5G single-stage and two-stage DCI
Downlink Control channel is an important new one for NR. Compared with LTE, NR has stricter requirements on Control signaling. DCI (Downlink Control Information) design is an important part to support NR requirements, such as forward compatibility, flexibility, low delay, transmission based on multi-beam, low energy consumption,In NR, the names for single-stage DCI and two-stage or multistage DCI are slightly different.From a gNB perspective, downlink control signaling can be located at the first OFDM symbol in a timeslot or mini-slot.NR supports at least downlink scheduling in the same time slot and across time slot.There can be one or more timing relationships between downstream data reception and corresponding validation. Which one is it?Reference: Dynamic indication through L1 signaling (e.g., DCI) the combination of semi-static indication to UE high-level indication through high level and dynamic L1 signaling (e.g., DCI) introduces single-stage DCI and two-stage or multistage DCI in NR.In single-stage DCI, all DCI content is transmitted in a single control channel (NR PDCCH), while in two-stage or multi-stage DCI, the first-level DCI is transmitted in the control channel (NR PDCCH1) and the second or higher level DCI content is transmitted in one or more separate channels.The channel carrying the second or higher level DCI can be an additional control channel (NR PDCCH2) or a data channel (NR PDSCH).One of the main advantages of single-stage DCI is spectral efficiency, as the CRC overhead of the control channel can be limited to just one data transfer per time.In addition, it can be beneficial in terms of reliability, because once a single DCI is detected, the data can be decoded, and a single DCI is usually protected more strongly than data.The potential disadvantage of single-stage DCI is that it may not effectively accommodate variable DCI sizes unless the number of blind decoding trials is increased.Another potential disadvantage of this approach is that it may not be easy to implement using higher MCS transport and DMRS with data sharing.While multilevel DCI designs could be further considered, single-level DCI should be supported as a baseline, as it is the proven solution for most use cases.In particular, for UL Grant, single-level DCI seemed a natural choice.One potential situation where a two-level DCI might be useful is as follows.First, if the UE needs to change the numerology used for data decoding or adjust the bandwidth it uses for data reception, a very short delay is expected to decode the control channel.If a two-stage DCI is used, with a potentially small BD on the first stage, and the first stage DCI is sent only at the beginning of the time slot, then the control decoding delay can be reduced and the gap between the control and the data for bandwidth/Numerology adaptive can be minimized.Second, if the control area is fairly fixed, or the maximum size is insufficient for the required control channel capacity, more control channels can be accommodated into the first level control area by making the first level DCI smaller and offloading the DCI content to the second level.However, similar problems can be solved by placing emergency controls (for fast decoding) at the beginning of the timeslot/search space, and additional control subbands may be configured if capacity becomes an issue.In this sense, no key reason has been found to support two-level DCI so far.Even if two-level DCI is supported, single-level DCI and multilevel DCI should coexist in order to support, for example, UL Grant.If two-level DCI is considered, the DCI content is divided into two levels and transmitted in a separate channel or in different resources on the same channel.While the first level DCI can be carried in a control area, the second level DCI can be carried in another control channel or in a different resource in the same control channel or in a data channel.Depending on the channel of the second-level DCI, several options can be considered for the two-level DCI.Option 1: The first stage DCI in the control channel (NR PDCCH1), the second stage DCI in the control channel (NR PDCCH2) In option 1, carry the first and second stage DCI in the NR physical downlink control channel.The location of the second level control channel (NR PDCCH2) can be predefined or indicated either by higher level signaling or more dynamically by the first level DCI.With this option, DMRS may or may not be shared between NR PDCCH2 and PDSCH.For example, this option will be valid when slot and mini-slot schedules are multiplexed, and the indication of the control area of the Mini-slot can be done in the first level DCI in the slot.In other words, the first-level DCI can indicate the resource location of the control area of the second-level DCI for mini-slot scheduling.Option 2: The first stage DCI in the control channel (NR PDCCH1), the second stage DCI in the data channel (NR PDSCH). Option 2: the second stage DCI in the data channel (NR PDSCH) is carried with the data.The exact time/frequency position of the second level DCI can be predefined relative to the NR PDSCH region or indicated by the first DCI.In option 2, the DMRS used for data demodulation can also be used for secondary DCI demodulation.When using this option, many aspects should be clarified, such as SU-MIMO, MU-MIMO, DMRS sharing in multiple layers, MCS of second-level DCI, processing of single or joint encoding between second-level DCI and data.Depending on where control and data are multiplexed in channel processing, different options and performance are expected.For example, control and data can share the same MCS or use different modulation (for example, fixed QPSK for control).