Satellite communication systems consist of two main segments, the space segment and the earth or ground station. The ground station system coordinates the communication process with satellites in space. A communications satellite is an artificial satellite that relays and amplifies radio telecommunications signals via a transponder; it creates a communication channel between a source transmitter and a receiver at different locations on Earth.
Unlike the classical single big satellite in geostationary orbit (GEO) and medium earth orbit (MEO), small satellites in low Earth orbit (LEO) are typically organized as a satellite constellation. The satellite number of a LEO satellite constellation would be several hundred or even tens of thousands to realize a fully global coverage. Groups of satellites are deployed at the same altitude and inclination in several orbital planes.
There are lots of advantages for LEO satellite constellation, such as small propagation delay, low propagation loss, and high fault tolerance as well as network robustness. With the low orbit of LEO satellites, the one way propagation delay is only up to 15 ms, which is commensurate with that of terrestrial links. Meanwhile, much less power is required to overcome propagation loss and other atmospheric attenuation.
LEO constellations and tracking challenges
LEO satellites are in constant motion as they orbit Earth, so an individual satellite can only cover (or capture) small areas of the planet with each pass. So, many LEO constellations will be comprised of dozens, hundreds or thousands of small satellites. Some of the better known and now in-development constellations include SpaceX (4,000 satellites), Boeing (1,300+ satellites), OneWeb (600+ satellites), and LeoSat (100+ satellites).
However, the challenge that results from a moving constellation is that each singular satellite only has line of sight to an Earth station for a short period of time. Once the satellite moves beyond the field of view, the Earth station must seek link to a different satellite that has come into the field of view.
Consequently, there are a number of key metrics to deal with and things you need to ask yourself when planning your ground network architecture and service support:
- What geographical coverage is needed to fulfill the mission? (Direct Reception)
- How often do we need to contact our satellite per day? What’s the longest duration without contact in minutes? (Gap Characteristics)
- How much data – terabytes per day – do we plan to bring down to Earth? (Daily Data Volume)
- How quickly do we need the data to get down? How critical is the delivery time of the data reception, in other words the age of the data? (Data Latency)
- How much time before flyover do we need in order to make late commanding of the spacecraft? (Reactivity Time)
Special Polar Missions
If your spacecraft is passing the polar region, this poses a couple of additional challenges that you, as a LEO satellite operator, need to address early in the planning process.
- The polar regions suffer greatly from radio frequency (RF) interference because of the high amount of satellites crossing paths there.
- The space-to-ground contact over the polar region is limited by the few ground stations’ locations and assets. The alternatives are sometimes scarce.
One solution to both these issues is to combine several ground stations for a united connectivity. Site diversity is the most effective mitigation of radio frequency interference. By using several polar stations in different geographical locations, there is a reduced risk of radio frequency interference from simultaneous X-band downlinks closer to the North Pole.
For ordinary management and control operation of a LEO satellite constellation, the ground station is required to communicate with the satellite and provide telemetry, tracking and command (TT&C) service. It should steer the antenna beam to follow the satellite during the brief interval from its rise above the horizon until it sets below the horizon.
Limitations of Reflector Antennas
The period of a pass is relatively short for a LEO satellite, with a typical duration of a few minutes, depending on the relative position between the satellite and the ground station. Traditionally, reflector antennas are used as satellite ground station with high performance and medium cost. However, it can only support a limited number of satellites in its very narrow beam. This problem can be alleviated by increasing the number of reflector antennas at the mission control site.
However, the total cost of land, infrastructure, and antennas forbid such a mode of ’antenna farms’. Hence, reflector antennas can not meet the requirements of ordinary management and control operation of the future ultra-dense LEO satellite constellation.
Tracking and communicating with LEO satellites is challenging. LEOs move very quickly and most are only visible for 20 to 30 minutes during each pass. This requires an antenna that can acquire the signal, track the satellite’s path, and upload or download as much data as possible in this short amount of time.
Second, with so many satellites flying within each constellation, antennas must be able to communicate through handoffs from one satellite to the next to the next. Conventional antennas may require tens of seconds to locate and track a follow-on LEO satellite. This type of communications outage, though brief and predictable, is undesirable for data communications, and in many circumstances, such as voice or video communications, unacceptable.
Due to the parabolic geometry of the reflector surface, the reflector antennas usually produce a pencil beam and have little way to reduce the sidelobe. The first sidelobe level (SLL) may be as high as -13 dB. Such high SLL sometimes may even cause mis-tracking by the sidelobe rather than the main lobe.
The high duty cycle (constant movement and continual use) requires antennas that are rugged and high performing. The excessive wear and tear that comes from continual movement, as compared to a stationary GEO application, creates a different set of performance criteria for LEO and MEO ground stations. As the structure of phased array antennas are stationary and no mechanical rotation is required, it is much more stable than that of traditional reflector antennas, whose motors and servo will suffer friction and loss during daily use.
Phased Array Antennas for LEO Constellations
To solve the aforementioned problem, phased array antennas are considered as attractive candidates for future ground station of LEO satellite constellation. The primary reason is that phased array antennas can produce numbers of beams by the beamforming network thus support multiple satellites simultaneously. There are also other advantages including faster beam steering, better interfere mitigation, higher reliability, and lower cost.
A phased array antenna is a collection of antenna elements assembled together such that the radiation pattern of each individual element constructively combines with neighboring antennas to form an effective radiation pattern called the main lobe. The main lobe transmits radiated energy in the desired location while the antenna is designed to destructively interfere with signals in undesired directions, forming nulls and side lobes. The antenna array is designed to maximize the energy radiated in the main lobe while reducing the energy radiated in the side lobes to an acceptable level. The direction of radiation can be manipulated by changing the phase of the signal fed into each antenna element. Similarly, it is possible to change from a radiated beam to an effective null to absorb an interferer, making the object appear invisible, such as in stealth aircraft.
For phased array antenna, the beam direction is electronically steered by setting the value of phase shifts of the antenna elements, with a much faster speed in the order of milliseconds. By amplitude tapering and phase control of antenna elements, phased array antennas can control the SLL and present the nulls on the interfere directions conveniently.
Semiconductor IC-based phase adjustments can be made in nanoseconds such that we can change the direction of the radiation pattern to respond to new threats or users quickly. An additional benefit of a phased array antenna over a mechanical antenna is the ability to radiate multiple beams simultaneously, which could track multiple targets or manage multiple data streams of user data. This is accomplished by digital signal processing of the multiple data streams at baseband frequencies.
Furthermore, ESA antennas typically have a +/- 60-degree field of view, and though this may provide adequate steering for a constellation of large numbers, a mechanical positioner may be required for infrequent macro-level movements in combination with the electronic steering that would occur continually.
One emerging technology is the use of phased array (flat panel) antennas. These low-profile form factors can be either electronically steered arrays or fixed beam antennas that utilize a mechanical positioner. Electronically steered array (ESA) flat panel antennas can be steered very quickly – perhaps instantaneously – which eliminates keyhole issues and minimizes losses, but they also offer less gain than other types of antennas.
In the past, multiple analog microwave components are used to realize analog beamforming for phased array antennas The connections and interfaces are very complicated, strict microwave parameters and specifications are required. With the advance of digital beamforming (DBF)
technology, the beamforming process is much more straightforward to be done in the digital domain, where digitized data from antenna elements can be duplicated and combined easily.
Previously the cost of phased array antennas are very high, thus early applications were confined
to military and defense areas. However, this has changed in recent years. Thanks to advancements in highly reliable solid state devices and Microwave Monolithic Integrated Circuits
(MMIC) technologies, phased array antennas now become much affordable for commercial and industrial applications by increasingly matured mass production and dramatically reduced cost.
Eliminating single-point failure. Some possible single-point failures exist in the reflector antennas. When one part of them fails, it will cause the whole ground station unable to work. However, when some antenna elements of the phased array antenna fail, it will only suffer part of performance degradation without loss of functionality, making them much more reliable than reflector antennas.
For reflector antennas, there are centralized high power amplifier to boost the output signal. Usually, the capability of several hundreds of Watts is required for a typical satellite ground station. While in phased array antennas, each antenna element has its own power amplifier of only several Watts. Such a decentralized mode is much more reliable than the centralized one.
Circular polarized active electronically scanned arrays (AESAs) are of great interest for satellite communications (SATCOM). Phased arrays which are capable of maintaining circular polarization (CP) over wide scan angles are of great interest especially for new satellite constellation that are being deployed in Low-Earth Orbit (LEO) and Medium-Earth Orbit (MEO). LEO constellations are closer to earth (∼500-2000 km) and therefore move overhead at a much faster rate and require fast hand-off from horizon to horizon.
The typical implementation of this array uses patch antenna elements configured in equally spaced rows and columns with a 4 × 4 design implying 16 total elements. This antenna array can grow from small 4 × 4 array to quite large in ground-based radar systems, with over 100,000 elements being possible.
There are design trade-offs to consider with the size of the array vs. the power of each radiating element that impacts the directivity of the beam and effective radiated power. The antenna performance can be predicted by looking at some common figures of merit. Often, antenna designers look at the antenna gain and effective isotropic radiated power (EIRP), as well as a Gt/Tn
There are many aspects of design trade of a phased array antenna based satellite ground station. As the total gain of the antenna array is the product of the gain of the single antenna element and the gain of the array factor, the choice of the type of antenna element will influence the array overall performance. The higher the antenna gain, the fewer the elements required. The choice of practical use of types of antenna elements comes down to engineering trade-offs of many aspects,
including size, weight, bandwidth, gain, efficiency, and power handling capability. For example, the dipole antenna is simplest and well known, while the microstrip patch antenna is with low cost and easy to integrate with the T/R module.
Another key aspect of phased array antenna design is the spacing of the antenna elements. Once we have determined the system goals by setting the number of elements, the physical array diameter is largely driven by limits to each unit cell being less than approximately one-half wavelength, which prevents grating lobes. Grating lobes amount to energy radiated in undesired directions. This puts strict requirements on the electronics that go into the array to be small, low power, and low weight. The half-wavelength spacing creates particularly challenging designs at higher frequencies where the length of each unit cell becomes smaller. This drives the ICs at higher frequencies to be increasingly integrated, packaging solutions to become more advanced, and thermal management techniques to be simplified in spite of becoming increasingly challenging.
Most phased array antennas that have been designed in past years have used analog beamforming where the phase adjustment is done at RF or IF frequencies and there is one set of data converters for the entire antenna. There is increased interest in digital beamforming where there is one set of data converters per antenna element and the phase adjustment is done digitally in the FPGA or some data converters. There are many benefits to digital beamforming starting with the ability to transmit many beams easily or even change the number of beams almost instantly.
There are many approaches to synthesize amplitudes and phases of the array elements
that reduce the SLL. Analytical algorithms are one of the simplest kinds of methods among them, which can be directly calculated from mathematical expressions easily. Recently, some evolutionary algorithms, such as genetic algorithm (GA) and particle swarm optimization (PSO), are found to be very effective for much more complicated array pattern synthesis.
A low-cost phased array antenna is the holy grail for low Earth orbit (LEO) broadband constellations, necessary to ensure that earth station can dynamically steer communication beams to continuously track and maintain uninterrupted and high-speed connectivity to multiple satellites as they zip through the sky.
Less land and infrastructure. To support multiple missions with reflector antennas, a large area of the site and many antenna bases need to be constructed, along with concrete roads, water pipes, electrical supply, and optical fibers, etc. Since phased array antennas can produce multiple beams
simultaneously, it reduces the need for land and infrastructure construction to meet the same system requirements.
Mass production. Over the years, the high cost and complexity of phased array antennas were the major obstacles to use in commercial applications. Recently, the cost of phased array antennas has been reduced drastically by leveraging the advanced manufacturing technology, volume production, and adaptation of COTS components.
The availability of commercial silicon beamforming chipsets has resulted in the ability for phased array antennas to be ubiquitous in future communication systems. The fully integrated chipset eliminates the need for discrete transceiver blocks and includes a polarization switch, a transmit/receive (T/R) switch, low noise amplifier, power amplifier, phase shifters, and variable attenuators. The fully integrated chipset reduces overall size, cost, and RF losses. Several commercial chipsets are now widely available operating in the Ku- and Ka- bands.
Significant cost reduction of the T/R module is contributed by the advanced integration technologies. The area of the multilayer RF boards and the number of connectors and cables are dramatically reduced to save the cost.
Easier maintenance. The maintenance and/or impairment processes for reflector antennas are complicated. Sometimes it requires dedicated tools (like derrick, crane, and scaffold), and also crews of highly skilled workers. However, if some antenna elements of the phased array antenna fail, they can be easily replaced with new ones. The maintenance process is rather simple, one worker with basic training is usually enough to deal with the ordinary maintenance work. A much
lower life cycle cost and budget saving can be expected.
Technical issues and challenges
For a specified system effective isotopical radiation power (EIRP), as the power from each individual amplifier increases, the number of required antenna elements decreases. The power amplifier is one of the most critical components of the T/R module. Currently, there are a lot of available technologies of power amplifiers, such as Silicon, Silicon Carbide (SiC), Gallium Arsenide (GaAs), Gallium Nitride (GaN), Indium Phosphide (InP), and Silicon Germanium (SiGe).
The practical election of power amplifiers is based on the combined consideration of power generating capability, efficiency, cost of the selected technology. When the required power is less
than 0.1 W, Si-based materials are preferred for high density integration. Meanwhile, GaAs and GaN are usually used for power requirement about 1 W. Moreover, the selected power level must falls in the power handling capability of the antenna element, to avoid saturation or even breakdown of it.
With the progress of solid-state microwave technologies, the radiated power of the power amplifier increases steadily. Meanwhile, the miniaturization of the T/R modules results in
very compact size. Both of these two trends lead to increased heat generation and a rise in the working temperature of the T/R module. These would negatively affect the performance and reliability of the module.
In order to keep the reliable operation of the T/R module, it is necessary to dissipate the generated heat and below the temperature at a reasonable value. Air and liquid cooling are two commonly used techniques to cool the T/R module. Although with simpler equipment and relatively low cost, air cooling has limited capacity for heat dispassion.
For liquid cooling, the T/R module is cooled by pumping cold liquid through cooling channels of the cold plate, which is made of high performance heat conducting material and installed closely with heat generating components. The cost of a liquid cooling system is relatively expensive because of more complicated system with pumps, pipe networks, non spill connectors, and heat exchanging units.
The large antenna array consisting of thousands of elements has the possibility of failure of some of them. These element failures cause sharp variations in the field intensity across the
array aperture, thus increasing both the side lobe and ripple level of the array radiation pattern. In such a case, the first thing that needs to know is the exact locations of the failed elements in the array. It may be too expensive to distribute a network of sensors integrated with the beamforming network for monitoring the array status in real time. Therefore, it is best to perform the antenna array diagnosis by measuring the array radiation field of a given number of spatial directions
of the distorted radiation pattern.
Several evolutionary optimization techniques have been demonstrated to be effective to locate the failed array elements. The optimization process is to find a solution that minimize the difference
between the measured array pattern and the supposed array pattern containing several failed elements. One of the most popular optimization methods is the genetic algorithm (GA) for its natural advantage to use the gene to represent the discrete status of array elements. A different
technique, applicable also to non-planar arrays, is based on matrix inversion computation.
Amazon says it has completed “initial development” on the Ka-band phased array antenna that is “smaller and lighter” than legacy antenna designs. The prototype is already delivering speeds up to 400 Mbps, with performance expected to improve in future iterations. Current available commercial phased-array antennas cost upwards of $20,000 or more, while SpaceX’s Starlink antenna is estimated to cost $2400 per unit to produce in quantities of 1 million. A number of start-up companies have been trying to crack the challenges in mass producing antennas to drive down costs to where it would be a consumer item. Amazon says its main accomplishment is overlaying Ka-band transmit and receive elements, delivering an antenna that measures 12 inches in diameter. A smaller antenna reduces production costs significantly, as well as having an impact on shipping and options for end-user installation.