Today, we talk about the 5G NR MIMO Transmission ways. What are the required transmission schemes for NR MIMO?
There are three implementation options for the 5G NR beam assignment, which are analog, digital, and hybrid beam assignment. For these implementations, both multibeam-based and single-beam-based approaches should be considered. But what are the required transmission schemes for the 5G NR MIMO?
Closed-loop MIMO of 5G NR MIMO
5G NR MIMO Digital beamforming
In this system, one RF beam will cover many propagation paths, so baseband beamforming gives the transceiver a high degree of flexibility to transmit one or more data streams based on CSI feedback. To facilitate understanding, a few examples are given below.
Case a: UE feedback from a pre-coder selected from the GoB codebook. The base station can adjust the baseband precoder to align with the optimal direction through beam focussing, see Figure 1(a), where the light and dark colors indicate the RF beam generated by analog beam focussing only and the final beam generated by hybrid beam focusing, respectively.
Case b: If multiple path directions are available, the base station can strive to transmit the same layer information by pointing multiple beams at independent paths, see Figure 1(b). If the beams can be weighted and coherently combined into a single stream, the receiver performance will be further improved, as in the case of using feature vectors for precoding.
Case c: If the rank is sufficient, the streams of these different layers can be transmitted simultaneously through multiple paths, i.e., spatial multiplexing, see Figure 1(c).
5G NR MIMO Hybrid beam assignment
In this architecture, the base station can use multiple TXRUs to transmit the same/different RF beams, optimizing both the analog beam and the baseband precoder. Similarly, the following cases are available.
Case d: the base station can use multiple TXRUs to transmit the same RF beams, and subsequently digital precoding can be associated with these TXRUs and further refine these RF beams, see Figure 2(d)
Case e: The base station can use different groups of TXRUs to transmit different RF beams, and the TXRUs within the same group can refine the final beam with a narrower beam width. The data stream can be delivered to the UE via several paths. The precise combination of multiple beams right can lead to a greater beam fugacity gain, see Figure 2(e).
Case f: similar to case e, but different beams can be used for multiple layers to achieve spatial multiplexing, see Fig. 2(e).
According to the previous analysis, different TXRU/TXRU groups can be regarded as virtual TRP, and the hybrid beam assignment can dynamically adjust the RF beam to achieve coordination among virtual TPs. Compared with the CoMP of LTE, vTP has smaller coverage and greater flexibility.
In LTE, single-layer/multi-layer/TP/UE transmissions use a unified framework. Due to the flexibility of DMRS, these transmissions have good transparency. In 5G NR, multi-beam transmissions can also be seen as cooperative transmissions between multiple virtual TPs, which shows that single/multi-layer/TP/UE transmissions can be implemented in a unified framework.
5G NR MIMO receiver beam assignment
Considering the cost and complexity, when there are more antenna units and the number of RXUs is usually smaller than the number of originals, the UE should first use the RF beam to select the receive beam pointing to one or more paths.
Case A: multiple RXUs shares the same RF beam and receive signals from the LOS path, with the baseband and receive weights of multiple RXUs further refining the beam, as shown in Figure 3(A).
Case B: Multiple RXUs/RXU groups correspond to multiple RF beams pointing in different directions. This approach allows more signals to be captured from different paths.
Case C: Same as Case B, but different beams correspond to different 5G NR MIMO layers.
When transmit and receive 5G NR MIMO beam assignments are implemented in a transparent way, some cooperation issues arise between receiving and transmitting. More specifically, the propagation path carrying the transmitted signal will not be received; or, there is no strong power in the path covered by the received beam. Some performance losses can also occur. Two possible solutions are proposed here.
1) Configuring or pre-specifying the beam range of selective transmit or receive schemes.
2) Employing a blind-detected multi-shot DMRS.
The former affects transparency, but the latter leads to more reference signal overhead.
Open-loop MIMO of 5G NR MIMO
In LTE, several spatial diversity schemes are supported, such as SFBC and FSTD, and for NR-MIMO, spatial diversity schemes are also required for the control and data channels. These spatial diversity schemes can be used for both single-beam and multi-beam approaches.
SF(T)BC:
Signals transmitted by multibeam/TXRU/TP can be coded with Alamouti, which is promising for LTE downlink transmission. However, this scheme is opaque and requires more standards efforts. Moreover, it is not scalable and hardly flexible to support the different numbers of beams/TXRU/TP.
Switching:
During data transmission, multiple beams/TXRU/TRPs can be switched with each other by FSTD or TSTD. FSTD seeks to switch between different subcarrier groups, but TSTD has to achieve this in different time-domain symbols.
In general, both options are opaque. The base station shall indicate to the UE information about the switching. Transparency for the above beam assignment can be achieved if the granularity of FSTD and TSTD is RBG or TTI, respectively.
CDD:
In different beam/TXRU/TRP transmissions, a certain time delay can be added in the time domain, which is equivalent to multiplying different phases of each frequency in the frequency domain to obtain more diversity gain.
Although it is not as good as SFBC in terms of diversity performance, it is a transparent method with minimal complexity of standardization.
The performance gain of closed-loop spatial multiplexing relies on accurate CSI feedback. In some cases, accurate CSI is not available, such as when the UE is moving at high speed, or when the CSI obtained based on non-ideal channel reciprocity is not accurate enough.
Therefore, the base station can use semi-open-loop 5G NR MIMO beam assignment for data or control transmission. An example is shown in Figure 4(a), where the base station uses multiple narrow beams within an adjustable range for transmission.
These narrow beams may be agnostic to the UE and the base station may determine the distance and beamwidth based on long-term/broadband PMI from the UE or coarse CSI based on SRS measurements. These beams can be switched in the time or frequency domain.
Another example is shown in Figure 4(b), where switching between beams implies switching between paths if multiple beams correspond to multiple paths, thus facilitating the robustness of the control channel when suffering from path blocking at high frequencies.
The transmission scheme should be as UE-independent as possible so that at least a unified framework of semi-open-loop 5G NR MIMO beam assignment and closed-loop 5G NR MIMO beam assignment should be used as a starting point for transmission. A unified framework here implies a little impact on the DCI and DMRS design.
The unification of the architectures for closed-loop and semi-static open-loop 5G NR MIMO depends on the requirements for the switching granularity, so the performance of semi-static open-loop 5G NR MIMO with different beam switching granularities and the trade-off between performance and complexity need to be further investigated and evaluated.
If you have any 5G NR antenna questions, please read our ANTENNA FAQ section, if you still cannot get the answer you need, please contact us.
You may also be interested in the below articles.
Internal Antenna VS. External Antenna
LTE Vs. 5G: Is 5G Better Than LTE?
