All
← Back to Squawk list
A NEW Trace! The FULL MH370 Story...So Far.
Petter from Mentour Pilot gives an excellent summary of MH370 so far and calls for a renewed search. (www.youtube.com) More...Sort type: [Top] [Newest]
WSPR is a low-power protocol, usually 5-watts of RF power output, used to test propagation on many, mostly HF (shortwave) frequencies. The transmitters key for 1-minute 50-seconds, normally every ten minutes although some operators key randomly. The signal is not strong to begin with, antennas are often omnidirectional, and scattering from metal surfaces weakens the signal further. Decoding a rapidly-moving airplane over the time period, even with interleaving of ID's within the protocol, is highly unlikely at best and impossible in practicality. Furthermore, there are simply not enough transmitters and receivers in that area of the world to capture enough useful date. What do I know? Very little actually; however, until recent retirement, I worked full-time in the wireless world as an RF (Wireless) engineer for over 50-years and been a ham radio operator since 1969. I have 13 WSPR beacons and 17 WSPR receivers. My call is NI5F. Look me up on wsprnet.org
Great info Bill, thanks. Presumably, they can test the accuracy of this technique any time and many times by trying it on existing aircraft flying through the Indian Ocean and compare the actual track with their WSPR-computed track.
Testing at 1-GHz using ADSB will act very differently as compared with high frequencies used by WSPR; 1-GHz has very short wavelengths and reflects differently than the much longer wavelength WSPR frequencies. Note that intercept radar use very short wavelengths (GHz frequencies) most of the time today for a reason, that is resolution. Also, for those mathematically inclined, f(doppler) = 2 * Velocity(aircraft)* COS(theta - angle of intercept) * [frequency(transmitter)/speed of light] in Hertz. Therefore, you can see doppler is directly proportional to frequency.
Furthermore, the WSPR dataset includes only the result over the 1:50-second interval. You cannot parse out the microsecond-to-microsecond frequency shifts, that is , doppler, without that data. Also, many WSPR operators are not time-locked to Stratum 1 clocks; the result is slightly inaccurate timing and frequency center. My WSPR transmitters are all GPS-3D-Fixed (slaved) using the 1PPS pulse to ensure timing and frequency; about half of my receivers, in contrast, are GPS-fixed. The non-GPS-locked receivers use NTP to keep the Raspberry Pi clocks fairly accurate. So my transmit WSPR data is nano-second accurate but my receivers may not be.
This may be more than you want to know. I am just illustrating the facts behind my statement that the WSPR database should not be used, or cannot be used, to determine anything but propagation paths and relative signal strengths.
Furthermore, the WSPR dataset includes only the result over the 1:50-second interval. You cannot parse out the microsecond-to-microsecond frequency shifts, that is , doppler, without that data. Also, many WSPR operators are not time-locked to Stratum 1 clocks; the result is slightly inaccurate timing and frequency center. My WSPR transmitters are all GPS-3D-Fixed (slaved) using the 1PPS pulse to ensure timing and frequency; about half of my receivers, in contrast, are GPS-fixed. The non-GPS-locked receivers use NTP to keep the Raspberry Pi clocks fairly accurate. So my transmit WSPR data is nano-second accurate but my receivers may not be.
This may be more than you want to know. I am just illustrating the facts behind my statement that the WSPR database should not be used, or cannot be used, to determine anything but propagation paths and relative signal strengths.
I can partially test it now with aircraft flying overhead here. I have 100,000's of daily hits to my fixed ADSB receiver with constant daytime air traffic flying from Atlanta to the Florida beaches, day and night east-west traffic to south Florida, and the consistent overflights of rotary-wing students from Fort Rucker (yeah, I know, it has a new girlly name). The biggest problem with aircraft is the Doppler shift caused by the motion of the aircraft; and the faster it goes, the more the frequency shift outside the very narrow WSPR bandwidth window ( a few Hertz wide ).
I can't speak to WSPR, but when I ran an ADS-B station from my Toronto balcony that contributed to FlightAware (I was in the top 15 at one point), I would sometimes get messages from well over the horizon in New Jersey up to 700 km away. My best guess for the source of propagation is high altitude trans-Atlantic airplanes reflecting the messages. My normal range of reception was about 430 km (being about 70 m up helped), so I could pick up traffic over Harrisburg or Albany but not Philadelphia or NYC.
I had a condo building go up at one point, blocking part of my eastern horizon, but it also reflected a fair amount of message from the east, presumably off the plentiful glass. Not traffic landing at Pearson, which would be blocked by the downtown towers, but higher flying aircraft. I learned a lot about microwaves running that station. I should really get a ham licence. I'd like to get into WSPR.
I had a condo building go up at one point, blocking part of my eastern horizon, but it also reflected a fair amount of message from the east, presumably off the plentiful glass. Not traffic landing at Pearson, which would be blocked by the downtown towers, but higher flying aircraft. I learned a lot about microwaves running that station. I should really get a ham licence. I'd like to get into WSPR.
Mentor Pilot is excellent! There is new RF science at work here and worth a new search effort. Jump to 42:00 min into video to understand the new data. The entire video is best summary yet.