SUNSAT: Solutions for Remote Sensing
Sias Mostert and Garth W. Milne
Dept of Electrical and Electronic Engineering,
University of Stellenbosch
Tel: (021) 808 4103 Fax: (021) 808 4981
email : mostert@firga.sun.ac.za
SUNSAT 1R, to be launched in May 1996, is a technology demonstrator of micro satellite bus technology and carries two experimental payloads in communication and remote sensing. These experiments have been defined serving the Radio Amateur Community, making use of the Departmet of Electrical Engineering knowledge base and possibly providing useful commercial services.
The SUNSAT micro satellite has since its first definition also become the platform to carry a NASA GPS receiver, a magnetometer and school science experiments.
This paper looks at possible use of functions on the SUNSAT satellites as part of a larger mission for wider application.
The purpose of this paper is to examine the fundamental constraints on a micro satellite in LEO (Low Earth Orbit) and to what extent the SUNSAT satellite bus approach these limits. The satellite bus is then briefly described followed by a more detailed investigation into the remote sensing payload experiment on SUNSAT. The paper closes with some speculation on the maximum functionality of future payloads of micro satellites.
Micro satellites provide a number of unique advantages not found in larger satellites, but also suffer from a number of constraints. The most significant advantage is the cost of launching which is around $6 000 per kg compared to $40 000 per kg for larger satellites. Given the cost advantage and the miniaturisation of electronics and mechanics to build in the required functionality, the constraints amount to power availability and opportunity for launching.
Micro satellites since Sputnik have been the way for entering space. The Delta 2 launch vehicle paved the way for the new entries into space, by offering piggyback launches. Ariane followed suite and defined the ASAP standard. Today the ASAP standard is supported by Russian and Chinese launchers.
The advantage of being a piggyback is that the major cost for the launch is recovered from the main payload. This implies an affordable launch for the secondary payload. However, because the secondary payload is not the paying customer, it can only fly when there is a major payload. Israel is trying to address this issue with their small payload launcher with multiple last stage restart capability to provide precision in orbit injection for any LEO orbit at any time. The price tag is planned to be commensurate to the service and the expected cost is $2 Million for a 50 kg satellite.
The available orbits for the next 5 years will be mainly GPS (Global Positioning System) transfer orbits and GTO (Geostationary Transfer Orbits). A very restricted number of sun synchronous polar orbits might become available if the small remote sensing satellites are to be launched with a launcher with extra capacity. The Russian launch vehicles welcome micro satellites, but they are restricted to a maximum inclination of 83.
The primary power source for micro satellites is solar panels illuminated by the sun. The sun provides power at a density of 1.3 kW per square meter. The most efficient way of converting this power is with solar panels with efficiencies varying from 14% for Si cells to 18 % for GaAs cells. One square meter of Si panels can deliver a maximum of 182W in space. Allowing for degradation of the cells and area for interconnects, this figure drops to 150W.
Micro satellites have three options for the construction of solar panels. The most reliable is body mounted panels which provide one quarter (for four panels) of the available power which can be collected by the solar area. The next option is deployable panels with a fixed configuration. The third option is deployable panels which can track the sun. The last option is the most risky, but also provides the most benefit in terms of power availability. See table 1 for a short summary of the benefit and cost of the options.
Construction Technology Cost Power
Body mounted Si $100k 29W
panels
(total area 0.8 GaAs $300k 37W
sq m)
Deployable panels Si $100k+deploy up to 116W
mechanism
(total area 0.8 GaAs $300k+deploy up to 149W
sq m) mechanism
Table 1. Solar panel technology and construction mechanism options.
Batteries are the single most limiting item on the lifetime of a LEO spacecraft. Both
in terms of mass and the number of charge/discharge cycles. It is
noteworthy that the spacecraft life is significantly longer at the
same discharge depth for prismatic NiCd cells ($4000 each) compared to round NiCd ($50 each) cells.
Micro satellite construction fall within two categories. The tray based construction was popularised by Surrey University, while many of the modern day micro satellites are built in a modular fashion (Bremsat, BADR-B).
Traditional space craft architectures consist of highly reliable subsystems constructed from space qualified components. These subsystems are interconnected and the control is done centrally from the ground. Some of these systems are duplicated for redundancy.
Micro satellites can typically not afford the cost or lead times of space qualified components. Further more, they are typically constructed for LEO use, which does not permit continuos visibility and hence not continuos control. Therefor a micro satellite almost always contains on board processing for autonomous operation in case it is not visible from the ground station.
Figure 1. SUNSAT Construction with boom and tip mass.
The SUNSAT micro satellite construction is tray based which has proven flexible enough for the addition of a GPS receiver at a very late stage of the program. Further more, the imager, reaction wheels and batteries are also contained within the satellite without sacrificing the tray structure, see figure 1.
The power generation of SUNSAT is with four body mounted solar panels which provide 29 W of power compared to the 116 W of solar panel area available. The choice was made for reliable operation in stead of maximum power.
The SUNSAT launch vehicle interface was designed for compatibility with the Ariane ASAP standard. The Delta II-P91 mission is a particular important window of opportunity for SUNSAT as the non-Delta II standard envelope of the main payload enabled SUNSAT to be included as piggyback on this Delta launch. The stiffness requirements at the launch interface are however more strict than with Ariane and the lowest fundamental frequency is 64 Hz compared to 50 Hz required by Ariane.
The SUNSAT architecture consist of three on board computers and a central telecommand and a central telemetry system. All components are commercial and where possible extended temperature range. The on board computers can control the mission and control from the ground station is only required in case of failure by the on board computers and for maintenance of the operational status.
The SUNSAT program was initiated in July 1991 with the first advisory board meeting. The next semester was spent doing a feasibility study which culminated in the first intake of 12 post graduate students in Jan 1992. Since then we have accepted 6 to 9 students every year and a number of the students graduate every year after two or more years on the project. To enable us to finish SUNSAT on time for the launch, we extended our team to include the intake of 1993, as they have the most experience and give us the best chance for completing SUNSAT on time.
It is worthwhile investigating why we have embarked on a project of this scale. Each of the original goals are stated and some comments are made on the achievement of these goals.
The goals of the SUNSAT program and how we are achieving them are:
To date we have established an Electronic Systems Laboratory where up to 35 graduate students can work together. At this time we have 27 students working there.
International collaboration are enabling us to find our first launch with NASA on a Delta II. In exchange for the launch, SUNSAT is to carry a GPS receiver of which the data is of scientific interest to NASA.
Stimulating the youth of South Africa is one of the major priorities and is to gain momentum with a nation wide competition for schools to actually design and make simple space craft experiments. The four best designs and constructed devices will actually fly with the satellite.
The communication payload/experiment is of use to Radio Amateurs. It can however also solve the most basic communication requirement of electronic mail. The imager on SUNSAT can be used for land resource monitoring.
The CSIR optics division within Aerotek is responsible for the near diffration limit optics of the SUNSAT imager. The Hermanus Magnetic Observatory is responsible for the orientation and the scientific magnetometer on SUNSAT.
The SUNSAT micro satellite is composed of six major subsystems as shown in the block diagram of Figure 2. Detailed descriptions of applications of the remote sensing payload is provided in this paper. An overview of the communication payload is given. In this section we discuss the subsystems of the bus that support the payloads and science experiments.
Attitude determination and control system
The attitude determination and control specifications on SUNSAT are stringent for a micro satellite. The design goal is to be able to point the pushbroom imager bore sight to within 1 km accuracy from an 800 km altitude, which amounts to a pitch and roll error of less than 1.2 m radians. To minimise geometric distortion of the images, the linear disturbance velocity at the sub satellite point must be less than 66 m/s, which is less than 1% of ground speed.
Five types of attitude sensors are used. A 3-axis magnetometer is used to measure the strength and direction of the geomagnetic field. This low power (100 mW) device can be operated continuously to provide attitude accuracy to within 1 degree. Coarse attitude information is also derived to within 5 degrees from six cosine-law solar cells mounted on each facet of the satellite.
Horizon sensors, a fine sun sensor and a star sensor serve as the accurate attitude measuring devices. Attitude control is achieved through a passive gravity gradient boom, combined with two redundant active actuation methods. Slow attitude motions and coarse pointing to within 1 degree is achieved through magnetorquers. Accurate pointing and stabilisation during imaging is provided by 4 servo-motor driven reaction wheels. The position control resolution of the satellite is 0.1 m radian.
Flight control
Flight control is provided by redundant computers of differing type
. General flight management tasks such as scheduling, CCD imager control and communications management are performed by an Intel 386 processor, backed up by an Intel 80C188 processor. Both have access to all peripherals, but the 386 is the preferred flight controller. A T800 transputer is dedicated to the fine attitude control system, but its tasks can be taken over by the 386 in case of failure. Seven additional embedded micro controllers provide further support for telemetry, telecommand, power control and attitude control subsystems. A separate RAM disc of 64 Mega byte, which is accessible by both processors, is provided for storage of imager data or large files for store-and-forward applications.
Figure 2. SUNSAT functional block diagram
Telemetry and telecommand
The telemetry data collection function and the data transmission functions are duplicated for redundancy. The telecommand system, being extremely critical, also has a backup system implementation with discrete logic components and a dedicated receiver.
. Reliability against natural occurring errors and illegal commands have been prime design objective.
Power system
The power system is kept as simple as possible while providing for as many component failures as possible. The solar panels are connected to the battery charge regulators and the batteries directly to the power bus on the satellite. Distributed regulators ensure that the required voltages is supplied to each of the subsystems. In the event of failure of the battery charge regulator or the batteries, the solar panel output will be supplied directly to the power bus.
The peak power capability is 90W to handle peak loads during imaging data transmissions. Depth of discharge of the batteries are limited to 20% to ensure a lifetime of 5 years.
The SUNSAT communication payload was designed to provide a useful service to the Radio Amateur Community. At the same time the Radio Amateur services provide a technology demonstrator for other applications of store and forward communication. This paragraph investigates the communication payloads and the ground stations for supporting the communication payloads.
Store and forward communication between user terminals across the globe is the driving force behind the current Orbcomm, Temisat, SAFIR and IRIS (Larock94) projects. Two specific forms of communication is supported i.e. data collection, -dissemination, and message transfer between people i.e. email.
The mission requirements for store and forward communication impose requirements on the space craft, on the ground stations and the operations management of such a system. The main requirements for each system component is listed below.
At 800km altitude, the 5 elevation footprint has a diameter of 5080km, which spans 45 in longitude. The radio range varies from 800km to 2800 km compared to the geostationary satellite ranges of 36000km. Data communication with 10 Watt or lower powered transmitters and dipole antennas is practical, permitting data interchange with low cost terrestrial transceivers. Since large quantities of data can be stored in the satellite, global data transfer is possible. The communication applications include single access, multiple access, and broadcast facilities. The link level protocols supports store and forward message passing, e-mail, large file transfers and data collection and dissemination protocols.
The high resolution image data will be transmitted in real time via the S-band down link to reception stations at Stellenbosch and Johannesburg. A large 4.5m diameter antenna is already erected at the Stellenbosch image reception ground station. The addition of a L-band receiver implements a transponder capable of 2 Mbit/s between two 1.5m diameter antenna ground stations.
An experimental television relay service is planned as an Amateur experiment using the mode S transponder. It shall permit relay through the satellite of a TV signal between the Antarctic SANAE base and Stellenbosch while the satellite is in simultaneous view of both.
The Amateur Radio payload definition was approved at the SA-AMSAT Spacecon 91 Conference.
Store and forward digital packet radio will be provided, including 1200 baud AFSK communication for compatibility with terrestrial equipment common in South Africa. To provide sufficient up link channels, one of the VHF receivers support four channels to be connected to the 1200 baud modems. Three 9600 baud modems compatible with the G3RUH standard are carried, and can be switched to various receivers and transmitters. The AMSAT Pacsat Standard Protocols as well as a bulletin board will be supported.
A VHF Parrot Radio is intended for Novice category users in schools. Up-linked speech will be digitally stored and re-transmitted on the same frequency. The VHF 1200 baud AFSK mode will enable an operator in a remote area to access the satellite with the minimal digital equipment. It is ideal for the low population density situations common in the RSA. The VHF down-link will also be used to distribute the South Africa Radio League's weekly bulletin.
Most of these payload elements were tailored to fit into the available Amateur frequency band, but any of these services could be provided on the approved commercial frequencies.
Figure 3. The communication payloads on SUNSAT.
amount of data to be transferred compared to non-tracking antennas.Image collection ground station can only receive the contents of the RAM disk or a direct image transmission at 40Mbits/s. It would consist of a 4.5 m tracking dish with the additional receiver and computers.The 2 Mbit/s commercial service ground station would consist of a 1.5m tracking dish, the associated transceivers and computer.The Parrot ground station can be used for speech transmission. A simple 5W hand held transceiver should be adequate.V. COMMUNICATION SERVICES BY SUNSAT-->
perature and rainfall into the station's periodic position report. Several APRS weather reporting stations can form a real-time weather reporting network. Ships, yachts and weather buoys can be included in the network.B. Messages serviceSUNSAT supports transmission of to and from the ground of digital data at rates of 1200, 9600 and 2 Mega bits per second. The contents of the data can be messages between any one or more parties. Messages can be handled as point to point communication via the satellite or point to gateway to satellite communication. The first mentioned is effective for a traveler who is not always in range of a gateway station, while second mechanism is more effective in utilising the satellite bandwidth.C. EducationSUNSAT can be applied for distance education as described in this paragraph. However in the short term, SUNSAT is used as a vehicle to enthuse young people into science, engineering and technology. Three school related projects are described in this section.It is important to note that the satellite will have direct line of site with the whole Southern Africa during a typical pass. Educational broadcasts can be made via the Geostationary satellites and the feedback can be received via a LEO micro satellite which would required low power levels and a simple ground station.Three school projects are already underway to contribute to the RDP of South Africa with regard to education. A school flight experiment, ground telemetry stations and the parrot constitutes the three structure opportunities for schools to get involved.. Space related projects will capture the child's imagination and stimulate them to pursue careers in wealth creating technical jobs such as Engineering.The school experiment is a 30 cm square, 100mW device which schools can enter to actually fly on SUNSAT when launched in May 1996. The experiments will be judged and acceptance tested before integration in the final model. The ground telemetry station experiment is a minimum modem and USART subsystem which can plug into a receiver and a PC. With the necessary software (which will be supplied) the status of the SUNSAT and other micro satellites can be monitored. Scholars with Amateur Radio licenses will be able to communicate with each other via SUNSAT's transponder. The PARROT radio will provide a unique service with which younger children could send a message to SUNSAT via radio. The message is then sampled and stored in on board RAM. After a short delay, the "Parrot" will convert the message back to a voice signal and transmit it to everyone listening.-->
Commercial remote sensing is only found in governments paying for weather services and military observation. There are various other applications proposed, but combined they have not been able to sustain the world remote sensing providers on a commercial basis. The major space faring nations and space companies do however continue to invest in remote sensing (Howes94). ie. SeaStar1, WorldView 1 and 2, Ofeq 3, Clark and Lewis.
The role of a micro satellite in remote sensing is to provide a very inexpensive platform which can be configured and launched for a very specific mission in a short time. Other roles include dedicated constellations for specific application areas due to the low cost of one such satellite. In all cases it would require strategic investments from governments or large organisations to provide them with a strategic or tactical advantage.
Spatial resolution is the diameter of a round object which will be represented within one pixel. The swadth width is the width of the ground track. This is a direct function of the number of pixels and the size of the pixel on the ground.
The number of bits which represent the intensity of each pixel. This determines the dynamic range of the signal. There is a tradeoff between the dynamic range of the signal and the bandwidth required to transmit it.
The number of days between consequtive visits to the same location on the ground. This is typilcally 26 days for the SPOT satellite at an altitude of 832 km.
The spectrum useful for remote sensing varies from ultra violet to visible (0.4 to 2.5 m), thermal infrared (3.5 - 4.2m and 8 - 12 m) and microwaves (from 5 mm). Within the broad bands there is a further subdivision possible based on the information available in a specific band. Refer to figure XXX for more details.
The pointing accuracy of micro satellites are to within 1 degree. This is adequate for a snap shot imager. The SUNSAT imager is a push broom imager which provides a larger swadth width, but the associated stricter requirements on pointing accuracy and orbit stability.
The amount of data from remote sensing satellites are enormous (64 Gbyte per day for the 4 hour operation every day of the SUNSAT imager). This require high bandwidth communication to convenient located ground stations. If a ground station is not in sight of the satellite footprint when imaging, then the image must be stored for later transmission.
The orbit requirements for applications where the same lighting conditions must apply, is polar sun synchronous. There are however cartography and reconnaisance applications where this restriction does not apply and revisit periode is more important which specifies a lower inclination orbit for locations with a lower latitude.
A user would like a reliable and predictable delivery time service. This can be more easily achieved with a number of satellites in stead of one large satellite.
The SUNSAT imager has a planned spatial resolution of 15 meter per pixel at 800km. This is with a focal length of 570 mm and a lens diameter of 100 mm. The swadth width is 51km with the sensor of 3490 pixels. The radiometric resolution is 8 bits which was evaluated as acceptable tradeoff between data rate and dynamic range (Moon92). The temporal resolution is a function of the orbit and with a sun synchronous polar orbit, 600km gives the best results, see fig 6. However, the orbit offered to SUNSAT 1R is elliptical 800 km to 400 km and not sun synchronous. With the view that SUNSAT 1R is a technology demonstrator, this is acceptable and should lead to various unique images possible due to the varying sun angle.
The spectrial frequencies chosen are compatible with Landsat Thematic mapper bands 2, 3 and 4. The three bands are green (520 to 610 nm), red ( 610 to 700 nm) and near infra red (700 to 900 nm). These bands were chosen as users are familiar with images in these bands and the bands are useful for biomass production prediction.
The pointing accuracy of the imager is of crucial importance for such a high resolution imager. The specification of a 1 km maximum error with a swadth width of 51 km will ensure that the correct target is located and photographed.
The ground segment of SUNSAT consists of a 4.5 m parabolic dish antenna which can download the data stream in real time. On board storage of 64 Mega byte enables SUNSAT to store a square stereo pair full colour image for later transmission to a ground station. Data compression and correction on the data stream leads to an improvement in the amount of data stored and the integrity of the data in the case of a marginal link budget.
Two types of ground terminals are required, ie. the full data reception ground station which should be transportable and an end user ground terminal which provides application specific images which are pre-processed and does not require the full bandwidth and complexity of a full data reception ground station. The last mentioned ground station can be mobile or fixed.
The operations management requires a 24h manned station to accept user requests, a full data reception groundstation to download images and a data dissemination system which can provide access to images via email or medium speed communication services aboard the satellites.
The operational reliability and service quality can be much higher with a constellation of SUNSATs. At a price tag of $6 Million per professional satellite, the cost of establishing a commercial constellation of six satellites, is small compared to the magnitude of the large single resource remote sensing satellites.
Satellite remote sensing has the fundamental advantage that it is available at a fixed rate with no additional effort. This enables remote sensing to be used for the monitoring of environmentatl processes which change with a periode of between 5 days and one year.
Timely distribution of the data is of utmost importance and all SUNSAT images (which were ordere) will be available within 24 hours of being taken on an FTP server from the ground station.
Remote sensing satellites can be classified into three categories:
SUNSAT 1R is the first research satellite in a series of commercial and follow on research satellites. The commercial satellites and the follow on research satellites will be developed in parallel.
Two options exist within the long term planning of SUNSAT. The first satellite is a duolicate of SUNSAT 1R with the specific functions required by the customers.
The second option is the next generation satellite which will be compatible with the next generaion Delta launch vehicle interface. This is a 12"X12"X16" structure weighing in at 35kg. This satellite could be tray based or of modular construction. To retain the current functionality it would have to be miniturised.
Future communication payloads will include encoding to improve the link budget for low speed, small ground terminals. At the upper end of the scale an X-band link is planned to cope with the expected higher data rate on the imager down link. Inter satellite links are already possible with the current architecture. Due to the lack of orbit adjustment, this is however only for experimental purposes.
The power consumption while imaging and transmitting the image in real-time, is 90 Watt. Future research SUNSAT satellites are planned to have a deployable solar panel, tracking the SUN to provide a 100 % duty cycle over the useful range of lighting conditions.
Imagers with a focal length of 1710 mm and a lens diameter of 200mm will lead to a spatial resolution of 5m at 800 km with the three spectral bands chosen. A sensor array of 10 000 pixels will then lead to a downlink data rate of 100Mega bits per second. This will require larger amounts of power and possibly a downlink in X-band to cope with the data rate.
This paper described the fundamental constraints on micro satellites, showed how the SUNSAT 1R micro satellite weighs up against the criteria and how the imager payload experiment can be utilised to provide useful services.
SUNSAT was conceived as a training vehicle for post graduate students and as such it has excelled. In the final leg to the launch date, the application of the SUNSAT services are investigated as a means of providing specific solutions which can only be completely satisfied with a space segment or a combined space and ground segment solution. In particular the remote sensing services were investigated. The communication services are described in (Mostert95).
We are looking for partners to develop specific applications based on the unique features which a low earth orbiting satellite provide. SUNSAT is of particular interest as it combines a high resolution imager with communication services in the same platform, making it suitable for combined communication and remote sensing missions.
SUNSAT incorporates a very innovative combination of a high accuracy attitude determination and control system with a push broom imager. This opens up new opportunities for the application of micro space to real-size problems.
Acknowledgments
SUNSAT is a product of past and present enthusiastic post graduate students, dedicated lecturers and devoted support personnel. Every single one of them is acknowledged in their quest for getting South Africa's first built satellite into orbit.
Our sponsors are acknowledged for staying with us from the beginning. In particular Grinaker Electronics, Plessey Tellumat, AMS, Siemens, Altech and First National Bank Technology Division. Other sponsors which have joined the band is the FRD and Reumech. Particular gratitude goes to ODF Technologies and Orbicom which have enabled us to establish our 4.5 meter dish for image reception.