As one of the three major application scenarios of 5G, TN929.5 ultra-reliable and low-latency communication (URLLC) has received more and more attention. The Internet of Vehicles is one of the main applications of URLLC, which has a high demand for the reliability of information transmission. In order to enhance the robustness and reliability of transmission, a variety of Multi-TRP based URLLC technical schemes are introduced from the perspectives of space division, frequency division and time division, and their advantages and disadvantages are compared and analyzed. In addition, a new signaling indication scheme for dynamic switching transmission mode is designed, and the performance simulation evaluation of various schemes is carried out. Simulation results show that the dynamic switching approach can provide significant performance gains.

The 5G NR will support three major application scenarios, namely enhanced mobile broadband (eMBB), massive machine type communication (mMTC) and ultra-reliable low-latency communication (URLLC). For URLLC, it is characterized by high reliability and low delay, the reliability can be as high as 99.999%, and the delay can be as low as 1 ms. The main applications of URLLC include: industrial application and control, traffic safety and control, remote manufacturing, remote training, remote surgery, etc.
For the transmission technology of Multi-TRP, preliminary progress was made in the early stage of NR Rel-15 discussion in 3GPP, but due to the limited discussion time of Rel-15 version, this topic was shelved. The Internet of Vehicles is one of the main applications of URLLC, which has a high demand for the reliability of information transmission. In order to enhance the reliability and robustness of transmission, NR Rel-16 restarted the discussion of Multi-TRP transmission technology [1]. Based on this, this paper introduces a variety of Multi-TRP based URLLC technical solutions from the perspectives of space division, frequency division and time division, and compares and analyzes their advantages and disadvantages. In addition, a signaling indication scheme for dynamic switching of transmission modes is designed, and performance simulation evaluation of various schemes is carried out.
1 Transmission scheme of multiple sending and receiving nodes in URLLC scenario
1.1 Multi-TRP based URLLC scheme For Multi-TRP transmission, the same transport block (TB) can be sent from two different TRPs, as shown in Figure 1. To support soft combining at the receiving end, different redundancy versions can be used for these repeated transport blocks. In addition, in order to further enhance the reliability of transmission, these repeated transport blocks may be scheduled by repeated physical downlink control channels.
For the repeated transmission mode of transport blocks, in the RAN1 96th meeting of 3GPP, it was determined to use space division (SDM), frequency division (FDM), intra-slot time division (TDM) and inter-slot time division (TDM) as Multi- Candidate technology of TRP based URLLC transmission scheme[2]
(1) Scheme 1 (SDM)
Multiple Transmission Configuration Indication (TCI) states are configured in separate time slots, overlapping in time-frequency resources.
①Option 1a
Each TRP transmits a layer or group of layers of the same transport block associated with a TCI and a group of demodulation reference signal ports. Furthermore, the spatial multiplexing layers of all TRPs come from the same codeword, which uses 1 redundancy version (RV).
②Option 1b
Each TRP transmits a layer or group of layers of the same transport block associated with a TCI and a group of demodulation reference signal ports. In addition, one layer or a group of layers of each TRP corresponds to an independent codeword, and different codewords can use the same RV or different RVs
(2) Scheme 2 (FDM)
Multiple TCI states are configured in separate time slots without overlapping in frequency domain resource allocation. Each non-overlapping frequency domain resource allocation is associated with a TCI state.
①Option 2a
One codeword uses one RV, and this codeword is mapped to all frequency domain resource allocation positions.
②Option 2b
One codeword uses one RV, different codewords are mapped to each non-overlapping frequency domain resource allocation position, and the RVs between different codewords may be the same or different.
(3) Scheme 3 (Intra-slot TDM)
Multiple TCI states are configured in separate time slots, which do not overlap in time domain resource allocation. The granularity of time domain is mini-slot, and the TB sent by each TRP corresponds to one TCI and one RV.
(4) Scheme 4 (Inter-slot TDM)
Multiple TCI states are configured between multiple different time slots, and the TB sent by each TRP corresponds to one TCI and one RV.
For the above four Multi-TRP based URLLC schemes, the advantages and disadvantages are shown in Table 1.
1.2 Scheme design of dynamic switching In order to give full play to the advantages of various schemes, the above-mentioned SDM and FDM methods can be combined with TDM methods to support FDM+TDM and SDM+TDM methods. In order to obtain greater flexibility, when the number of repetitions in the time domain is greater than 1, the three modes of TDM, FDM+TDM and SDM+TDM can be dynamically switched; when the number of repetitions in the time domain is equal to 1, FDM and SDM can be switched dynamically. These two methods are dynamically switched.
In addition, when the number of time domain repetitions is 1, it can be Multi-TRP transmission for URLLC, such as SDM/FDM, or it can be traditional Multi-TRP transmission for eMBB. Compared with the repetition scheme of URLLC, eMBB-oriented Multi-TRP transmission can be called a non-repetition scheme. The non-repetition scheme can improve the transmission efficiency and increase the throughput, while the repetition scheme can enhance the reliability and robustness of the transmission. For greater flexibility, the repetition scheme and the non-repetition scheme can be dynamically switched based on channel conditions.
In order to realize the dynamic switching of the above repetition scheme and non-repetition scheme, as well as the dynamic switching of TDM, FDM+TDM and SDM+TDM in the repetition scheme, a new DMRS signaling indication table is designed in this paper, as shown in Table 2 shown. In Table 2, the indices under a codeword are 0 to 7 corresponding to the non-repetition scheme, and the remaining indices are divided into 3 groups, corresponding to the 3 schemes of TDM, FDM+TDM and SDM+TDM in the repetition scheme respectively. The corresponding indexes under the two codewords in Table 2 can also use a similar method to dynamically switch between various schemes. When the number of TCI states indicated by the TCI field in the DCI is 1, the DMRS in Rel-15 NR is used to indicate the table (that is, the old table); and when the number of TCI states indicated by the TCI field in the DCI is greater than 1, this information The advantage of the command indication scheme is that it can maintain the same signaling overhead as Rel-15 NR, and is backward compatible with users of Rel-15 NR, realizing non-repetition scheme and repetition scheme (including TDM, FDM+TDM and SDM+TDM) dynamic switching function.
2 Performance evaluation and analysis In order to verify the performance of Multi-TRP based URLLC, this paper first uses link-level simulation to evaluate the performance of repetition and non-repetition schemes relative to Single-TRP:
(1) Repetition: SDM repetition based on coherent joint transmission (NCJT)
Each TRP transmits data of layer 1 independently. The MCS and resource allocation configuration used by the two TRPs are the same. The two TRPs use the DMRS port 0 and port 2 respectively.
(2) Non-repetition: eMBB-oriented Multi-TRP transmission Each TRP transmits data of one layer separately, and the configuration of resource allocation is the same. These two TRPs use DMRS port 0 and port 2 respectively, and the data parts of the two layers are repeated.
Link-level performance comparison between a single TRP, two TRPs without repetition, and two TRPs with repetition. The main parameter configuration of the simulation is shown in Table 3, and the modulation mode of 64QAM is adopted. It can be seen from the simulation results in Figure 3 that the Block Error Rate (BLER) performance of the repetition scheme is better than the other two schemes, because the receiving end can perform soft combining. For spectral efficiency, the scheme with two TRPs performing repetition can achieve the highest spectral efficiency in the low SNR region compared with the other two schemes, while in the high SNR region, the scheme with two TRPs without repetition has the best performance. This is because the transmission performance in the high SNR area mainly depends on the multiplexing order, while in the low SNR area, the reliability of transmission is more pursued.
Table 4 shows the system-level simulation performance of single cell, NCJT with repetition and NCJT without repetition in indoor scenarios. From the two indicators of 5% throughput and interruption probability in Table 4, it can be seen that compared with the performance of a single cell, the two schemes of repetition and no repetition between two TRPs can bring great performance gains . In system-level simulations, the performance of dynamic switching of non-repetition and repetition schemes is evaluated simultaneously, and it can be seen that dynamic switching can provide significant performance gains.
This paper introduces a variety of Multi-TRP based URLLC technical solutions from the perspectives of space division, frequency division and time division, and compares and analyzes their advantages and disadvantages. In addition, a signaling indication scheme for dynamic switching of transmission modes is designed, and performance simulation evaluation of various schemes is carried out. Simulation results show that the dynamic switching approach can provide significant performance gains