One way to comprehend the 5G ecosystem is to break it up into segments based on frequency. Figure 1 is a diagram of this. There are two main areas that will capture the attention of network operators, device manufacturers and system solution providers such as Skyworks. First, there will be significant activity in the sub-6 GHz frequency domain. Both 4G LTE as it is currently practiced, and 5G NR will be deployed concurrently in spectrum below 6 GHz. Second, the devices, techniques, and general radio protocols will be very familiar to those addressing this segment of the market. In other words, below 6 GHz we expect the same look and feel with respect to the RF content with allocations for additional content to meet new bands and 5G features.
So far, 3GPP has decided to tether 5G NR to the existing 4G LTE environment. As such, we do not expect standalone 5G NR devices to exist in the market for quite some time. The implication here is that the framework for mobile devices remains on a steady track, albeit with some increased functionality, for the new 5G.
Another implication not readily discernible is that the backhaul network will need to increase capacity to cover all the front haul increase in data.
Most of the recent 5G headlines have focused on mmWave spectrum and the introduction of new techniques like beam forming to overcome the higher path losses at these higher frequencies. As with any new technology in its infancy, there are many unknowns regarding the business and usage case, which is compounded by new commercial technology deployments.
Despite the unknowns, we see 5G NR usage cases centered around small cell densification in fixed wireless application as the most likely outcome. This aligns with the views proposed by many of the advanced LTE mobile network operators, with Verizon Wireless being a very strong advocate. Due to the differences in mmWave radio transmission and the sub-6 GHz ecosystem, Skyworks believes the business case for high-bandwidth mmWave fixed wireless communications is the correct focus for initial deployment of 5G NR.
Currently, there are many obstacles to overcome that would prohibit mmWave deployment in consumer devices as a first step in the technology roadmap. These include battery life, beam tracking and management, and radio propagation challenges, to name a few.
Another way to think about the differences between the sub-6 GHz and the mmWave ecosystems is to think about the area coverage. Sub-6 GHz frequencies are used in the macro and small cell densification networks to provide users with data rates approaching and exceeding multiple Gbps covering a large geographical area. Millimeter wave deployments are targeting several tens of gigabits per second data rates in close proximity to small cell base stations with very narrowly focused beams, hence the fixed wireless application.
Mobility applications from mmWave devices will come much later, after the initial deployments have proven successful. Technology barriers for beam tracking and SNR still need to be overcome to achieve a feasible resolution.
Spectrum in 5G
In addition to the key LTE features, another important aspect to consider over the next several years is the availability of spectrum and new bands, which will be utilized to deploy some of these new techniques. A quick survey of the proposed bands indicates a preponderance of new TDD spectrum becoming available globally in the 3 to 6 GHz range (LTE sub-6 GHz) for both LTE Advanced Pro and 5G NR phase 1. For 5G NR phase 2, the intention is to utilize mmWave frequencies with much larger bandwidth for new 5G applications.
The chart shown in Figure 2 shows new frequencies being discussed in relation to 5G NR. There are two frequency regions of interest playing an important role in 5G NR. In the 3 to 6 GHz band, there is generally clear spectrum globally in the 3.3 to 3.8, 3.8 to 4.2, and 4.4 to 4.9 GHz region. These bands, which are all based on TDD or unpaired spectrum, generally have wider bandwidth than their 4G predecessors. They will be particularly important in user equipment plans that will use 4G LTE anchors and new 5G NR radio transmissions. In addition to the licensed bands in the 3 to 6 GHz range, it is very likely there will be additional use of supplemental unlicensed bands to squeeze out even more usable bandwidth.
The other aspect of the 5G NR, which is much more revolutionary, will utilize mmWave spectrum. The mmWave bands have the widest achievable bandwidths and are available with some regions offering multiple GHz spectrum. Industry consensus is building around the usage model of mmWave spectrum for fixed wireless applications. Applying mmWave technology to mobile devices will represent a very high technological challenge for the near future.
5G will bring very high capacity and low delay (ultra-low latency) using both sub-6 GHz and mmWave spectrum. In conjunction, we will need to determine how much capacity is needed for backhaul. Network Function Virtualization (NFV) is likely to emerge on last mile links and will leverage more intelligent switching protocols at the network edge.
A key component of the 5G landscape is going to be outdoor small cells. We referred to this earlier as small cell densification of the 5G network. These small cells are essential, as mmWave will leverage directional beams in a short range (see Figure 3). A secondary effect of densifying the network with high data rate links is the need to improve backhaul characteristics.
Today, fiber is the backhaul option of preference. However, as networks grow to become more dense with the deployment of small cells, licensed fixed or point-to-multipoint (PMP) mmWave may emerge as the most flexible solution. For example, an operator (leveraging guaranteed QoS with licensed PMP) with a >10 Gbps hub site could aggregate backhaul traffic from multiple base stations. The economics of that scenario improve as more base stations are added progressively with densification.
Millimeter wave communications technologies in the 60 GHz and 70-80 GHz range for high capacity at the last mile and pre-aggregation backhaul were explored in ‘Advanced Wireless and Optical Technologies for Small Cell Mobile Backhaul with Dynamic Software-Defined Management’ (Bojic, et al., 2013).
New Technologies that are Required to Serve 5G
Figure 4 represents existing technology and spectrum, as well as the planned 5G NR spectrum. We mapped those applications to the technologies needed to implement both power and low noise amplifiers, RF switching, filtering, and antenna integration functionality.
One key takeaway is that the entire new spectrum for 5G NR, whether sub-6 GHz or greater than 6 GHz, are all TDD bands. Due to the fact that they are time division duplex, frequency duplexers are not required to implement a front-end solution. Filtering, as needed, is accomplished in bandpass filters. In the 3 to 6 GHz region, filtering can be accomplished in acoustic, IPD or ceramic technologies. In the sub-6 GHz region, most of the 5G NR activity will be deployed in the 3 to 6 GHz frequency spectrum. In all cases, 5G NR will require a 4G LTE anchor, typically in the below 3 GHz region. In other words, 5G will exist as an overlay to the existing 4G network. Implications are that user equipment will use very similar techniques in both the 4G and 5G NR sub-6 GHz domain. This leads to increased complexity, however the techniques and technology remain essentially the same as our current 4G devices.
For the 5G mmWave fixed wireless applications, requirements for massive MIMO and multiple beamforming mean that transmit and receive functions will most likely be in distributed array formats. As a result, there will be multiple PA streams and multiple receive chain streams to accomplish transmit and receive functionality to fixed wireless devices.
Filter technology for mmWave 5G is likely to be based on transmission line and waveguide cavity technology. Microelectromechanical systems (MEMS) cavity resonators are also an attractive choice to avoid handtuning filters and leveraging silicon wafer-based manufacturing approaches. Multi-pole filters suitable for operation at frequencies of 20 to 100 GHz have been demonstrated in ‘RF-MEMS Based Passive Components and Integration Concepts for Adaptive Millimetre Wave Front-Ends’ (Gautier, 2010).
User equipment PA technology for 5G sub-6 GHz will leverage traditional GaAs heterojunction bipolar transistor (HBT) technology, but it will require some improvement or innovation in dealing with wide bandwidth signals, especially in biasing. Skyworks will leverage its current expertise in Wi-Fi 802.11ac and 802.11ax (up to 160 MHz bandwidth with extremely tough error vector magnitude requirements). Above 6 GHz, we will leverage our experience in System-in-Package (SIP) technology to create antenna front-ends on organic laminates.
To download full technical paper, please visit www.skyworksinc.com/5G.