7 HARQ protocol
25.3083GPPHigh Speed Downlink Packet Access (HSDPA)Overall descriptionRelease 17Stage 2TS
The HARQ protocol is based on an asynchronous downlink and synchronous uplink scheme. The ARQ combining scheme is based on Incremental redundancy. Chase Combining is considered to be a particular case of Incremental Redundancy. The UE soft memory capability shall be defined according to the needs for Chase combining. The soft memory is partitioned across the HARQ processes in a semi-static fashion through upper layer signalling. The UTRAN should take into account the UE soft memory capability when configuring the different transport formats (including possibly multiple redundancy versions for the same effective code rate) and when selecting transport formats for transmission and retransmission.
If blind HARQ retransmissions for HSDPA are configured by upper layers, when the UE receives a MAC PDU within 5 sub-frames from the reception of the previous MAC PDU intended for the same HARQ process, the UE combines the received data with the data in the soft buffer and generates a feedback for this HARQ process.
7.1 Signalling
7.1.1 Uplink
In the uplink, a report is used indicating either ACK (positive acknowledgement) or NACK (negative acknowledgement).
7.1.2 Downlink
7.1.2.1 Shared control channel signalling
The following HARQ protocol parameters are carried on the HS-SCCH:
– HARQ process identifier:
– Every HARQ process is assigned an identifier, which is used to couple the processes in the transmitter and the receiver. For dual stream, three stream and four stream transmissions a pair of HARQ process identifiers is indicated by the HS-SCCH.
– New data indicator:
– It is used to distinguish between data blocks. It is specific to the HARQ process. It is incremented for each new data block. For MIMO mode and MIMO mode with four transmit antennas, the new data indicator is implied by the redundancy version indicator carried on the HS-SCCH.
7.1.2.2 In-band signalling on HS-DSCH
The following parameters are signalled in-band in the MAC-hs header to support in-sequence delivery and priority handling at the UE. These parameters are protected by the same CRC as the Data block. For MAC-hs the parameters are:
– Re-ordering Queue Identity:
– It is used to identify the re-ordering buffer destination of a MAC-hs PDU.
– Transmission sequence number:
– It is incremented for each new data block destined to a re-ordering buffer. It is used for reordering to support in-sequence delivery.
For MAC-ehs, the parameters are:
– Logical Channel Identity:
– It is used to identify the logical channel and the re-ordering buffer destination of a MAC-ehs SDUs or segments of MAC-ehs SDUs.
– Transmission sequence number:
– It is used for reordering to support in-sequence delivery.
– Length of the MAC-ehs SDU or the segment of the MAC-ehs SDU included in the MAC-ehs PDU.
Additionally the following parameter is signalled in-band in the MAC-ehs header to support MAC-ehs SDU segmentation in Node B and reassembly at the UE.
– Segmentation indication.
7.2 Void
7.3 Void
7.4 Error handling
The most frequent error cases to be handled are the following:
– NACK is detected as an ACK. The NW starts afresh with new data in the HARQ process. The data block is discarded in the NW and lost. Retransmission is left up to higher layers.
– ACK is detected as a NACK: If the network retransmits the data block, the UE will re-send an ACK to the network. If in this case the transmitter at the network sends an abort indicator by incrementing the New Packet Indicator, the receiver at the UE will continue to process the data block as in the normal case.
– If a CRC error on the HS-SCCH is detected, UE receives no data and sends no status report. If the absence of the status report is detected, NW can retransmit the block.