Receipts / S10

Reel cover: Your Wi-Fi router is bouncing radio off every wall in your house.
S10 · · posted 19 Oct 2026

Your Wi-Fi router is bouncing radio off every wall in your house.

    Line by line

    Every claim, with its source.

    Every spoken or on-screen factual claim, with the wording it rests on. Sources fetched 2026-10-11. The 1978 Nature paper is paywalled (abstract, authors and dates only). CSIROpedia was read from a Wayback capture; the Federal Circuit opinion from the Wayback copy of the court's PDF. Unit conversions marked "(our conversion)" are ours.

    ShotClaim as spoken or shownSource and verbatim support
    hook"Your Wi-Fi router is bouncing radio off every wall in your house."US 5,487,069: "Reflections from the walls (and floor and ceiling) of the room 1, items of furniture 2, and the like ... cause the multiple path transmissions". PM's Prize 2009: "In the confines of buildings and rooms, radio waves bounce off many surfaces, so that a transmission arrives at a receiver followed by a series of echoes."
    hook"That should scramble everything."US 5,487,069: each symbol must last "substantially longer than the delay time in order that the received echoes of a first symbol not mask the receipt of a subsequent symbol. This requirement has provided a severe upper limit to the rate at which data can be transmitted". PM's Prize: "a fuzzy, ambiguous signal, akin to 'ghosting' on a television".
    hook"Part of the reason it doesn't goes back to astronomers hunting exploding black holes."PM's Prize: "the search for radio waves from exploding black holes—predicted in 1974 by Stephen Hawking"; the "techniques he and his collaborators developed to clean up intergalactic radio wave distortion eventually found expression as the technology in the wireless LAN". O'Sullivan (EPO): "the problems we had to solve in radio astronomy back then with the black holes and later with the WLAN were remarkably similar, and FFT was a key component". "Part of the reason": the patent is one technique in 802.11a/g/n, not all of Wi-Fi. The link is O'Sullivan's own telling; hedged at close.
    answer"They never found one."PM's Prize: "John didn't find them". Nature 1978 title: "Limits on cosmic radio bursts with microsecond time scales". Lazio 2008: searches "have not yet found a convincing signal (O'Sullivan et al. 1978, Phinney & Taylor 1979)".
    answer"The maths they used to look became one of the tricks that made Wi-Fi fast"O'Sullivan (EPO): "FFT was a key component". Patent: "the preferred method of generating and demodulating multi-carrier modulation schemes uses a device capable of performing Fast Fourier Transforms (FFTs) and Inverse Fast Fourier Transforms (IFFTs)". E.D. Tex. 2007: 802.11a and 802.11g "embod[y]" CSIRO's invention. "One of the tricks", not "invented Wi-Fi": O'Sullivan (EPO): "we did not invent wireless networking per se".
    answer"and earned Australia hundreds of millions of dollars."CSIROpedia: "total revenue earned from this technology in excess of $430 million" (Australian dollars; CSIRO is the Australian government's science agency).
    answer (screen)"0 FOUND"As "They never found one."
    answer (screen)"A$430M+ (CSIRO, BY 2012)"CSIROpedia: "By 2012 CSIRO had licence agreements with 23 companies representing around 90% of the industry, with total revenue earned from this technology in excess of $430 million".
    hunt"The nineteen seventies. Stephen Hawking had predicted that tiny black holes should slowly evaporate, and finally explode."PM's Prize: "exploding black holes—predicted in 1974 by Stephen Hawking". Nature 1978 abstract: "black holes of mass ≤10^15 g evaporate in ∼10^10 yr, ultimately annihilating into a burst of energetic photons and particles." ("Tiny": 10^15 g = 10^12 kg, about a billion tonnes / 1.1 billion US tons, our conversion; tiny in size, not mass.) A theoretical prediction; none has been observed.
    hunt"One idea: the blast might send out a blip of radio, a millionth of a second long."Nature 1978 abstract: "a radio pulse with a characteristic frequency of ∼3 GHz and an energy of ∼10^32 erg may also be generated"; title: "microsecond time scales". The radio-pulse idea is Rees 1977, not Hawking (dossier), hence "One idea".
    hunt (screen)"radio pulse 1 µs wide"Nature 1978 title: "Limits on cosmic radio bursts with microsecond time scales".
    john"An Australian engineer, John O'Sullivan, was working at the Dutch radio observatory at Westerbork."PM's Prize: O'Sullivan "set off in 1974 to work for a year on the Westerbork radio telescope in The Netherlands. Nine years and two children later ... he became head of the group responsible for the development and maintenance of the Westerbork's receiver systems"; "1974–1983 Head of Receiver Group, Netherlands Foundation for Radio Astronomy". Which telescope the 1978 search used is unverified (paywalled), so the line says where he worked, not where the search ran.
    john"He and two colleagues went looking for those blips."Nature 276, 590–591 (1978): J. D. O'Sullivan, R. D. Ekers & P. A. Shaver.
    john"They recorded hundreds of metres of film."O'Sullivan (EPO): "We used complex technologies, finetuned radio telescopes, improved spectrometers and recorded hundred of meter of film". (Hundreds of metres = hundreds of yards, our conversion.)
    john"In his words, they did not find the smoking gun."O'Sullivan (EPO): "but although we looked at the characteristic traits of the signal, we did not find the smoking gun".
    john (screen)"Westerbork dishes (25 m / 82 ft each)"NWO-I: "fourteen 25-metre steerable dish antennas"; Apertif paper: "14 identical steerable 25-m dishes". 25 m = 82 ft (our conversion).
    john (screen)"NO BURSTS · NATURE, 1978"Nature 276, 590–591, published 7 Dec 1978; title "Limits on cosmic radio bursts ..." (an upper limit, i.e. no detection).
    fourier"What he kept was the method."EPO: "the problems ... were remarkably similar, and FFT was a key component". PM's Prize: techniques "developed to clean up intergalactic radio wave distortion eventually found expression as the technology in the wireless LAN".
    fourier"A signal from space arrives smeared."PM's Prize: "intergalactic radio wave distortion".
    fourier"A piece of maths called a Fourier transform pulls it apart into its frequencies, so you can clean it up."Standard definition of a Fourier transform (not quoted). PM's Prize: "clean up intergalactic radio wave distortion".
    fourier"He even had it built into a chip."PM's Prize: "To simplify and speed up the transform task he worked with Austek Microsystems to create a computer chip". AU 610,934 "A transform processing circuit", inventors incl. "John David O'sullivan", priority 21 Aug 1987.
    fourier (screen)"160,000 TRANSISTORS (1988)"PM's Prize: "This VLSI chip consisted of 160,000 transistors and performed real time transforms at rates up to 2.5 Msamples/s". EPO: "In 1988 the team had built ... a first chip for FFT".
    goal"Then, back in Australia at the science agency CSIRO, a new target."EPO: "In 1983, Dr. O´Sullivan returned to Australia to head the Signal Processing Group at CSIRO Radiophysics".
    goal"Around nineteen ninety: cut the office network cable, and still move a hundred megabits a second."PM's Prize: "By 1990 CSIRO was looking for ways to commercialise its capability in radio physics". O'Sullivan (EPO): "At that time, wireless LANs already existed ... we would need to reach transmission rates of 100 mbit per second". Patent title: "Wireless LAN".
    goal (screen)Ethernet cable snipped; target chip "100 Mbit/s"As above (EPO).
    echo"Indoors, radio bounces off walls, desks, even someone opening a briefcase."Patent: "Reflections from the walls (and floor and ceiling) of the room 1, items of furniture 2"; the channel changes from "the simple act of someone opening a briefcase positioned on a desk".
    echo"Your receiver hears the signal, then a crowd of late echoes."Patent: the receiver gets "an undelayed signal 5 ... and a number of delayed signals 6" (Fig. 2).
    echo"Send bits fast, and the echoes of one bit land on top of the next."Patent: "in order that the received echoes of a first symbol not mask the receipt of a subsequent symbol". Fed. Cir. 2008: "The "echo effect" caused by the bouncing signals".
    echo (screen)"BRIEFCASE" labelPatent: "someone opening a briefcase positioned on a desk".
    rival"Motorola's answer was hardware: antennas with thirty-six possible paths."Patent background on Motorola's ALTAIR: "a directional antenna system with 6 beams for each antenna resulting in 36 possible transmission paths".
    rival"The Australians' answer was maths."Patent claim 1 (sub-channel modulation) and preferred embodiment (FFT/IFFT).
    rival (screen)"two antennas × 6 beams = 36 PATHS (MOTOROLA ALTAIR)"Patent background: "a directional antenna system with 6 beams for each antenna resulting in 36 possible transmission paths".
    trick"Don't send one fast stream. Split it into many slow ones, side by side, each on its own frequency."Patent claim 1: modulate the data "into a plurality of sub-channels". Fed. Cir.: "transmit different portions of a series of signals containing the data ... over a number of different frequency channels".
    trick"Each one is slow enough that the echoes die away before its next symbol arrives."Patent claim 1: "such that the period of a sub-channel symbol is longer than a predetermined period representative of the time delay of significant ones of non-direct transmission paths". Fed. Cir.: "sufficient temporal separation between each signal transmitted on each frequency to avoid intersymbol interference".
    trick"And the Fourier transform does the splitting and joining in one go."Patent: "the preferred method of generating and demodulating multi-carrier modulation schemes uses a device capable of performing Fast Fourier Transforms (FFTs) and Inverse Fast Fourier Transforms (IFFTs)"; example "uses a 16 point complex IFFT". (Preferred embodiment, not claim 1.)
    trick (screen)"FFT box at both ends"Patent Fig. 7: "Inverse Fast Fourier Transform 47" (transmit); FFT for demodulation per the quote above.
    honest"To be fair, they didn't invent that idea. Splitting data into many carriers was already known, and their patent cites the earlier work."Patent cites Cimini 1985 "Analysis and Simulation of a Digital Mobile Channel Using Orthogonal Frequency Division Multiplexing" and Casas 1991 "OFDM for Data Communication over Mobile Radio FM Channels". Fed. Cir.: prior art incl. Rault 1989 COFDM and Bingham 1990 "Multicarrier Modulation ...: An Idea Whose Time Has Come".
    honest"What they patented was a way to make it work for fast wireless networks indoors."CSIROpedia (court's reading): "a combination of three key techniques: parallel sub-channels (ensemble modulation) ..., data reliability enhancement through Forward Error Correction (FEC), and data reliability enhancement through bit interleaving". Fed. Cir.: Rault did "not disclose an indoor environment".
    honest (screen)"1985, 1991 · PRIOR ART CITED IN THE PATENT"Cimini 1985; Casas 1991, both cited on the patent.
    honest (screen)"US 5,487,069"Patent number, "Wireless LAN", assignee CSIRO.
    standard"The patent was filed in nineteen ninety-two."Patent face: "Nov. 27, 1992 [AU] Australia ... PL6069" (Australian priority filing). PM's Prize: "Within a year, in 1992, CSIRO applied for an Australian patent". (US application filed 23 Nov 1993.)
    standard"The standards body built the method into Wi-Fi: eight-oh-two eleven a, then g, then n."E.D. Tex. 2007: "In 1999, the IEEE ratified the 802.11a standard, which embodies the core technology invention by CSIRO ... In 2003, the IEEE ratified the 802.11g standard, which also embodies CSIRO's invention." Science in Public 2009: "802.11a, 802.11g, and the new 802.11n standard"; CSIROpedia: "IEEE 802.11 draft n". (802.11b "differs from CSIRO's invention", so not listed.)
    standard"And almost nobody paid. Twenty-eight companies said no."E.D. Tex.: "In 2003, CSIRO contacted companies who practiced the '069 patent ... None of the potential licensees accepted". EPO: "Out of 28 approached firms, none agreed to license."
    standard (screen)"1992 FILED · 1996 US GRANT"Patent: priority "Nov. 27, 1992"; "Date of Patent: Jan. 23, 1996".
    standard (screen)"28 COMPANIES · 0 LICENSES"EPO: "Out of 28 approached firms, none agreed to license" (2004).
    court"So CSIRO went to court in Texas, won the first round,"E.D. Tex. (Tyler), 15 June 2007: summary judgment of infringement against Buffalo, and "the Court GRANTS CSIRO's Motion for Entry of Permanent Injunction". "The first round" because Fed. Cir. 2008 "AFFIRMED IN PART, VACATED IN PART, and REMANDED" (obviousness sent back for trial); the companies settled before a final judgment.
    court"and the companies settled in waves."EPO (2009): "around AUD$ 200 million"; Patentology: "HP was the first to cave, and the other 13 companies soon followed suit"; April 2012: AT&T, Verizon, T-Mobile "around $220 million" (Patentology); minister: "CSIRO will receive more than $220 million from this round" (Beef Central).
    court"By twenty twelve, CSIRO says it had collected more than four hundred and thirty million Australian dollars."CSIROpedia: "total revenue earned from this technology in excess of $430 million". Hedged ("CSIRO says") because it is CSIRO's own figure and others differ: Patentology "over $420 million", EPO "some AUD 450 million".
    court"The patent ran out in twenty thirteen."Google Patents: "2013-11-23 Anticipated expiration", "Expired - Lifetime"; EPO: "By 2013, the patent in question will expire". (US patent.)
    court (screen)"A$430M+ (CSIRO)"; stamp "EXPIRED 2013"As above.
    close"So, the way O'Sullivan tells it, the hunt that failed, found something else."O'Sullivan (EPO): "remarkably similar, and FFT was a key component". Hedged because the link is his and the PM's Prize telling: EPO footnote "Some people argue ... there is only an indirect connection"; the bridge is general Fourier/FFT know-how, 14 years separate the 1978 paper and the 1992 patent, and the Wi-Fi patent had five inventors.
    close"Next time your Wi-Fi works through three walls, thank a black hole that never showed up."Rhetorical. The patent technique handles echoes (multipath), not wall penetration; "three walls" is a scene-setter, not a technical claim.

    What the drawing simplifies

    • No openly licensed photo of John O'Sullivan or the CSIRO team exists; all people are line-art figures or typography.
    • The Westerbork photo (Bert Verhoeff / Anefo, Nationaal Archief, CC0) shows the array at its opening on 18 June 1970 (12 dishes then; 14 later). It is the place and era only: not O'Sullivan, not the 1978 search, and which telescope that search used is unverified.
    • Black hole, radio pulse, film reel, Fourier bars, the chip die, the office, the room-echo diagram, the antenna-array vs equation panel and the lanes animation are our own illustrative animation. The chip die is generic, not the Austek chip.
    • Patent sheets 1, 2 and 6 of US 5,487,069 (public domain) are the real patent drawings: Fig. 1 room and paths, Fig. 2 direct + delayed signals, Fig. 3 60–61 GHz fades, Fig. 7 transmit chain with IFFT 47.
    • Court scenes (gavel, settlement waves) are typographic/illustrative; no court document images are used.
    • "PRIOR ART" stack is drawn titles, not reproductions of the Cimini or Casas papers.

    Checked, and cut

    • "CSIRO invented Wi-Fi": not said. O'Sullivan: "we did not invent wireless networking per se"; 802.11b "differs from CSIRO's invention".
    • "CSIRO invented OFDM": not said; the patent cites 1985 and 1991 OFDM work and the script says so.
    • "They found exploding black holes": false; none found.
    • "Hawking predicted the radio blips": not said; the radio-pulse idea is Rees 1977 ("One idea").
    • "The black-hole search led directly to Wi-Fi": hedged as O'Sullivan's telling; no claim that a black-hole gadget became a Wi-Fi part.
    • "Patent filed 1987": not used (that is the FFT-chip patent AU 610,934).
    • "Australia is paid for every Wi-Fi device": not said; patent expired 2013, never covered 802.11b.
    • "A$1 billion+": not used. The union's "nearly $500 million" and the PM Prize's "nearly a billion people use his invention every day": not used.
    • Settlement amounts per wave (A$200m / US$205m currency conflict): not put on screen.
    • The 1977 adaptive-optics Fourier paper is not merged with the 1978 black-hole paper.

    Sources

      Found an error?

      Tell us at hello@workingshown.xyz or in the comments. We check it against the sources, pin the correction, fix the caption, and log it on the corrections page.