It does surprise that most of those investing in technology shares often only have the sketchiest grasp of what a company does. I remember writing about WANdisco (WAND; 955p) a year ago and at the time I felt nobody really understood what the business was about. There was a dearth of analyst coverage and the few journalists who attempted to cover it only had the vaguest notion of its products. Who can blame them? Behind WANdisco’s products, scientists and researchers were proving what it did mathematically and through simulations but it was clearly beyond the ken of many. That’s why I have this month decided to look at Filtronic, a supplier of cellular equipment for mobile phone base stations. Not many presently understand “Massive MIMO” and “beamforming”, two of the company’s technologies, which are going to be at the heart of 5G, the next generation of cellular connectivity (well perhaps that is apart from a certain “Mrs Dixon,” a high net worth individual who has accumulated 14% of the shares - Editor).
Relationship with Nokia going back to the 1990s
Founded in 1989 by professor David Rhodes, an academic at Leeds University turned entrepreneur, Filtronic’s business has historically comprised two divisions: Wireless and Broadband.
Current chief executive, Rob Smith, who I have met several times, argues that the single greatest opportunity facing the group is its Wireless business and the introduction of 5G cellular technology. At the heart of the opportunity is Filtronic’s relationship with Nokia stretching back to the 1990s and the two have collaborated on the development of a Massive MIMO (mMIMO) antenna. Smith references the “announcement on 30th July by T-Mobile and Nokia stating that Nokia will supply T-Mobile with US$3.5 billion of early stage 5G Network equipment as part of T-Mobile's first US nationwide 5G rollout” as an example of the trends in the industry as network operators rush to upgrade base stations. Although stopping short of confirming that this deal is part of an initial US$2m mMIMO contract that Filtronic announced last month, he does go on to say that “if you were to assume Massive MIMO had to be applied to T-Mobile’s 110,000 masts, in theory it would be several $100 millions of business for Filtronic.” Not all will be upgraded; some are rural sites; OEM’s often buy from more than one vendor; and Filtronic is not the sole source of mMIMO to Nokia but a convoy of potentially mind boggling orders looks set to emerge.
Filtronic Wireless
Filtronic Wireless supplies custom designed but high volume sub-assemblies including antennas for use in mobile phone base stations, to transmit and receive digital radio frequency (RF) signals at cell sites. This includes microwave filters used for combining and separating signals as well as a range of signal amplifiers and combiners that improve performance.
The days of 2G and 3G data and network are long gone and the world now thrives on the speed and luxury of 4G networks. With each generation of cellular technology Filtronic experiences a nasty pincer effect to its sales as operators stop ordering whilst they await the next generation technology.
When 4G was first mooted, marketers were quick to use the term “4G LTE,” a term deployed with early 4G networks that presented a substantial improvement on 3G, but did not fully qualify as 4G. Right now, we have something similar happening. 5G mobile network specifications are still in the process of being solidified but nobody in the mobile industry is presently waiting for 2020 to start building their networks because the 4G cellular bandwidth is already overcrowded as there are now millions of devices (smartphones, tablets, smart home devices) operating on it and the use of data hungry video apps continues to grow.
Cisco estimates that by 2020 - when 5G is expected to roll out - there will be 5.5 billion mobile users around the world, each consuming 20GB of data per month. The expectation is that 5G will see data rates rise from 100 Megabits per second to an anticipated 1 Gigabit per second.
What is happening, therefore, is that the network operators are beginning to look for technologies to upgrade their existing towers so they can improve network performance in the time gap between 4G and 5G.
Focus on densification of existing networks
Existing 4G networks therefore remain the dominant technology and the network operators have begun making a big investment to increase capacity by “densification” of existing networks - i.e. to multiply the capacity of a wireless connection. Such dense networks, primarily at frequencies less than 6GHz, are being marketed as 4.5G, 4.9G and 5G evolution by MNOs and is where Filtronic expects to see the majority of its demand over the next few years.
There are two ways to achieve this densification, says Smith. One technique is to use more spectrum in a cell - this can be done by using ultra-wide band (UWB) antennas. Another approach is to use existing band width more effectively using more spectrally efficient technologies such as mMIMO.
Old timers will recall that Filtronic had supplied UWB integrated antennas to Nokia in the past. However, product life cycles proved short and although it won some huge orders in 2016/7, Filtronic’s UWB program ended abruptly last year due to a change of priorities for the end customer and the market has moved on since that product was launched. But all is not lost because as Smith says, a significant amount of the IP developed for the UWB antenna has been utilised in the new mMIMO product. As Smith quips, “the King is dead, long live the King.”
Massive MIMO technology at the heart of 5G
The advantage of mMIMO is that a mobile operator doesn’t need to look for new sites and towers. With mMIMO, they’re upgrading existing sites, basically setting them up to broadcast 4G LTE and 5G at the same time in future. When 5G is ready they will be “software-upgradable” to 5G without additional tower climbs and able to operate in split modes to support 4G and 5G devices simultaneously, boosting the capacity of existing 4G LTE but also forming the foundation for 5G.
OK, so what is MIMO?
The acronym MIMO stands for Multiple-Input Multiple-Output. It is a term that is used to describe the ability of a wireless network to achieve more capacity without more spectrum by transmitting and receiving more than one data signal at the same time using the same data radio channel. Standard MIMO networks tend to use two or four antennas that are physically separate but used in combination to form part of the network. Specifically, MIMO organizes the network traffic flowing between two points into individual streams, transmits the streams in parallel and enables the receiving device to re-assemble it back into single messages.
mMIMO takes this concept to a whole new level by equipping the transmitter/receivers with dozens of antennas on a single array; the more possible signal paths, the better the performance in terms of data rate and link reliability. In this instance it is desirable for the antennas to be housed in a single unit and such a dense array of elements is referred to as a massive MIMO antenna. From a performance perspective the higher the number of antenna elements, the better the network performance.
Filtronic/Nokia have in the past referred to its Massive MIMO being a 32-antenna array in a 16-transmit/16-receive configuration and Filtronic and Nokia believe their mMIMO can provide an eight-fold increase in peak cell throughput and up to a five-fold improvement in average cell capacity at 2.6GHz bands with LTE. There’s no set figure for what constitutes a massive MIMO array however and systems could have tens or even hundreds of antenna elements, says Smith. Because of practical considerations, for example space on masts, wind loads and weight, mMIMO antennas are usually constrained to use at higher frequencies, which have shorter wavelengths and therefore produce physically smaller antennas. Right now, 4G LTE can only operate up to 6GHz, whereas the radio bands that 5G will be able to handle will be anywhere between 30GHz and 300GHz.
Upgrading existing masts
There are numerous factors involved but some writers estimate that mMIMO could potentially yield as much as a 50-fold increase in capacity in the future. In a world where spectrum licences are in the hundreds of millions of dollars, if not billions, this kind of increase in capacity is highly attractive to mobile operators and justifies the infrastructure investment necessary to deliver mMIMO.
In last year’s financial results, mMIMO didn’t exist at Filtronic. Filtronic Wireless division’s revenue was £18.4m down from £30.5m whilst operating profit fell from £3.5m to £2.4m as UWB sales tailed off although Filtronic made up some of this by selling higher margin filter and combiner products to public safety market customers (mostly Tetra), a new market for it. Filtronic also took the opportunity of a quiet period to consolidate two sites in the US into one and is now lean and mean and even despite small sales is generating free cashflow once more - so much so that last year cash went up by £1m to £3.6m. For the future, manufacturing for mMIMO is with its sub contract partner, which will allow it to ramp up production quickly.
So what about beamforming?
mMIMO therefore looks very promising for the future of 5G. As well as freeing up tons of bandwidth there is another advantage. We may all have experienced signal interference when we drop connections usually when local masts become oversubscribed. Installing more antennas to handle cellular traffic also causes more interference if those signals cross, which might happen in, for instance, a densely populated road.
As Smith explains, as the number of antennas and the corresponding number of data streams in a mMIMO system increases, it requires highly advanced signal processing and system design for the network to function well. Until very recently, the computing power just wasn’t available to do this but by arranging elements in a grid pattern, it is possible to get round this problem using a technique known as 3D beamforming.
Unlike 4G technology, which shoots radio signals in all different directions, beamforming enables a radio signal to be focussed at a single point rather than across the full sector covered by an antenna. To achieve 3D beamforming with a mMIMO antenna it is necessary to integrate the OEM’s radio and antenna technology together. A really close working relationship between the radio and antenna manufacturer is therefore essential.
Chinese OEMs barred in US
mMIMO is already live commercially in China and Japan within a 4G LTE context. Chinese vendors, like Huawei, are not allowed to supply into US telecommunications networks on security grounds so the US network goliaths, T-Mobile/Sprint (the two are in the process of merging), AT&T and Verizon, who are all conducting their own ambitious mMIMO tests, therefore have to buy from Western companies.
Smith says T-Mobile has historically bought from Ericsson and Nokia as have Sprint who I understand also buys small amounts from Samsung. Verizon and AT&T buy from Nokia as well as Ericsson. Beyond that there is no real competition.
Filtronic Broadband
The smaller business that exists is Filtronic Broadband, which provides backhauling products. If you imagine a network comprised of fibre optic cables, there may be areas on the periphery to which the network physically does not extend but where a network operator might want coverage. Rather than dig trenches and lay more fibre, the MNO can use a microwave point-to-point link to backhaul data to the fibre. Filtronic’s modules, branded as Orpheus, do this and are basically high frequency radio transmitters. That has been the historical mainstay for the division but the business has also developed transmit receive modules (TRM) going into phased array radars in the defence sector. These are computer-controlled arrays of antennas used by the military in aircraft, which creates a beam of radio waves that can be electronically steered to point in different directions, without moving the antennas.
Last year, the Broadband division grew sales to £5.6m (2017: £4.9m) and moved back into the black with an operating profit of £0.2m (£0.9m operating loss) after it won two multi-year (3-8 year) defence orders for TRMs. H1 had seen a few component supply issues but both contracts were in full production in Q4 18. The superb news is that the aircraft that these TRMs go into have seen more orders being placed and so more contracts will flow down the supply chain in due course.
With both sides of the business firing up strongly, I am a buyer.
