Australian CubeSat to use 76 GHz The IARU Satellite Coordination Panel has announced the amateur radio frequencies for the Australian 76 GHz CubeSat CUAVA-1 that is expected to launch in July 2019 CUAVA-1 is a 3U CubeSat and the first CubeSat project of the new ARC Training Centre for CubeSats, Uncrewed Aerial Vehicles (UAVs), and their Applications (CUAVA), whose primary aim is the education and training of people, mostly PhD students, for the space sector. With significant heritage from the QB50 CubeSat INSPIRE-2, CUAVA-1 is a 3U CubeSat that will link with the international radio amateur community for outreach, training, and increased data downloads, observe the Earth with a novel multi-spectral imager, use a GPS instrument to explore radio occultation and the reception of GPS signals scattered off the Earth as well as provide a backup determination of the CubeSat location, investigate plasma environment and associated space weather with radiation detectors, and explore the performance of a new communications payload. This mission addresses issues of radio technique interesting to the radio amateur community in the following ways: 1) Global Radio Amateur Participation in Mission and Data Downlinking We will work with radio amateurs and other groups to receive and decode the spacecraft beacon and downlinked data, with subsequent transfer to the internet database (ideally the SatNOGS database). In detail, the CubeSat will transmit data, especially recent images over the terrestrial footprint, to participating radio amateurs across the globe. This will directly involve radio amateurs in the mission and its success, by greatly increasing the overall amount of downlinked data available and having the images be directly relevant to the receiving people. The receiving station and people would be identified in the database and then acknowledged in any publications resulting. The mission’s success will thus be directly tied to the involvement of the international radio amateur community. In addition, the mission should provide multiple opportunities for enhanced outreach and training for both the global amateur radio satellite communities and CUAVA. 2) Student and Radio Amateur Participation in the Groundstation We will train students and desiring radio amateurs in the setup and use of a groundstation hosted by the University of Sydney and then have these people operate the groundstation (including control of the satellite and managing the uplink and downlink) and transfer downlinked data into an internet database (ideally the SatNOGS database). This will involve existing radio clubs in the training, increasing their memberships and leading to new clubs and people familiar with the international radio amateur and satellite communities. 3) Radio Wave Propagation The ionosphere, thermosphere, and lower atmosphere have multiple effects on the propagation and absorption of radio waves and microwaves. This mission will study the electron number density as a function of position, time of day, and space weather events using the ``radio occultation’’ of GPS signals and their associated refraction and attenuation. These data will be published and made available for ionospheric research via a website, and provided to Australia’s Bureau of Meteorology and other space weather organisations worldwide. These data are used to predict maximum and minimum usable frequencies for radio amateurs (and both commercial and government users). In addition, the GPS signal attenuation and electron number density profiles can be used to extract the amount of water as a function of height and used to predict ordinary weather. This work will also add to knowledge of the orbital environment via the drag forces and decay of satellites depending on the gas and plasma densities. 4) Communication Protocols Modulation techniques that will be investigated for the high-speed communications experiment include QPSK, 16-QAM and CPFM. If successful, this technology for wavelengths below 10 cm will increase the data transfer rates by at least 4 orders of magnitude while also decreasing the sizes of antennas and the associated spacecraft. This experiment will be relevant to spacecraft-toground and inter-spacecraft communication links and is particularly relevant to radio amateurs, universities, and their students and staff, due to the dramatic increases in data rates and capabilities and associated dramatic reductions in costs. In addition, the use of multiple frequencies is important for rain (and moisture content) attenuation mitigation techniques, as well as to provide another data stream for weather prediction. 5) Radiation Effects on Electronic Components The Low Earth Orbit (LEO) environment is protected from cosmic rays, solar particles, and particles trapped in the Van Allen Belts by Earth’s magnetic field. Some portions of LEO do harbour regions of enhanced radiation, in the auroral zones and the South Atlantic Anomaly (SAA) for example. In addition, transient solar and magnetospheric particle energization events, a major component of space weather, can change the radiation level by orders of magnitude. This radiation can adversely affect spacecraft which pass through them. This mission will directly measure the counts of energetic particles as a function of space weather activity, position, and time of day, thereby characterising the Earth’s radiation environment. It will also study the effects of the radiation on the computer and other onboard electronics. Examples of effects include single event upsets (SEUs), degraded solar cells, and non-functioning electronics such as radio receivers and transmitters. 6) Attitude and Position Determination Reception and analysis of GPS signals by the onboard GPS receiver will determine the spacecraft’s attitude and location as a function of time, thereby determining the satellite’s orbit. Comparisons with NORAD radar-derived orbits will test the on-board GPS receiver and measure drag and other effects. These orbits are vital for radio amateurs interested in testing and characterising their radio equipment, as well as in downloading the satellite beacon and data signals for transmission via the web to the satellite project and the international community. Proposing to downlink telemetry on 9k6 GMSK AX25 on UHF and high speed downlinks on 2.4 GHz, 5.6 GHz and 76 GHz. Planning a launch from Japan in July 2019 into a 400km orbit. These frequencies have been coordinated by the IARU: Downlinks: 437.075 MHz, 2404.000 MHz, 5840 MHz and 76.800 GHz Uplinks: 145.875 MHz, 2404.000 MHz and 5660.000 MHz More information on CUAVA-1 can be found at https://www.cuava.com.au/ https://twitter.com/Arc_Cuava IARU Satellite Frequency Coordination Panel http://www.amsat.org.uk/iaru/ http://www.southgatearc.org/news/2019/april/australian-cubesat-to-use-76-ghz.htm
This is very good news for those who are interested in the millimeter wave Amateur Radio bands! I wish the team the best of luck in the launch and hope to copy the downlink on 76GHz when they are QRV.
LEO cubesat at 400km orbit with its 76GHz will have excessive Doppler. I’m pretty sure it won’t be easy to work it. Even 13cm band will cause some plus-minus 60kHz or so. At 76GHz Doppler could reach I suppose plus-minus 2GHz.
Hmmmm using 76ghz would be like using a pinhole to see the grand canyon while the pinhole travels 17,000mph as you try to hit the spinning pinhole with a laser pen back on earth... but other than that let’s see what happens ...
I used an online calculator to approximate the Doppler shift. Just using the 17,000 mph number given by W0IW (7,600 m/s), the number I got was about 3.3 GHZ. This does not use any combined effects of the earth's rotation, and the altitude of the orbit to find the apparent or relative velocity which could make it either larger or smaller. Luckily, 76.8 GHz is only a downlink frequency. I assume the reason for this is item No. 4 in the press release: Experiments with data modulation techniques for achieving higher data-rates and using relatively small, high gain antennas at this frequency. I doubt anyone will be using 16-QAM on the 2m uplink. The modulation techniques they listed are pretty common on geostationary Ku and Ka band satellites (zero Doppler). 76 GHz would be classified as V band by IEEE and EHF by the ITU. The next ITU band up is THF ("Tremendously High Frequency") and then it's IR. Back to the Doppler problem for a moment. I'm just thinking as I type here... Unless I'm wrong, you can downconvert the signal into something manageable but you're still stuck with the full Doppler shift. There is no way to scale that. For example, if you downconverted it to let's say 4 GHz, you'd need to be able to tune from 0.700 GHz to 7.3 GHz (+/- 3.3 GHz). A daunting task to make a receiver that covers that range. Instead, my approach would be to have the LO at the LNB (Low Noise Block) be a VFO. If the LO/VFO was centered around 75.5 GHz, you could downconvert the signal to approx. the center of the 23 cm band. In this range you could begin to use conventional ham gear to do the rest of the reception/demodulation. Also, signals in the 23 cm (aka "L band") region can go a modest distance on good coax without being completely lost. An LO/VFO in this range would need to be able to tune from about 72.2 to 78.8 GHz (+/- 3.3 GHz) and would need to have good stability. I'm not sure how you'd go about making that but it's the first approach that comes to my mind. https://www.omnicalculator.com/physics/doppler-effect
The antenna is actually quite easy to imagine. The only thing that makes sense is a parabolic or spherical reflector dish antenna. With these antennas, gain increases inversely proportional to wavelength. As the frequency goes up and the wavelength gets shorter, the reflector is a larger multiple of the wavelength. Hence, larger dishes have more gain than smaller ones. Take a look at this PDF document referencing work by W3HMS and K4ITO: http://f1chf.free.fr/hyper/Parabolic dish calculations.pdf There is a link in it to a spreadsheet created by the same people: http://f1chf.free.fr/hyper/ERP Calc 20 April 2010mod by JL.xls (The actual hyperlink embedded in the PDF is truncated and therefore busted.) If you enter some basic information at the top of the spreadsheet, you can quickly get the gain, ERP and beam width for a variety of dish antenna sizes. The current generation DirecTV antenna is 26" x 36" oval. It's not 100% accurate, but I'll average the mean diameter to 28" which is about 0.71 m. I inserted some blank rows in the sheet and created calculations for dishes from 0.1 m to 0.9 m in 0.1 m increments by copying the formulas. Everything is based on the value in column A. According to the spreadsheet, a 0.7 m dish should produce 42.1 dBi of gain at 76.8 GHz (76,800 MHz). As further reference points, a 1.0 m dish would create 45.2 dBi and a dish of only 0.2m (about 8") would still give 31.2 dBi. By comparison, at with a Ku band frequency of 13 GHz, a DirecTV antenna would only have approx. 26.7 dBi. The comparison is a little unfair because DirecTV satellites use "fully saturated" transponders of 27, 36, or even 54 MHz. This means more energy under the total curve. However, gain is still gain. A good commercial satellite LNB (Low Noise Block, the active circuit that the RF is focused onto), has a typical gain in the vicinity of 62 dB with a noise figure of 0.7 to 0.9 dB so about 61 dB of usable gain. Combine this with the 42 dBi from the DirecTV sized antenna and you have over 100 dB of gain. One more point for comparison, I work with 1.2 m antennas receiving Ku signals of approx. 11 GHz. Our antennas yield about 30 dBi of gain combined with 61 dB from the LNB, so around 91 dB total and a beam width of 5.3 degrees. We use these with geostationary satellites, so we don't have any Doppler to deal with. It seems at least plausible that an 8" antenna used with a not-too-exotic LNB could yield over 90 dB of gain and a beam width of 4.5 degrees. This is what I'm guessing they're trying to experiment with, but there's still that tremendous Doppler shift to be wrestled with. At this point I'm in danger of getting into the deep end of the gain pool because to do this all properly, you need to do what is known in the satellite business as a "link budget" calculation and many more factors come into play including things like transmitted power up and down, distance to the satellite, sky noise, etc.
The basic mechanics of tracking a LEO are not particularly complicated per se. No more so than any other satellite. In theory, conventional ham AZ-EL rotators with satellite tracking software could do this. As I see it, where this gets complicated is that even if we can use an 8" dish like I mentioned, the 4.5 degree beam width could be tricky to aim. It's fairly narrow and will be hard to keep on target. Typical ham rotators aren't precise or accurate enough to maintain this degree of aiming precision. You'd also need something that can slew it's rate (accelerate and decelerate) more gracefully than run-of-the-mill ham gear. I'd probably try to do something with a higher degree of precision, accuracy, and better slew rate control. Perhaps using stepper motors. Since the antenna will be relatively small, I could enclose the whole thing in a radome to weatherproof it and it would still be less than 2 ft. in diameter. Maybe even under a foot if I were clever about it. So there's still that tricky Doppler problem that was discussed above. I don't have a good solution for this but I'm figuring that someone else has or they wouldn't be trying it. Yesterday I mentioned making the LO (local oscillator) in the LNB a VFO. More likely it would be a VCO (voltage controlled oscillator) than a VFO. Some web surfing validates that work is being done on VCOs in the vicinity of 75 GHz that I suggested: http://cc.ee.ntu.edu.tw/~jrilee/publications/75G_pll.pdf https://ieeexplore.ieee.org/document/782080 https://www.semanticscholar.org/pap...hner/bfaf33e8693a6f8dbbf7bb38de092d5c533adbf6 The last paper claims to have a tunable range of 4.5 GHz. Not the full 6.6 GHz (+/- 3.3) that I calculated earlier but enough range that the Doppler could be tracked over a wide section of the orbital pass. Remember that at the relative apogee of the pass (not the satellite's orbital apogee, but the highest relative point it achieves during a pass), the Doppler shift is zero since the satellite is, at that moment, neither approaching you or receding from you. For an appreciable period of time either side of this, you could track the satellite with +/- 2.25 GHz of Doppler shift. While this would not always give you horizon-to-horizon coverage, it would give you enough to experiment with. My idea so far has just been direct down conversion from 76 GHz. They could add other mixers and steps in the down conversion process but now you're essentially into the area of heterodyne receiver design. This is all me just typing as I think. Presumably they have a more elegant plan than what I've imagined. This is a group back and funded by 2 Australian universities, 4 Australian companies, 4 government agencies (Australian?), and 2 U.S. Universities. I have to imagine they have some technology at their disposal. http://www.acser.unsw.edu.au/sites/acser/files/u118/TUE-1-5-JamesHarpur.pdf In the presentation found at the link above, it mentions that this satellite will have a Reaction Wheel https://en.wikipedia.org/wiki/Reaction_wheel This explains how they will be able to stabilize the satellite and keep it pointed accurately at the ground. On a non-technical note, I wonder, given the makeup of the group behind this and that there seems to be a large commercial involvement, is this possibly industry and/or government using the amateur bands for research on a basic level? Even if the raw science gained is of limited value, is industry using this (and therefore the ham bands) as a training ground for brilliant minds who will someday apply the lessons learned for real commercial applications. By using the ham bands, they avoid a lot of red tape to get commercial or government frequency allocations.
To KE2D: I think that your Doppler shift calculation contains a big mistake. The maximum Doppler shift is possible +/- 2.02 MHz on 76.8GHz.
The SpaceX Starlink LEO satellites will be using frequencies as high as 30 GHz initially and up to 50 GHz eventually. I'd say the Doppler problem is solvable.
UA4HAK, Thank you. I stand corrected. Your number is correct. I had used an online calculator and left a default value set as one of the parameters which I should have checked and I misread the result. I redid the calculation with the original calculator and another one and I now concur that the number is about 2 MHz. This makes the Doppler shift problem far more manageable. Since the net Doppler shift is now only 4 MHz (+/- 2 MHz), none of the approach I spoke of above with a VCO in the LNB would be necessary. You could simply use an LNB with an LO of 75.53 GHz and it would produce an L band output in the middle of the 23 cm band (1270 MHz). This could be easily tracked by tuning the VFO of a conventional amateur radio such as the Kenwood TS-2000X. Alternatively, you could use an LO of 76.365 GHz and get an output at 435 MHz in the middle of the 70 cm band and there are many radios which could tune this range easily and cover the whole Doppler shift. Using 435 MHz as the IF may be simpler since the signal will survive better over coax. Of course you could also calculate other LO frequencies which would produce IF frequencies in the 2m and 6m bands if you preferred to use those (76.654 and 76.747 GHz). Each of those bands has 4 MHz of bandwidth and the majority of HF rigs today also cover 6m. I imagine LO stability will be challenging at that frequency. Now we just need someone to make us 0.2m dishes with tiny little LNBs at 76.747 GHz and a precise motorized tracking system. Calibrating the tracking system could be a challenge also I imagine. I wonder if the sun puts out anything in this range? It is only about 1/2 a degree in angular diameter and it's position is well known. MEANWHILE... If anyone was following the news about the SpaceX Starlink launch the other day (24 May) and your read the Wikipedia article about Starlink, you'll see that they plan to use the V band (75 GHz) in the future with initial operation on Ku and Ka. https://en.wikipedia.org/wiki/Starlink_(satellite_constellation) This would explain why Australian companies and government agencies are interested in experimenting with satellites operating in this frequency range. I suspect there is probably a lot of commercial and other interests which would like to gain some quick knowledge in this area. I don't think it's a coincidence that Starlink plans to use V band and that this satellite will be experimenting with it also.
