XC04 Security, defence and resilience
Depth of coverage
At its target depth of L2. What the depth levels mean.
What this layer covers
Dual use, attack surface, jamming and interference, physical vulnerability, energy security, supply chain security
Where it ends
Cross-cutting
Deep area
This layer is studied in more depth than most. It belongs to the deep area Power electronics and ground segment.
Security, defence and resilience
The risks in these records include attack from the ground and a forged control signal. Secure World Foundation's 2026 open-source assessment judges it unlikely that China now has a working ground-launched anti-satellite missile that can reach geostationary orbit, the high orbit where a satellite stays over one point on Earth, though one is likely in development (CLM-0741). For a target there, the assessment estimates several hours of warning after a detected launch. No source held here shows whether a large structure could use that warning (CLM-0744). A 2019 assessment co-written with US defence-affiliated researchers judges that a solar power satellite could become more vulnerable as space becomes more contested, and suggests many smaller satellites in place of a few large ones (CLM-0230). The control link is the radio signal from the ground that points the power beam. A patent filed in 2017 for the CASSIOPeiA design discloses no check that this signal is genuine. Patents often leave out security measures, so this says little, and nothing about the current design (CLM-0899). An analysis of the published design, made for these records and not a test, finds that of the failures analysed, a forged signal is the only one that would steer the beam toward the forger (CLM-1478). Viasat, a satellite operator, says the February 2022 attack on its KA-SAT network came through a ground network (CLM-0757).
No quantified threat assessment for a large power satellite has been found. Two programme studies judged the beam alone not to be a real weapon, though it might be used to frighten people. The debris and cyber guidance held here advises rather than binds. On a reading of these records, the most specific risk on paper is a forged pilot signal.
Scope. How a power satellite and its ground link could be attacked, taken over or misused, and what rules apply. It covers weapons in orbit, debris, cyber attack and dual use of the beam.
What is settled: the beam. Diffraction, the spreading of any wave sent from an antenna of finite size, sets how small the beam's spot on the ground can be. This site's own calculation applies it to the 1978 US design, a 1 km antenna at 2.45 GHz (gigahertz, billions of cycles a second). The spot is about 10.7 km across to its first null, the edge of the main beam (CLM-0889). At a given frequency, peak intensity can rise only with more power or a bigger antenna, not by focusing. This sets the spot size, not a safe level (CLM-0889). Two studies of the military implications, written for the US programme in 1978 and 1980, judged that the beam alone is not a real weapon. Both said it might still be used to frighten people (CLM-1039). The 1980 study estimated that twenty satellites aimed at one point in the same hemisphere as their receivers would give 460 mW/cm2 (milliwatts per square centimetre, equal to 4,600 W/m2), about 3.3 times the sunlight above the atmosphere. It showed no arithmetic, and it warned that many captured beams together could reach "incendiary levels" (CLM-1039).
What is settled: the rules. The guidance in these records advises rather than binds. The international debris guidelines are advice, but national licences may make them binding conditions (CLM-0756). US Space Policy Directive-5 of 2020 asks that operators consider protecting command links, based on their own risk assessment, and states that it creates no enforceable duty (CLM-0764). The civil standards for securing a spacecraft's command link (CCSDS) name no required encryption method, so conforming to them does not by itself show any level of security (CLM-0763).
What is disputed. The first dispute is how exposed a large satellite in geostationary orbit would be. That is the high orbit where a satellite stays over one point on Earth. On one side, a 2026 open-source assessment by Secure World Foundation judges it unlikely that China currently has a working ground-launched anti-satellite missile that can reach geostationary orbit. Only one test, in May 2013, is known to have targeted altitudes above low orbit (CLM-0741). The same assessment estimates several hours of warning after a detected launch, against 5 to 15 minutes in low orbit (CLM-0744). It also argues that a Chinese kinetic attack, one that physically strikes the satellite, would leave large amounts of long-lasting debris and be relatively easy to attribute (CLM-0745). The 1980 study found the satellite inherently no more vulnerable than other civilian infrastructure, but named the electromagnetic pulse of a nuclear burst as its real weakness. Both studies predate every anti-satellite test in these records (CLM-1040).
On the other side, the 1978 study found the satellite vulnerable to an enemy with good space capability. It said the transmitting antenna may be the weakest part (CLM-1040). A 2019 assessment co-written with US defence-affiliated researchers judges that a power satellite could become more vulnerable as space becomes more contested, and suggests many smaller satellites in place of a few large ones (CLM-0230). The 2026 assessment says the Chinese programme is likely still experimental, with too little evidence to say whether it will become operational soon (CLM-0741). No cited source gives a large structure's ability to move out of the way within hours (CLM-0744). The debris argument comes from a low-orbit test and is not extended to geostationary orbit (CLM-0745). A collision there, at speeds of the order of 500 m/s, would still leave a satellite little chance of surviving (CLM-0748).
The second dispute is the control link. The pilot signal is a low-power signal sent up from the ground, which the satellite's antenna array, its grid of many small transmitters, uses to aim the beam. A 2017 patent for the CASSIOPeiA space solar design contains no authentication wording, that is, no check that the pilot is genuine. Patents often omit security measures, so this says little, and nothing about the current design. The 1978 US design did include a coded pilot, as a 1981 NASA report records (CLM-0899). The reports that describe it do not say how the satellite would tell a forged pilot from a real one (CLM-1478). An analysis of the published design, made for these records, finds that losing the pilot or the reference signal defocuses the beam, so that it spreads out and its power falls away, which fails safe. Two failures do not. A forged pilot would make the array aim at the forger, and drift in the internal reference would weaken and mispoint the beam. The analysis is from the design on paper, not a test, and the fail-safe evidence comes from a different 1978 design (CLM-1478). The 1980 study listed safeguards such as pilot encryption and a dual-key shutoff (CLM-1039).
One real attack on a satellite network in these records came from the ground. Viasat, a satellite operator, says the February 2022 attack on its KA-SAT broadband network entered through a misconfigured VPN appliance (a remote-access device) on the ground, and it shipped nearly 30,000 replacement modems (CLM-0757). It found no evidence that the satellite itself was directly involved (CLM-0758). This is the operator's account, written while its inquiry was open, and it concerns a consumer broadband service (CLM-0757).
What is missing. No cited source gives a quantified threat assessment for a large geostationary power structure. The 2019 assessment is qualitative (CLM-0230). No cited source reports a test of these failure modes on CASSIOPeiA hardware (CLM-1478). No cited source assesses jamming, energy security or supply-chain security for a power satellite.
Read next. Threats to a GEO solar power satellite; Phase coherence across a flexible array; From direct current to beam; The beam; What a space power plant is worth to the grid; Running it, wearing out, and retiring it (for disposal at end of life).
Records
41 of the 49 records tagged to this layer are published. The rest are not yet published. A record is published only after a second reader has checked it and it has passed every publication check. What the gate checks.
| Record | In plain language | Tier | Confidence |
|---|---|---|---|
| CLM-0230 | A 2019 US study warns that large solar power satellites could be more at risk as more countries build anti-satellite weapons, and suggests using many small satellites instead of one big one to reduce that risk. | T3 | MED |
| CLM-0232 | Some of the radio frequencies being considered for beaming power down from space are the same frequencies already used by military airfield equipment worldwide, which makes it harder to pick a site. | T3 | HIGH |
| CLM-0741 | A 2026 independent assessment says China is unlikely to have a working weapon that could shoot down a satellite as high as geostationary orbit, because only one test, in May 2013, is known to have targeted altitudes above low orbit, and such a weapon would likely need several tests to be militarily useful. The assessment adds that this capability is likely still being tested or developed. | T4 | HIGH |
| CLM-0742 | A 2026 assessment lists thirteen Chinese anti-satellite tests since 2005; only one is recorded as going far above low orbit, to about 30,000 km in May 2013, short of the roughly 36,000 km of geostationary orbit given in the same assessment, and the assessment records it as probably just a rocket test with no known target. | T4 | HIGH |
| CLM-0743 | Two separate expert bodies looked at the same 2013 Chinese launch and disagreed about what it was: one calls it probably a rocket test, the other says it hints at a basic ability to attack satellites very far from Earth. Neither says a satellite that high has ever actually been shot down. | T4 | HIGH |
| CLM-0744 | The higher a satellite orbits, the longer the warning it gets: minutes in low orbit, but several hours at geostationary altitude, and the 2026 assessment doubts a missile could still track a target that had moved during those hours. | T4 | MED |
| CLM-0745 | The 2026 assessment says that if China physically destroyed a satellite, it would create long-lasting debris that would also endanger its own satellites, would be relatively easy to identify as the attacker, and would face international condemnation, so China would have to weigh these costs against less destructive options such as jamming or blinding. | T4 | MED |
| CLM-0746 | A 2025 assessment by a US think tank lists which countries hold which kinds of anti-satellite capability: China and Russia hold nearly all of them, while Iran and North Korea hold jamming, spoofing, cyber attack and attacks on ground infrastructure but not missiles that can reach satellites. North Korea is also listed as able to set off a nuclear explosion in space or the upper atmosphere. | T4 | HIGH |
| CLM-0747 | A 2025 assessment by a US think tank says a Russian satellite believed to collect signals has spent time near eight commercial communications satellites in geostationary orbit, and that some analysts think it may have come within about 5 km of one in July 2024 and less than 1 km of another in January 2025. | T4 | MED |
| CLM-0748 | The international radio standard says collisions in geostationary orbit happen at around 500 metres per second, far slower than in low orbit, so they are less catastrophic, but it also says a satellite has little chance of surviving any collision. | T3 | HIGH |
| CLM-0749 | A 2010 international radio recommendation said the smallest object that could then be tracked in geostationary orbit was just under a metre across, against 30 centimetres in low orbit, so smaller debris up there went unseen, and it said other break-ups at that altitude had quite likely gone unnoticed. | T3 | HIGH |
| CLM-0750 | The 2020 edition of the international debris guidelines says a retiring geostationary satellite should climb at least 235 km above its working orbit, plus an extra allowance that depends on how large and light it is, and the 235 km is made of 200 km of protected zone plus 35 km for the way gravity nudges the orbit back down. | T3 | HIGH |
| CLM-0751 | Working the international disposal formula through by hand shows that for a very large and very light satellite the extra height it must climb at retirement is dominated not by the fixed 235 km allowance but by the push of sunlight on it, and that for anything lighter than about 8.5 kilograms per square metre facing the sun that sunlight term is the larger of the two. | T0 | HIGH |
| CLM-0752 | ESA's 2026 report says that nothing naturally removes objects from geostationary orbit the way the atmosphere slowly clears low orbit, so a retiring satellite there has to be deliberately moved out and kept out for at least a century. | T3 | HIGH |
| CLM-0753 | ESA's 2026 report says the world is not clearing satellites out of orbit well enough, that even the usual 90 per cent success target would not reduce the amount of debris on its own, and that the figure needs to rise to nearly 100 per cent. | T3 | HIGH |
| CLM-0754 | ESA's 2026 report projects that collisions between existing pieces of debris will keep making more debris in low Earth orbit even if nothing else is ever launched, and it makes no such projection for geostationary orbit at all; nothing in the register says whether the same runaway effect would happen up there or not. | T3 | HIGH |
| CLM-0755 | NASA's July 2026 debris bulletin counts 33,098 tracked objects in orbit as of 4 May 2026, of which 18,189 are satellites and 14,909 are spent rocket stages and debris, but it gives no figures at all for geostationary orbit specifically. | T3 | HIGH |
| CLM-0756 | The international space debris guidelines are advice, not law: ESA's own report says they are a starting point for national rules and technical standards, and that it is up to each country, operator and manufacturer whether and how to apply them. | T3 | HIGH |
| CLM-0757 | In its own report on the February 2022 attack, Viasat says the attacker got in through a misconfigured VPN device, reached the network management system, and then sent the network's own ordinary management commands to tens of thousands of customer modems at once, wiping enough of their memory to knock them off the network without permanently breaking them. | T4 | HIGH |
| CLM-0758 | Viasat's own report, about five weeks after the attack, says it had found no evidence that the satellite itself or its main ground stations were directly involved, and that its analysis of affected modems found no impact on their physical or electronic parts and no evidence of tampering with modem software or firmware images, yet the service still went down across much of Europe. | T4 | HIGH |
| CLM-0759 | The UK's national cyber security agency said in May 2022 that Russia was almost certainly behind the attack on Viasat, that it started about an hour before the invasion of Ukraine began, and that wind farms in central Europe were among those affected. | T3 | HIGH |
| CLM-0762 | The civil space command-link security standard says exactly which parts of a message it protects and which it does not: it fully protects the command payload, only partly protects the message header, and leaves several control and status channels unprotected altogether. | T1 | HIGH |
| CLM-0763 | The civil space link security standards deliberately do not say which encryption to use, do not describe how to build a real system, and leave the choice of algorithm to the mission; a companion document adds the key management, rekeying and replay protection needed to run the protocol. | T1 | HIGH |
| CLM-0764 | A 2020 US presidential directive sets out what good cyber security for satellites looks like, including making sure a satellite only obeys commands from an authorised sender, but it says agencies and operators "should" do these things and states in its own text that it creates no legally enforceable obligation. | T1 | HIGH |
| CLM-0765 | Searching all fourteen security and defence sources for anything about insurance or liability turns up no premium, no cover limit and no liability figure for any satellite; the only real mention is one reference to the 1972 space liability treaty. | T0 | MED |
| CLM-0884 | The February 2022 KA-SAT attack was a break-in to a ground computer network. Nobody needed a rocket, and Viasat, which runs the network, said it found nothing showing the satellite itself was touched - which is what it could check, not a guarantee. On whether China can currently attack a satellite in high orbit with a missile, this register holds two outside views that differ: a research foundation judges it unlikely, and a US defence intelligence body reads a 2013 Chinese launch as suggesting the capability may exist. | T4 | MED |
| CLM-0889 | A beam sent from geostationary orbit cannot be focused to a small point. Diffraction sets a floor of kilometres across, and that floor is fixed once the antenna size and frequency are chosen. The only ways to make the beam more intense are a bigger antenna or more power, both of which are large and obvious. | T0 | HIGH |
| CLM-0899 | The patent never says the steering signal is protected in any way. There is a difference between scrambling that signal so nobody can read it and signing it so the satellite can tell a fake from the real thing. Only the second would stop someone aiming the beam. The patent has neither. | T3 | HIGH |
| CLM-1039 | The programme's own military studies did not treat the beam itself as a real weapon. The 1980 study said one captured beam would carry rather little power and has no known harm to living things, so at most it might be used to frighten people, as the 1978 study also suggested. Twenty satellites pointed at one spot would reach 3.3 times the sunlight figure the study used, which is about the strength of sunlight above the air; that is about 4.6 times standard sunlight at the ground. What the studies worried about was the lasers a satellite could power, and someone faking the pilot signal, for which they listed the countermeasures. | T3 | HIGH |
| CLM-1040 | The programme's first military study said the satellite could not be fully protected; its second said it was no softer a target than a power station or a refinery, that it could not be made to fall on anyone, and that the one real threat was the electromagnetic pulse of a nuclear detonation in space. | T3 | HIGH |
| CLM-1084 | When the programme asked the public in 1979-80, one respondent in seven wrote back; the environmental and consumer groups were mostly against, the students split, the space enthusiasts for, and most people had never heard of it. | T3 | HIGH |
| CLM-1113 | This paper does not describe a laser power satellite, though two NASA-funded designs for one exist from 1978-80. It does show how a laser receiver behaves: it needs light about a thousand times brighter than the microwave beam to work well, and even at the level the authors call eye-safe it is about a third as efficient as the microwave receiver. | T0 | MED |
| CLM-1183 | ESA's 2026 report counts about 960 catalogued objects in geostationary orbit at the end of 2025, about 800 of them satellites, but nearly 9,700 objects pass through the protected zone around it at some point in their orbits. A separate chart counts about 1,870 satellites near geostationary orbit in 2025, more than half of them drifting. | T3 | HIGH |
| CLM-1188 | Over the past ten years, almost every satellite retiring from the geostationary belt has tried to move itself out of the way as the debris rules ask, but only about seven to nine in ten have managed it, and the report notes two bad years, 2015 and 2022. | T3 | MED |
| CLM-1223 | In 2019 China's state news agency said the country's main space-technology academy planned a megawatt-scale solar power station in space by 2035, weighing about 200 tonnes, and was spending about 28 million US dollars on a ground test site in Chongqing. | T5 | MED |
| CLM-1227 | China's state broadcaster said in 2022 that China planned a 10 kW orbital test in 2028, 10 MW by 2035 and a 2 GW commercial plant by 2050. | T5 | MED |
| CLM-1243 | The company says it is lining up partners in Japan, the US, Saudi Arabia and Canada, and working with the European Space Agency. It pitches beamed power as something Britain could offer its allies, including for defence and polar sites. It says the ground receivers would be cheap enough to finance even in poorer countries. None of this is backed by named deals or figures in the post. | T6 | MED |
| CLM-1245 | In 2016 a senior Chinese military official told state media that China planned to use the space between the Earth and the Moon for solar power after its space station was built, but that a full-scale plant, over 10,000 tonnes, was beyond current technology. | T5 | MED |
| CLM-1246 | In 2019 a Chinese state science paper said China had started building a test base in Chongqing, planned high-altitude solar stations by 2025, a megawatt-scale test station in space from 2030 and the ability to build a gigawatt commercial plant by 2050. | T5 | MED |
| CLM-1267 | ESA also wanted to know whether beaming solar power from a satellite could power Mars and Moon bases or distant probes, because Europe had no nuclear space power sources of its own and building them would be costly. | T3 | HIGH |
| CLM-1478 | Almost everything that can go wrong with this design makes the beam fall apart and become harmless, which is the right way to fail. Two things do not. If someone fakes the ground signal, the satellite aims at the fake. And if the internal reference cabling drifts, the beam weakens and wanders rather than switching off. | T0 | MED |