“5G” technology is all about big ideas: incredibly high speeds, low latency and complex multitechnology networks expected to support new, demanding use cases and billions of new devices.
One of the fundamental challenges for 5G will be fitting those big ideas into very small spaces – including the radio frequency front end within chipsets.
While 5G is essentially a marketing term for the wireless industry right now, the vision that is being laid out has some major divergence from 4G, according to Mike Eddy, VP of marketing for Resonant, which designs and produces filters for the RF front end.
“Rather than focus more on consumer mobile broadband, [5G] is much more about many, many different use cases … that put a huge amount of demand on the network,” Eddy said.
In a recent white paper (pdf) on 5G and the FR front end, Resonant concluded some of those drivers for 5G are conflicting: that increased data rates and capacity and ultra-high reliability in 5G also vie with expectations for low device cost and an explosive increase in the number and types of connected devices. A 5G target latency of less than 1 millisecond is “an exceptionally aggressive target for a cloud service with communication between the access network, core network servers and the wireless device,” Resonant went on to say, adding “current average latencies in the U.S. are 70 to 85 milliseconds.”
Resonant sees core technologies for 5G enablement as new frequencies, particularly those above 6 GHz; massive multiple-input/multiple-output antenna technology; network densification and interference mitigation; and new, adaptable waveforms.
Those new approaches mean new challenges for the RF front end, however. Increasing use of carrier aggregation (in 4G as well as in 5G) is driving more complex designs. In a 5G RF front end, Eddy said, “you have a huge number of RF paths, so that you can aggregate different pieces of spectrum but also so that you have a [device] that can operate anywhere in the world.”
In particular, he added, “the piece of the front end that gets complicated is the filter.”
More MIMO demands and spectrum proliferation mean more filters will be needed, but that will impact power efficiency of the device. Eddy said a state-of-the-art smartphone currently has about 50 filters, but may increase to as many as 200 filters with more antennas and more RF paths – so filters need to be smaller in order to meet design size requirements. While the industry is already shrinking filters compared to their size even two years ago, Eddy said he expects them to become smaller still to meet 5G requirements. However, the high performance needs for 5G will make it difficult for all of that to be done at a low cost, he noted.
Resonant also expects the use of millimeter wave frequencies will require different filter technology than the acoustic wave filters that have been used in mobile devices at cellular frequencies. The company also predicted that due to the need for optimization of the RF chain and careful assessment of interactions between elements, there will be more integrated approaches for filtering – and that the overall jump in complexity for 5G subsystems may mean only a limited number of manufacturers can actually produce those products.
“With 5G coming along, the number of design requirements is huge, and more and more complex,” Eddy said.
Image copyright: mycteria / 123RF Stock Photo
