In order to avoid this situation, non-UE-specific (i.e., cell-specific) beam training is required. All UE needs to know the timing of cell-specific RS, so it is not appropriate to send the timing through PDC. Other options for transmitting the timing information include RRC signaling, MIB/SIB or designation in the air interface.
Because RS is useful for the maintenance of candidate beams, it is more beneficial to appear these RS periodically. These RS enable the UE to discover and monitor multiple candidate beams of any alternative feasible path (for example, diversity to overcome congestion, beam combination, neighbor cell discovery and handover, etc.). ). These beams may come from several Trps in the cell. This facilitates "beam switching" from one TRP to another when the UE crosses the cell. Therefore, this RS is called mobility reference signal (Mrs).
There are several other advantages to using MRS to perform such tasks. MRS can also help to measure the CFO of other Trps and neighboring cells. It is a feasible choice to let CFO make proportional numerology estimation on a regular basis. The UE maintains a database that captures the frequency and timing offset of each found candidate beam. When the UE receives a signal different from MRS on a known beam, the database can also be updated.
RSRP or RSRQ measurements over the entire frequency band or sub-band are useful. In particular, these measurements of n best candidate beams are reported, where n is the free flow rate. In addition, the UE indicates which beams arrive at similar arrival angles (azimuth and elevation). These beams are called QCL (quasi-localization). GNodeB can switch between QCL beams without notifying UE. On the other hand, QCL beams are not suitable for spatial MIMO transmission, because UE cannot analyze between beams. In addition to the n candidate beams, the UE should report which of these beams are QCL, which will support the gNodeB to select the beam associated with the CSI-RS antenna port. Using the same principle, GNOBE can also exclude QCL beams seen from GNOBE side (the deviation angles of azimuth and elevation are similar).
If the relative frequency or timing offset of these beams exceeds a predetermined threshold, MRS can be used to exclude these candidate beam sets from MIMO transmission.
Based on the identified set, if the influence of phase noise and CFO is moderate, the UE can report the related CQI of the selected transmission mode. This happens if the power of the phase noise /CFO is small relative to the power of the total noise and interference. GNodeB can use this information to reduce CSI-RS transmission and related CSI reporting.
According to P-2 and P-3, CSI-RS can be refined by Apollo beam. Therefore, multiple symbols can be used when the beam of the antenna port sweeps across a narrow coverage area. The UE should then be configured to report RSRP or RSRQ measurements. If the gNodeB antenna port keeps the same beam on multiple symbols, the UE should be informed so that it can improve the beam it receives.
It may be necessary to support beamforming across multiple gNodeB subarrays to achieve higher EIRP. This can be achieved if the antenna ports associated with these subarrays are beamformed in the same spatial direction and transmitted during the same CSI-RS symbols. Then, the UE can measure the phase difference between the received signals of each antenna port. One way to convey these phase differences may be to report PMI based on the Class A codebook if the subarrays are equally spaced and any phase mismatch between subarrays is adjusted by the calibration mechanism. Otherwise, other reporting mechanisms need to be supported to report this stage in a quantitative way.
For all CSI-RS transmissions, it is important for the UE to use the correct subarrays and beams for correct reception. To achieve this, the gNodeB informs the UE using the index in the previously reported candidate beam list. By identifying candidate beam sets, the UE will know exactly which of its subarrays and antenna weights are suitable for receiving upcoming CSI-RS symbols.
CSI-RS mechanism should be able to evaluate different Trps of MIMO transmission. In this case, beams of different antenna ports with different Trps should be provided. Because different Trps may have different frequency and timing offsets, it is suggested that antenna ports of different Trps be placed in different symbols. If the UE knows the candidate beam for each symbol, it can query its database to know the frequency and timing offset to be considered for each CSIRS symbol. The antenna port of each TRP may be configured as a CSI-RS resource, and the UE may be configured to report a resource indicator of the best performance resource (TRP). In addition, it must report the CQI and PMI of the best performing TRP.
In the case of signal strength loss due to sudden interruption, such as main path congestion or beam synchronization loss due to packet loss, beam recovery mechanism is very important.
The UE may still have a suitable beam to the gNodeB, or it may find a beam from the Mrs. In either case, gNodeB may not know the beam.
If the UE loses its uplink timing, it can re-establish the link to the gNodeB using the RACH signal in the appropriate beam direction. Alternatively, if the UE still has uplink timing, it can transmit in a channel similar to RACH, during which the gNodeB also scans its receive beam. In this channel, each UE is allocated a resource, so that gNodeB knows which UE has transmitted. After correct reception, the gNodeB knows which of its reception beams is suitable for connecting to the UE. This channel can also be used to receive scheduling requests, so it is called SR channel. The advantage of this channel is that it can provide more collision resources than the RACH channel because it utilizes the uplink synchronization of the UE.
After GNOBE and UE thus establish a beam pair for the working link, GNOBE can accelerate the beam discovery of UE by providing CSI-RS bursts containing beams throughout the sector.
For communication above 6GHz, hybrid beamforming is considered to overcome the high path loss between transmitter and receiver. In such a system, both Node gNB and UE can use multiple beams for control/data communication. The same or different beams can be used for control channel and corresponding data channel transmission.
User movement, angular rotation and blocking will lead to signal attenuation of data or control beam. Beam management monitors and switches beams to ensure reliable transmission and reception between gNB and UE. However, in some cases, if enough windows are not provided to switch beams, the signal quality may drop rapidly, resulting in beam misalignment. In this case, the control channel performance on UL or DL may be affected, which may eventually lead to wireless link failure and connection re-establishment. These processes cause additional delays that affect data throughput.
In order to overcome these problems, when beam misalignment is detected at UE, the beam recovery process is considered. The detection of beam misalignment is very troublesome and needs further study. Through beam recovery, gNB and UE can use alternative beams to reconstruct data and control channels. There are two cases to be considered in beam recovery: UL synchronization and UL asynchrony.
UL synchronization
When gNB and UE synchronize through UL, SR resources are used to perform beam recovery. GNB monitors the scheduling request, and after receiving the beam recovery message, GNB and UE re-establish the data and control channels.
Beam restoration using scheduling request has the following advantages: (1)SR area may contain more resources (cyclic shift); Compared with the contention-based RACH method, the beam recovery speed using SR is faster.
UL out of sync
When gNB and UE are out of synchronization, UE sends random access preamble for contention-based RACH process. After the RACH process is successful, gNB and UE re-establish data and control channels.