Increasing numbers of low earth orbit (LEO) satellites are being developed and deployed to address the challenge of providing ubiquitous broadband data coverage around the globe. These ‘mega constellations’ of new satellites will help to deliver the full potential of 5G and provide high-speed connectivity to remote parts of the world, including moving platforms such as aircraft and ships, as well as providing low latency connectivity for business users.

New opportunities and limitations

For the developers of satellite technologies and components, the proliferation of LEO satellite constellations presents new challenges, as well as exciting opportunities, not previously encountered with traditional Geostationary Equatorial Orbit (GEO) and Medium Earth Orbit (MEO) satellites.

Conventional geostationary satellites remain in their fixed orbits for 25 years or more. To achieve such lengthy operational lifetimes in space, all components must meet stringent requirements for reliability and radiation tolerance. The testing procedures are arduous and, where component reliability is in doubt, redundancy is achieved at the component level, by installing two or more identical sub-systems as a back-up.

Finding the right standard

For new LEO satellites used in mega constellations, redundancy is achieved at the satellite level, meaning operators launch replacement satellites to repair the constellation. These satellites have much shorter operational lives than their traditional geostationary counterparts, remaining operational for five to ten years. Nevertheless, there remains a need for the components installed on these satellites to offer high reliability, quality and performance in the high-radiation environment of space.

Using the stringent standards set for GEO and MEO satellite components would result in costly over-engineering for LEO satellite applications. So, to achieve the high quality and reliability levels required, LEO operators have tended to look to other demanding, high-performance applications – such as automotive – to set the standard for their satellite sub-systems. However, these general standards are not always fit for purpose, and there is a lack of suppliers with the necessary expertise at higher frequencies. These factors, combined with a lack of space heritage among suppliers, mean the LEO market requires a fresh approach if it is to meet growing demand for space-qualified RF components.

Meeting demand for more capacity

The volume of data being consumed worldwide is increasing apace, so there is an urgent need to increase satellite capabilities and network capacity. One example is through phased array antennas that make it possible to steer beams and target areas where extra capacity is needed, creating multiple beams at the same frequency. Care is required to avoid congestion and interference caused by overlapping beams, but in general it is possible to increase the capacity of satellites by reusing frequencies for different geographical areas.

Deploying digital processing capabilities within the satellite payload is another way to expand capacity. This enables data being brought to the satellite from users to be repackaged and consolidated on the satellite, creating extra capacity and allowing more efficient use of the frequency spectrum.

Exploring higher frequencies

However, as demand for data increases worldwide, the ultimate constraint on expansion will be RF capacity. Delivering the extra bandwidth needed will require expansion into higher frequencies, not currently widely used for earth-to-satellite communications. Ku and Ka bands offer around 2GHz of available bandwidth each, although there is some protected bandwidth within these bands. This creates additional obstacles when it comes to developing frequency plans and implementing hardware. What’s more, communications channels are becoming very congested at Ka and Ku bands, which are also shared with geostationary satellites. The orbits of any new LEO constellations therefore must be very carefully plotted to avoid interference with existing geostationary satellite transmissions.

Future satellite systems will move to Q and V bands, and indeed the International Telecommunication Union (ITU) has already approved these bands for use in forthcoming constellations. These higher frequency mmWave bands are currently little used for satellite communications and provide an important way to increase the capacity of the feeder links between satellites and the terrestrial network. Q and V bands each provide up to 5GHz of additional bandwidth, with a few excluded sub-bands. Looking further into the future, even higher frequencies will offer greater scope for expansion, with E band providing two wide-open 5GHz segments of contiguous bandwidth.

Challenges posed by mmWave

As frequency increases, so does atmospheric absorption, which, along with the difficulties in generating power in these bands, makes the links more sensitive to environmental conditions. However, the wider bandwidth available means that modulation levels can be reduced and output power increased to maintain the link, retaining data rates comparable with lower frequencies. The capacity gains thus make it an attractive option that’s worth investing in to secure long-term broadband connectivity via LEO satellites.

Currently, there are very few RF mmWave payload systems available with space heritage. That’s a limiting factor in the market, but also a huge opportunity for RF component manufacturers with the necessary expertise and track record in critical terrestrial applications. With new and well-resourced satellite operators joining the burgeoning market for LEO satellites, the demand for high-quality, space-compliant mmWave components is set to grow significantly.

UK expertise in mmWave components

High-reliability mmWave transceivers and SSPAs for communications and defence applications are already being designed and volume manufactured in the UK – and these are precisely the components needed for next-generation satellite products. Proven UK expertise in producing high reliability u-wave and mmWave modules for defence and communications systems can be applied directly to satellite applications. And because these devices have been rigorously tested and successfully deployed in terrestrial networks, satellite operators can be confident in their capabilities for non-terrestrial applications. The semiconductor processes already used in mmWave devices for terrestrial applications are inherently tolerant to radiation, so do not need further testing for space compatibility. The microprocessors and transistors used to provide power and control can be sensitive to radiation, but cost- effective radiation-tolerant alternatives are readily available.

Recently, Filtronic has designed high-capacity transceiver modules for use in high altitude pseudo-satellite (HAPS) systems, providing communications links of up to 40Gbps at E-Band. HAPS stations are unmanned aerial vehicles that provide moving 5G base stations, operating in the stratosphere at an altitude of around 20km to provide a ‘base station in the sky’. They play an important role in expanding connectivity around the globe and directly to LEO satellites at much higher altitudes where atmospheric absorption is very low. Filtronic RF technology provides solutions for both HAPS and LEO satellites, and the functionality of its terrestrial communication systems has been expanded to provide the environmental suitability for satellite applications. Filtronic is actively involved in supplying LEO demonstration hardware for both ground and payload applications.

Bold steps into new bandwidths

The solution to meeting the rapidly increasing demand for global broadband via LEO satellite constellations lies both in developing and manufacturing more commercially viable, high-performance RF components that can withstand the rigours of space, and in transitioning the bandwidths used for satellite communications into higher frequencies.

Since the u-Wave spectrum is limited and subject to many conflicting and overlapping demands, the development of feeder links operating in mmWave will be an important factor in the success of new satellite constellations.

While exploiting the possibilities of mmWave bands presents technological challenges, the experience we have in the UK of developing long-range terrestrial mmWave transceiver solutions with high data rates provides a solid foundation for further development to meet the demands of new LEO satellite systems.

Further reading:

Ground Segment Architectures for Large LEO Constellations with Feeder Links in EHF-bands, Iñigo del Portillo, Bruce Cameron, Edward Crawley. Massachusetts Institute of Technology, 2018 IEEE Aerospace Conference, March 2018.

Using E-band for Wideband SATCOM – Opportunities and Challenges. Sam Morrar Hughes Network Systems. Microwave Journal, August 2021.

Application of mmWave technology in High Altitude Pseudo Satellites (HAPS). Mike Geen, Filtronic. Microwave Journal, Feb 12th 2021.

The post Time to step up to mmWave to unlock potential of LEO satellites for global data connectivity  appeared first on Filtronic.

New TTA specification delivers greater consistency, audio quality and reliability for first-responder radio communications in North America

In 2018, one of the leading OEMs of critical communication equipment in North America changed its specification for the Tower Top Amplifiers (TTAs) installed in critical communications network base stations. This has now become their only specification of TTA available for new or replacement installations throughout the land mobile radio (LMR) network in North America. The rationale for the switch to this new specification was not widely publicised, and there remains limited understanding in the market about the benefits of the change.

Filtronic was instrumental in developing a TTA product to meet the new specification. Working closely with the OEM gave us a clear understanding of the thinking behind the new specification, the advantages it offers to radio system installers, and the improvements in LMR communications it brings to first responders.

In this paper, we outline the background to the specification change, the problems associated with the old system, the changes brought about by the new specification, and its benefits for public-safety communications in North America.

Background: Public-safety communications in North America

The current public-safety communications network in North America was established in the 1980s and early 1990s. The network of base stations aims to give 100% area coverage across North America, providing guaranteed communications connections for police and other emergency responders using land mobile radios in any location.

When the network was initially designed and implemented, it was based predominantly on analogue radio technologies, which were state-of-the-art at the time. This system proved invaluable, however, it was insecure, enabling new, easily obtainable radio scanners to pick up signals and listen in to police radio and other first-responder communications. Over the next decade or so, the radios were gradually updated to use digital modulation schemes and eventually the data was encrypted, making them secure and preventing casual eavesdropping. At the same time, channel bandwidths were reduced to enable more channels to be accommodated. More recently, the individual communication channel bandwidth is being halved again in order to double the number of channels – so the evolution continues.

However, while the radios themselves have been constantly upgraded, the radio frequency (RF) components and base stations that support the communications network have not really altered since the 1990s. This means that some components installed in base stations, such as power amplifiers, power combiners, filters, antennas and other RF conditioning products, have remained unchanged since their installation almost 30 years ago.

Technology: The role of Tower Top Amplifiers

One crucial component used in the majority of network base stations is the Tower Top Amplifier, which is used to improve receiver performance. TTAs incorporate a low noise amplifier installed at the top of the mast and a control/distribution unit installed at the base. The tower-top element incorporates a low-loss, bandpass filter to protect the receiver from out-of-band interference and a low noise amplifier to boost the received signal. The control/distribution unit at the base amplifies and splits the output signal to feed multiple different radio receivers. The tower top and base units are connected by a coaxial feeder cable running down the mast.

Over the years, many different companies have made the components for tower-top and base units, and there was some level of interoperability between them. But, over time, manufacturers opted to produce both the tower top and base elements together, meaning they had to be purchased as a pair which helped improve reliability and system performance.

An important feature of the original TTA specification was that the gain of the receive signal could be set to achieve the desired performance. That meant using attenuators to adjust the level of amplification and prevent the receiver from being overloaded. So, for example, in a dense urban environment where there is a lot of signal traffic, the gain can be reduced to prevent the receiver from picking up too many signals. This higher level of attenuation can also reduce system performance as well as restrict the range of the receiver. Conversely, for base stations in sparsely populated areas, the receiver needs to be far more sensitive so it can pick up signals from far away – so the attenuation would be reduced to achieve higher gain. At each base station, the gain would be set at the point of installation, according to the location of the site and its application.

Challenges: Complications caused by dual attenuators

The problem with the original specification for TTAs was that they featured two separate attenuators, providing two locations at which gain could be set. Both were housed in the control/distribution unit. The first ‘Reserve Gain Attenuator’ was located before the amplifier, and the second ‘Distribution Attenuator’ was located after the amplifier. As there are two locations for setting the gain, there are almost infinite possibilities for altering the ratio between them to achieve the same required level of overall gain. So, while the overall gain achieved might be the same, you would get different system performance depending on the balance between the settings of the two attenuators.  

The Reserve Gain Attenuator influences the sensitivity and noise figure of the system and its performance in the presence of high-power interfering signals. The Distribution Attenuator effects the system linearity. So, setting the first attenuator high and the second attenuator low, produces poor noise figure, poor range, but high immunity to interference. The other extreme is to have little or no attenuation at the input and all the attenuation after the distribution amplifier. In this case, you would achieve the same overall gain as the above scenario, but with very good noise figure and increased range, but greater susceptibility to interference.

Because there were multiple ways to set the two attenuators to achieve the same level of gain, different installers could set up sites differently. Each manufacturer provided guidance on how to set attenuation levels for different locations, with little constancy from one product to another, meaning it was up to individual installers to achieve the required gain levels for each site by setting the attenuators in whatever way they chose. This meant that once the base stations were operational, it was sometimes difficult to diagnose the cause of any underperformance, since the attenuators could have been set in many different ways. Any problems, such as audio drop-outs or poor call quality, were difficult to rectify without knowing how the two attenuators at each base station had been configured.

Solution: Delivering a new TTA specification

The potential problems caused by this uncertainty were one of the reasons why the leading OEM requested changes to the TTA specifications in 2018. The two significant changes made to the specification were:

1. To remove the second attenuator completely, and instead have a single adjustable attenuator located before the amplifier in the control/distribution unit. That gave installers a single way to adjust gain for the whole system.

2. To improve the low noise amplifier in the tower top by making it more linear, with better noise figure than previously required.

The new specification for Tower Top Amplifiers proved particularly challenging to achieve, and several RF suppliers attempted the task before a successful product was developed. Previously, TTAs had incorporated stand-alone filters, amplifiers and other connectorized components, which were cabled together in a waterproof housing. In the new product, all components are fully integrated into a single cast housing. As well as meeting all the performance standards stipulated by the OEM, this new product provides a lighter-weight solution in a smaller footprint, enabling better utilisation of space at the communication tower.

Benefits: Improved performance, control and reliability 

There are significant benefits of the new TTA specification for installers, end users and – ultimately – the general public. Fundamentally, each base station can now be set up very simply via a single control to optimise gain levels for the site, according to its location and the density of base stations in the region.

Having a single attenuator means there is only one point of adjustment to set the overall gain for the site. That immediately removes any ambiguity about how gain levels should be achieved. It makes setting gain levels far simpler and more consistent across the entire network. It means that if you’re in a remote rural area, you can simply set the attenuator to achieve a high system gain of 15dB, while in an urban area you set a low gain of 5dB. That eliminates idiosyncrasies in the way different installers set up gain levels at different base stations.

To compensate for the loss of an attenuator, the specification for the amplifier at the tower top has been significantly improved. This means there is no loss of performance, despite only having a single attenuator in the base unit. It gives you the best of both worlds, having the benefit of a tower-mounted amplifier with the gain effectively set to maximum, while providing the control simplicity of having a single attenuator to set the overall gain level for the site.

Outcome: Better communications for public safety

The new TTA specification supports better LMR performance for all emergency service providers. It gives radio system operators the confidence that their mission-critical networks will operate reliably with resilient connections and higher quality audio, especially in congested urban environments. Ultimately, that means greater peace of mind and security for the citizens who rely on these vital rapid-response services.

The post How upgrading tower top amplifier specification has improved first responder radio communications appeared first on Filtronic.

Originally printed in Microwave product digest May 2021

High-power gallium nitride (GaN) devices are becoming increasingly prevalent in the transmit chain of phased array antennas, not only for radar but more recently also for 5G active antennas. GaN’s high breakdown voltage and high power density, combined with its good thermal conductivity, have made it the technology of choice for power amplification in phased array antenna designs where size, reliability and efficiency are the prime considerations. This article explores the challenges in producing GaN-based products for these applications and discusses assembly techniques for improving yield and performance.

Why GaN for phased arrays?

Modern phased array radars can incorporate more than 10,000 elements, with the gain of the overall antenna structure being a function of the number of elements as well as the gain of each element. With element spacing fixed by electrical design constraints at around half a wavelength [1], packing this many elements into a defined space outline inevitably generates a large amount of heat, and this must be safely dissipated without affecting performance and reliability.

Massive MIMO (mMIMO) antennas for 5G base stations—currently with up to 64 transmit and 64 receive elements—are smaller, but power consumption and weight are major considerations. Already tower companies are becoming concerned about the weight of these active array antennas requiring a large crane for installation, and for tower construction to be reinforced.

Although still sometimes considered a ‘new technology’ GaN has already been used extensively for both military and commercial applications, and yet still has considerable potential to extend into the mmWave region.

Wolfspeed [2] has summarized the advantages of GaN as follows:

• GaN can achieve much higher power levels for a specific die size, due to its higher power density
• GaN-on-SiC has demonstrated reliability at higher junction temperatures than other technologies, and also has superior thermal handling capability due to the high thermal conductivity of SiC
• Due to having a wide bandgap, GaN can operate at wider bandwidths, as its smaller size for a given power level reduces output impedance and makes wideband matching easier
• The higher power density and optimal output impedance of GaN also allow it to achieve better efficiencies at higher power levels, which is particularly important in a phased array where multiple PA modules are densely packed.

Further to this last point, Qorvo [3] has quantified the thermal advantage of GaN over GaAs. For a one-million-hour median time to failure (MTTF), a GaN device can comfortably operate at a channel temperature 50°C higher than GaAs. All of these considerations mean that GaN is now the preferred technology for power amplification in phased array T/R modules.

Design challenges

In addition to the normal constraints of microwave design, the additional challenges in designing high-power, densely-packed transmit/receive (T/R) modules for phased arrays are both mechanical and thermal. Any type of packaging will have an impact on both performance and thermal impedance, and therefore reliability. One approach for the high power and dependability demanded by phased array systems is to integrate bare die into modules using surface-mount laminate as subsystem-level packaging, as shown in Figure 1. The commercial laminate material used for packaging can often be the same or similar to that of the PCB, thereby providing a good thermal match. Figure 2(a) and (b) show in more detail the features of the package.

Care is required to ensure that the thermal conductivity of the die attach is as high as possible, to dissipate heat and minimize the junction temperature. The material to which the die is bonded also needs to be closely matched in coefficient of thermal expansion (CTE) to avoid placing unnecessary strain on the chip and potentially causing damage to the die bond.

In the RF design phase, state-of-the art microwave design tools including 3D electromagnetic modeling, along with AWR Microwave Office and Keysight ADS, ensure sound electrical performance to meet demanding specifications, including maximizing power added efficiency to keep the junction temperature as low as possible for a given power level. Thermal simulation software is also used to optimize the dissipation of generated heat.

Direct integration of bare GaN die into microwave assemblies under cleanroom conditions requires in-house automatic die attach and wire bond capability. Vacuum picking from waffle packs or reels and fully-automated placement equipment (Figure 3) enables tightly controlled, accurate and repeatable die placement with X/Y accuracy of ±10µm. The accuracy of bare die placement directly impacts the bond wire length, which in turn may impact RF performance especially at mmWave frequencies. Using a bare die technique ensures that the highest level of performance is achieved without the need for the additional complexity of packaging. Hybrid construction methods offer the ability to mix and match technologies, and to include other components like quartz filters and printed microstrip components as well as packaged SMT passive components.

Addressing thermal challenges

It is critical for reliability that die-attach should be optimized to keep die temperature, and thus channel temperature, to a minimum. Three common methods of die attach are available—epoxy, eutectic, and pressure-less silver sintering. High conductivity epoxy typically has a thermal conductivity (k) value of up to 60W/mK, while gold-tin (AuSn) eutectic solder has a k value of approximately 57W/mK. Both of these methods have been used for attaching GaN die, but superior thermal performance can be obtained by using pressure-less silver sintering, which has a higher k in excess of 200W/mK.

The design of the module packaging is also crucial, and a solid copper paddle will disperse the heat as rapidly as possible, as shown in Figure 2(b). This provides a good thermal conductivity of typically 401W/mK, and eliminates ground via inductance since there are no vias in the ground paddle itself.

The challenge is that the high differential between the CTE of the SiC substrate of the GaN chip and that of copper potentially causing delamination under temperature cycling. Mounting the die onto copper molybdenum (CuMo) or copper tungsten (CuW) alloy improves the die-to-package CTE match. If epoxy die bonding is used, then a thicker epoxy layer can absorb some of the mismatch, but unfortunately at the expense of reducing the thermal conductivity of the bond.

It is often the case that a newly-assembled module will exhibit extremely good die shear performance, exceeding specification by an order of magnitude. When this is put through 1,000 temperature cycles, however, it can degrade to the extent that it will fail the die shear test. Considerable work has taken place on optimizing the die attach technology to prevent such failures, and test pieces are put through cycling to ensure this.

Minimizing voiding

Voiding—areas where the die is not attached—is one of the key reliability metrics for a GaN PA. MIL-STD-883 J TM 2012.9 specifies that the area of voiding underneath the die must not exceed 10% of the die area. Producing batches of between 5,000 and 10,000 high-power GaN amplifiers, Filtronic is currently achieving yields of 97% against the void specification.

With 2,500 square feet of Class 100,000 clean rooms at its UK hybrid microelectronics assembly facility, Filtronic has the capability to assemble microwave and mmWave devices into surface mountable (SMT) modules, in the form of system-in-package (SIP) and multi-chip modules (MCM). The microwave and mmWave devices could include chip scale ball grid arrays (BGA) and flip chip die.

In addition to low-void die attach and precision component placement, Filtronic’s assembly and test capability includes fully-automated wire and ribbon bonding with deep-access multi-level capability, sub-assembly manufacturing, skilled manual assembly, hermetic sealing, and automated test to 90GHz. MIL-STD-883 processes and procedures are supported by Six Sigma, along with high levels of production automation and strict traceability. As well as optimizing die-attach and heat-sinking for power devices, particular attention is also focused on minimizing wire-bond parasitics for products that operate at higher frequencies. Reliability and tolerance to harsh environments are evaluated in-house using high-temperature operating life (HTOL) testing, highly accelerated stress test (HAST) and humidity testing.

To date, over 60,000 T/R modules for phased array radar have been successfully shipped, at production rates in excess of 3,500 per month

References

[1]    Microwaves 101 Encyclopedia, https://www.microwaves101.com/encyclopedias/phased-array-antennas

[2]    ‘GaN MMICs for mmWave applications’, Jeremy Fisher, Wolfspeed, at 2019 Interlligent RF and Microwave Seminar https://youtu.be/xiGCylz6_aw [3]          ‘RF Applications of GaN for Dummies’, Qorvo eBook, https://www.qorvo.com/design-hub/ebooks/gan-for-dummies

The post GaN design and manufacturing challenges for phased array applications appeared first on Filtronic.

Mike Geen – Chief Scientist – Filtronic

Along with low earth orbit (LEO) satellite constellations, high-altitude platform station (HAPS) systems—or high-altitude pseudo-satellites—operating in the stratosphere, have the potential to address the challenge of providing ubiquitous connectivity. Although there has been great progress in rolling out high speed mobile networks to serve major centers of population, terrestrial connections will never realistically cover every part of earth’s surface. To deliver the full promise of 5G and address the ‘digital divide,’ it is essential to provide coverage to low population areas where terrestrial mobile networks are not viable. This will be particularly important not only for improving personal communications, but also because many Internet of Things (IoT) sensors will need to be in these regions. This article gives an overview of the role of HAPS and satellites in forming “networks in the sky” and describes some of the RF challenges in designing the high data rate (10 Gbps-plus) communication links needed to backhaul data between earth and the satellite or HAPS, and between the platforms themselves.

CURRENT LANDSCAPE

Successive generations of mobile communications technologies have been effective in covering the most highly populated areas of the world, and modern life has come to depend on this ubiquitous connectivity. Mobile network operators in most developed countries have worked hard to meet their targets of connecting typically around 98 percent of the population in terms of where they live. However, the 5G goal of being able to connect everything —wherever it is on the planet—remains elusive for terrestrial mobile systems. A ‘digital divide’ has been created between populations who have a broadband connection, whether fixed or mobile, at an acceptable speed and those who do not. The FCC currently defines this benchmark as 25 Mbps, which places some rural parts of the U.S. on the wrong side of the divide. Furthermore, mobile platforms such as aircraft and ships, along with many IoT devices located in isolated areas, will most likely fall outside the range of terrestrial 5G telecommunications systems.

This ambition to connect everyone and everything, wherever they are located, therefore cannot be fulfilled by ground-based communication networks alone. Therefore HAPS, operating in the stratosphere at an altitude of around 20 km, along with constellations of LEO satellites at an altitude of between 350 and 1,100 km, are beginning to be deployed to help address the challenge of providing ubiquitous connectivity. A HAPS station comprises an unmanned aerial vehicle – which may be either a gas-filled balloon, an airship structure or a fixed wing aircraft—and a payload that is essentially a moving 5G base station with onboard solar panels or fuel cells to provide power. As the technology evolves, we can expect to see non-terrestrial networks being converged with terrestrial infrastructure, as shown in Figure 1. This introduces some interesting technical challenges in creating reliable links between each of the elements of these converged networks.

SPECTRUM ALLOCATIONS

As demand for broadband capacity from these space and airborne systems grows, additional spectrum will be needed to support it. The telecommunications industry has successfully lobbied for more spectrum for HAPS, and the allocation of new wider bands around 26 and 38 GHz was agreed to during the 2019 World Radio Conference. In addition, experimental licenses were granted for E-Band (71 to 86 GHz), supporting the growing interest. Figure 2 summarizes the current and proposed bands, along with showing the geometry of the HAPS link to its ground station. There are commercial transceivers and high-power amplifiers available in these frequency ranges. Many have proven effective in XHaul (fronthaul, backhaul and midhaul) applications. Further development and qualification of these systems are exploited in the links that feed data between earth and the HAPS constellations.

While HAPS networks will mostly communicate directly with existing mobile phones, LEO satellite constellations will also generate the need for huge numbers of new fixed consumer terminals. As a result, this will create a demand for high volumes of components operating at frequencies up to 55 GHz where production volumes are historically low. The increase in manufacturing volume presents a challenge not only for OEMs but also for test equipment manufacturers. Simultaneously, it also offers new opportunities for companies who design and manufacture mmWave components and subsystems.

INDUSTRY INITIATIVES

The HAPS Alliance was formed in February 2020 with the aim of creating an ecosystem to promote the use of high-altitude stratospheric vehicles to extend connectivity to more people and locations worldwide. The alliance is based on an earlier initiative by HAPSMobile (a joint venture between Softbank and AeroVironment) and Alphabet’s Loon* to work together to advance the use of HAPS.

In addition to these companies who are developing the high-altitude platforms themselves, the alliance now counts among its members: telecommunications providers, including AT&T, Bharti Airtel, China Telecom, Intelsat, T-Mobile and Telefónica; aviation and aerospace companies like Airbus Defense and Space and Raven; and technology providers, including Ericsson, Nokia and Filtronic.

Further evidence of the growing interest in HAPS technology has been demonstrated by the formation of a Non-Terrestrial Connectivity Solutions project group within the Telecom Infra Project, an engineering-focused collaboration sponsored by Facebook.

HAPS USE CASES

There are several use cases for which HAPS technology is uniquely well suited, mostly related to their potential for extending enhanced mobile broadband to areas that are difficult to reach or not served by terrestrial mobile for economic reasons. These can include mountainous terrain, remote islands, marine regions and developing countries. A further application is mobile cell connectivity for passengers on board vessels or aircraft, or hybrid connectivity to reach passengers on board public transport vehicles like high speed trains, buses or river boats. They can also deliver fixed wireless access for users in isolated villages or remote industrial premises that cannot be reached by fiber.

Yet another use case is to supplement the capacity of existing networks to meet accelerating demand, and to supply “instant infrastructure” in emergency situations or for disaster relief. This can ensure network resilience for critical network links that require high availability. They can be rapidly deployed to cover a footprint of approximately 100 to 140 km diameter, over any type of terrain, requiring only minimal ground infrastructure.

HAPS can potentially be used to rapidly deliver media and entertainment content in multicast mode to RAN equipment at the network edge, to reduce latency for 5G cellular users, or in Direct-To-Node broadcast, whereby TV or multimedia services are delivered direct to home premises or to users on board a moving platform.

Utilizing their capacity rather than bandwidth, HAPS can be exploited for Massive Machine Type communications, to communicate with both local area and wide area IoT services.

SATELLITE AND HAPS SYNERGY

While it may appear that HAPS and LEO satellites both perform the same function, they are complementary. LEOs move rapidly in orbit relative to the earth so are more difficult to coordinate but have a wider area coverage. HAPS, being closer to the earth’s surface, are much easier and less costly to launch and deploy and give much lower latency. They feature “persistence”—the ability to remain stationary relative to the ground—and their transmissions are less affected by obstacles such as trees and buildings than terrestrial base stations. Unlike satellites, HAPS can be returned to earth for maintenance or payload re-configuration. They can quite literally be used as “base stations in the sky” providing additional capacity and wide cellular coverage in low density areas. They can also use the same frequency bands as terrestrial 3G, 4G or 5G networks, subject to regulatory approval, which means that the design of HAPS payloads should benefit from technology developments in 5G terrestrial base stations.

Experimental work with HAPS for communications relay has been taking place since the late 1960s,¹ but it is only since the growth of mobile Internet that their commercialization has accelerated. There are some differences between the two types of HAPS. For example, the solar-powered unmanned aircraft manufactured by AeroVironment and deployed by HAPSMobile are heavier than air and can only carry relatively small payloads. Another, fuel cell-powered fixed wing platforms, like those from Stratospheric Platforms, are emerging, which can carry much larger payloads. Finally, the airship-style platforms, along with the balloon platforms being used by Loon, are lighter than air and can usually carry a heavier payload as shown in Figure 3. These airships have been deployed to extend mobile networks over some regions of Africa and South America. What the two styles of platforms have in common is the flexibility to be upgraded to newer technologies, and their multi-purpose capability, as they can be adapted for institutional, commercial communications or civil security applications. They are payload-agnostic and can readily be networked together and linked with other platforms.

HAPS LINK CHARACTERISTICS

Breaking the converged network down into its individual elements, the characteristics that will be required for each of the links between them can be examined. Rain attenuation is one of the key factors to consider. Looking more closely at the HAPS feeder link geometry in Figure 2, for a platform at 20 km of altitude, the footprint can reach 70 to 80 km in diameter, and the link length will typically be about 40 km at an elevation of 30 degrees. Figure 4 shows the maximum rain attenuation per kilometer against frequency to permit 99.99 percent availability. For most of North America and Europe (Zone K), where rainfall is around 42 mm/hour, attenuation at 16 to 17 dB/km in E-Band is significantly higher than for the currently-used bands below 50 GHz. Nevertheless, as this service will be targeted at areas that have previously had no coverage at all, the demand for availability is likely to be lower than the 99.9 percent or 99.99 percent that is demanded of terrestrial networks, so some outage due to rain attenuation can be tolerated.

Potential data rates for HAPS feeder links in the different frequency band allocations are shown in Table 1, where they are compared for 256QAM and QPSK. Since antenna sizes are smaller at the higher frequencies, fade margins at 48 and 86 GHz are similar for the same antenna size and channel bandwidth. It is also striking that the data rate of over 10 GB/s that is available at E-Band using 256QAM is substantially higher than for the lower bands, due to the higher available channel bandwidth. Even if forced to drop back to QPSK to improve the fade margin, 2.5 GB/s data rate can still be achieved. The key figures for comparison are highlighted. Availability could be improved by combining E-Band with either 31 or 39.5 GHz.

Figure 5 shows the formula for calculating the maximum distance between HAPS, showing the distances calculated for the different limitations. The distance normally used between HAPS is around 200 km. Rain and clouds have far less effect when the platforms are at this altitude, and the greatest limitation is the curvature of the earth’s surface. Figure 6 shows the transmitter power that would be required for inter-HAPS links in the different frequency bands, indicating that high-altitude atmospheric losses at E-Band are very similar to those in V-Band below 50 GHz. The higher antenna gain for a particular size helps to compensate for free space loss.

SATELLITE LINK LIMITATIONS

Performing a similar analysis of the limitations on the feeder links for LEO satellites, Figure 7 shows the bands available and the geometry calculations involved. The orbits being used for SpaceX satellites are in the range of 350 to 1,100 km, and with an elevation of 35 degrees at the earth station this gives a link path length of between 600 and 1,700 km. Free space losses now become much more significant in comparison to atmospheric losses, as a higher proportion of the path is above the atmosphere. The free space loss is several tens of dB higher than the figures for HAPS feeder links – around 195 dB for a 1,700 km link length at 86 GHz.

When this is translated into system requirements, system gains between 180 and 200 dB are viable for mmWave satellite feed links in clear weather conditions, and this is achievable in E-Band with antenna gains of less than 58 dBi (equivalent to a 1 m parabolic antenna) and transmit powers less than +40 dBm. However, rain is a severe issue that can limit availability in both E-Band and V-Band, where increases in system gain of between 20 and 40 dB would be required to ensure acceptable availability. Inter-satellite links have no atmospheric limitations and fewer horizon issues, so LEOs can be many hundreds of kilometers apart.

MMIC TECHNOLOGY AND ACTIVE ANTENNAS

Although many power amplifier technologies at microwave frequencies now use GaN technology for better efficiency and higher power, it is rare to find GaN devices that work in V- or E-Band. Some experimental GaN devices have shown promising results up to around 100 GHz, but have not been commercialized yet, and it is challenging to find commercially available GaN devices above 40 GHz. Also, although SiGe and CMOS devices can work at the higher frequencies, their power levels are low, and many more elements would be needed to reach the required EIRP of around +60 dBm.

While phased array and active antenna technology has proved effective at increasing the gain and EIRP of antennas, as well as providing beam-steering, as the frequency increases the half-wavelength dimension becomes smaller and they become difficult to fabricate. Increasing the number of elements also increases power consumption, so using high-power GaAs devices with fewer elements has become the optimum solution for E-Band links.

E-BAND TRANSCEIVER TECHNOLOGY

Figure 9 shows a typical payload for a satellite or HAPS. As well as the eNodeB (LTE) or gNodeB (5G) base station, there are transceivers for the inter-platform links and the ground link, meaning that each platform would require three links at Ka-, V- or E-Band.

Figure 10 shows a block diagram of the basic transceiver that is being used for these E-Band links. It is fully integrated, for use in the 71 to 76 GHz and 81 to 86 GHz bands. It gives a highly linear transmitter output of more than 20 dBm and supports 256QAM modulation and above, along with a channel bandwidth more than 2 GHz. Phase noise is typically -112 dBc/Hz at 1 MHz. An integrated diplexer facilitates a single T/R port for the antenna interface, and there is also a single 50-way connector that supplies all communication between the module and the modem, as well as DC power, baseband data and control signals. A small, lightweight form factor is required, due to the number of elements to be accommodated and to optimize the overall weight of the payload. Morpheus II transceivers have a footprint of 90 × 80 mm and weigh only 110 g. For higher power levels, power combined amplifiers can be used – for example the Filtronic E-Band Cerus power amplifier can deliver output powers greater than 2 W. Figure 11 shows an example transceiver module, which is being used in E-Band links for terrestrial applications and has also been customized for HAPS/LEO links.

CONCLUSION

Global mobile data usage is growing rapidly, and the new use cases enabled by 5G will create demand in areas that are underserved – and difficult to serve – by terrestrial cellular networks, such as more remote and sparsely-populated regions. Terrestrial networks will not be able to provide 100 percent coverage, and this creates a clear case for converged networks that will integrate satellites and HAPS with terrestrial mobile networks. Although hybrid satellite solutions are not expected before 2023 to 2024, the standardization work for integration of the satellite in 5G networks is underway. Following the finalization of 3GPP Release 16 at the end of 2019, it is expected that provisions for satellite and HAPS systems could form part of Release 17. Since spectrum is limited and is subject to many conflicting and sometimes overlapping demands, mmWave frequency bands are expected to form a key part of the solution for the links between satellites, HAPS and earth terminals. The FCC has approved several trials in V-, E- and W-Bands, confirming the growing role for mmWave bands in the future.

While the use of mmWave bands presents some technological challenges for semiconductor devices, RF systems, antennas and network architectures, some long range mmWave transceiver solutions with high data rates up to 40 Gb/s and above have been developed and successfully demonstrated in trial systems and will improve for future systems.

For more information on this material, visit YouTube at https://youtu.be/bDGmkIF4-no.

The post E-Band mmWave technology for HAPS and LEO satellite systems appeared first on Filtronic.