detected?

A checksum error indicates buffer desynchronization and triggers the UDC buffer reset procedure.

Related PDCP Pages