5G mMIMO Βeamforming Overview

Beamforming is one of the most efficient and practical solution to either increase throughput for cell-edge users in sub6 GHz or overcome higher path loss and atmospheric attenuation in mmWave bands.

mMIMO technology and its beamforming efficiency is a tradeoff between antenna panel cost, antenna panel size and antenna panel’s RF module complexity. Due to mMIMO peculiarities there are two main technical solutions, analog beamforming and digital beamforming. However due to some exotic and useful characteristics it is beneficial to combine analog with digital beamforming to give rise to a new hybrid solution sharing all benefits for both sub6GHz and mmW bands.

A.?????Analog beamforming

Following the history of antenna beamforming, most likely in the military applications of phased-antenna arrays for radar beam-tracking, beamforming in early years started as an analog solution with phase shifters (analog tappers), one technology that was improved gradually in multiple stages. Some flexibility was later incorporated through a switching architecture with several phase shifters, each for a different beam pattern. Finally, mostly for civilian applications adjustable phase shifters at each antenna element were introduced to enable flexible and adaptive beams that could look into any direction, requesting high power (EiRP) transmissions.

Figure 1. Analog beamforming with phase shifters

?

A simplified analog beamforming diagram is illustrated in figure 1, where the functionality is presented in expense of the RF details. The weights are adaptively computed in digital domain and updated in analog domain using specific DSP algorithms for a target criterion.

?

B.?????Digital Beamforming

Digital beamforming on the other hand is more flexible for spatial multiplexing, emphasizing the user throughput mostly on the multiplexed number of layers rather than the beam gain. In this case, each antenna element has its own dedicated RF chain as well as individual DACs and ADCs. Recalling the sampling analogy, this implies that the gain and phase of each spatial sample is adjusted using a digital precoder in the signaling processor before RF up conversion at Tx or after down conversion at Rx in order to decide about the proper multiplexing and phase shifting. From signal processing point of view it is more powerful as a technique since it enables an algorithmic approach for each antenna output separately as an accessible sample figure 2.?

Figure 2. Digital beamforming with baseband digital Precoder

?

Digital beamforming, instead of using the beam forming gain to counter-effect pathloss and low SINR conditions it combines beamforming to certain directions, interference suppression by using nulls in the un-desired interference aggressor’s angle of arrival, along with Multiple spatial streams (Spatial Multiplexing?- SMUX) can be simultaneously created for spatial multiplexing.

One of the major drawbacks of Digital beamforming is the requirement of having almost equal RF chains as the number of antenna element groups (sub-arrays) in the antenna panel, a restriction which leads to very expensive firmware implementations especially when the number of antenna elements increases.

?

C.?????Hybrid Beamforming

Hybrid beamforming comprises both Analog and Digital beamforming benefits. Following the implementation there is a compromise between low power transmission and spatial multiplexing, where precoding or combining is done first in the analog part and later in the digital domain.

In such a solution the antenna elements are still driven by analog phase shifters, however the number of RF chains and the subsequent ADC/DAC are much less than the number of sub-arrays in the panel, thus making the antenna implementation feasible.

Precoding is then performed in the digital domain at the level of radio chains, figure 3. This sufficient rank reduction in the number of data converters and corresponding chains results in less cost, computational load and power consumption.?

Figure 3. Hybrid beamforming

References

1.???????https://wirelesspi.com/what-is-the-difference-between-analog-digital-and-hybrid-beamforming/

2.??????3GPP TR 36.871 “Downlink Multiple Input Multiple Output (MIMO) enhancement for LTE-Advanced”

3.??????R1-164708 : 3GPP TSG-RAN WG1 #85 - NR Massive MIMO Design Considerations

4.??????R1-165179 : 3GPP TSG-RAN WG1 #85 - Multi-Antenna Technology for NR Interface

5.??????R1-165362 : 3GPP TSG-RAN WG1 #85 - Multi-antenna Architectures and Implementation Issues in NR

?

For example, this flexibility is evident in the following scenarios.?

?

The BLER calculation is based on evaluating the CRC on each transport block, meaning that an erroneous block is a Transport Block, of which the cyclic redundancy check (CRC) is wrong , following next figure 1 (reference:?https://telecompedia.net/block-error-rate-in-lte/ ).