macpacheco

The primary issue with GNSS (GPS, Galileo, GLONASS, ...) is that everything is probabilities.Any accuracy level is accompanied by a probability.Even if GPS+WAAS delivers a 2 meter accuracy with 95% probability and 10 meter accuracy with 99% probability (not actual numbers, just for arguments sake), there is still that 0.01% chance of it being 20 meters off. 20 meters off laterally and you could be outside of a typical GA airport runway. All those probabilities are based on theoretical worst case scenarios, for those 99% exceptions (what might happen in the worse 1% situations).That's why there's no autoland even with a full LPV 200 approach available.Now, that doesn't mean I agree with FAA's judgement. I'm just stating what I understand is their thought process behind those regulatory decisions. Millions of LPV 200 approaches have been flown, and I've never heard of such problems. If the FAA applied the mathematical requirements they apply to WAAS to ILS, those pessimists might find a way to forbid even ILS CAT I approaches. But since ILS has been used for 60+ years, it's a trusted system. The people that write aviation regulations think much more like lawyers than real world people. Any really tiny concern and something gets prohibited.The main issue is if you're 10+ miles away from a reference station, the reference station doesn't know of any possible GPS spoofing. Jamming isn't a huge issue, since the signal would simply go out (or become corrupted), raising the VPL and HPL (protection levels) above the requirements, generating an alert and aircraft would go around if in IMC, but spoofing is ultra serious stuff, since it would give erroneous information (spoofing is sending an apparently valid GPS signal, but with the wrong data in it).Add to that the fact that today the highest error factor with GNSS augmentations is iono (ionospheric = very high atmosphere) error.With multiple GPS signals on multiple frequencies, iono errors can be calculated with accuracy ten times better than current WAAS supplied iono parameters, since iono errors change from place to place.Those two reasons are why additional civilian signals (each on a separate frequency) is so important for GNSS. WAAS reference stations today use a 20 year old method to use one civilian signal + 2 military signals to very accurately calculate GPS corrections. But the FAA considers those features (semi codeless) taboo for usage in WAAS aircraft receivers. The main reason is the FAA looks down on the L2 GPS band. GPS military signals use L1 and L2 bands. L1 is a civil aviation protected band, but L2 is not. L2 is used by some radars for instance. This never stopped WAAS/EGNOS/MSAS and other aviation oriented GNSS augmentations from being implemented (fully depending on the L2 military signal), but never mind using that on board civilian aircraft.Then there are the new GPS civilian signals, L2C (L2 Civilian) and L5 (there's no military L5 signal), L1C is still years ahead and doesn't help aviation much, since it runs on the same L1 frequency that we already have a useable civilian signal on.A lot of GPS satellites have the L2C (L2 Civilian signal) transmitting today. That's 8 satellites so far. But the US Air Force doesn't have the capability of monitoring that signal on its ground stations, and their requirement that a signal needs to have at least 18 satellites with a given GPS capability before it can be considered useable, and L2C is right now only available for testing (GPS ranging can be calculated, but its transmitted in a format that is forbidden from any safety of life scenario like aviation).The next thing down the line is the L5 signal. L5 is located in an aviation protected band. This band is used by DME for instance. But L5 signals today are only available from one GPS satellite plus some augmentation satellites (all WAAS satellites, QZSS transmit it for sure, probably all aviation augmentation satellites transmit it ). L5 is a high grade, purely civilian GPS signal, that is tuned for high quality (with its 10MHz bandwidth versus 1MHz bandwidth for L1 and L2 civilian signals).With dual (L1 + L5 or L1 + L2C), and triple (L1 + L5 + L2C) frequency receivers, your on board GNSS receiver could perform a lot of extra validation that it can't do today. For any spoofing to go undetected, all frequencies in question would have to be spoofed simultaneously. That's why I think that L2C should be an integral part of the future WAAS. By requiring that all available GPS signals are received by all future receivers, and requiring that they validate all available frequencies, GNSS+augmentation becomes a ultra accurate system, with an extremely high extra integrity. Besides, no spoofing has been documented so far, outside of intentional/planned testing situations.Ultra accuracy GPS systems that don't suffer from FAA induced regulatory limitations are showing inch level accuracy today, even without using L2C and L5 (but using L1 civilian + L1/L2 military signals). All we need for autoland is 4 meter lateral accuracy and 8 meter vertical accuracy (also sufficient for CAT II landing).With triple frequency receivers, a couple inch worst case accuracy can be achieved, and conditions that prevent such incredible accuracy can be detected.Add to that the brand new introduction of chip scale atomic clocks (with a little bit of cost reduction would be affordable for usage in aviation certified GNSS receivers), and future GPS receivers can detect spoofing 100% of the time, knowing the local time down to the nanosecond allows a GNSS receiver to validate each individual GPS signal received with ultra high precision, any spoofing would cause very noticeable discrepancies between local atomic clock time and calculated time from received GPS signals. Hopefully the FAA will create a new regulatory opening for triple frequency receivers with chip scale atomic clocks that would allow for CAT III autoland operations.Marcelo PachecoPP-IFR + computer wiz + GPS enthusiast + always a huge critic of government induced inefficiencies