Integrity and RAIM
Before WAAS was introduced you were required to do an integrity check before each flight when you planned to navigate with a GPS receiver certified under TSO C129 (supplemental navigation). Each such certified GPS receiver must be able to perform a check of the integrity of its position solution.
AC 90-100 stated that if equipment (without WAAS) is used for RNAV (area navigation), RAIM Availability must be confirmed for your route of flight. This was extended by AC 90-100A, which required (starting July 2009) that a pilot must check GPS RAIM availability for Area Navigation (RNAV) routes (Q and T routes), departures, and arrivals if RNAV compliance is based solely on TSO C129 equipment. These requirements led to the so-called RAIM system (Receiver Autonomous Integrity Measurement).
With a WAAS GPS receiver the picture changes significantly -- RAIM checks are no longer required unless you lose WAAS coverage. With WAAS, the receiver can now be used for primary navigation. But these receivers, certified under TSO C146, must still check for integrity of your GPS position solution, and this check is more sophisticated since there are new integrity requirements for approaches with vertical guidance that are more stringent than for LNAV approaches. First we'll describe what RAIM is, and then discuss Integrity in TSO C146 equipment.
So what is RAIM Availability, a RAIM check, or RAIM Prediction? RAIM Availability means your receiver (with the number and geometry of satellites in view) can validate its position calculations. A RAIM check determines whether you meet the Horizontal Protection Limit (HPL) requirements for your phase of flight, which are 2 nm Enroute, 1 nm Terminal, or 0.3 nm for non-precision approaches. These checks are being done continuously (autonomous measurements) to determine your integrity at all times (by comparing the pseudoranges to a number of satellites). If you fail to meet the requirements, messages must be given to the pilot (each receiver does this differently).
RAIM Prediction is based on the current almanac of the coarse orbit information of the entire satellite network. When you select an arrival time at a specific waypoint, as in the GNS 530W below, it advances the satellite positions along their orbits to determine which satellites are in view there (which ones and how many). From the geometry of their location and yours, a dilution of precision can be calculated and an estimate made of how good your solution will be. This prediction uses certain criteria to determine whether RAIM will be available or not. On the 530W it will do this calculation and give you an answer. That prediction is shown below, along with RAIM prediction from an FAA website (raimprediction.net). The FAA prediction is for a 24 hour window.
On the website, the "Summaries" pull down menu allows you to choose Enroute, Terminal, and Non-Precision Approach (NPA). Each of these has two options; with or without Baro-Aiding. The snapshot above is for a NPA without Baro-Aiding. Garmin 430/530/1000 receivers without WAAS have Baro-Aiding, which uses your altitude in place of the GPS altitude of the position solution. Then, the receiver only solves for the horizontal position and needs one less satellite, giving a higher likelihood of having the required integrity for the time frame of the graphical prediction.
The Raim Check
There are different algorithms for doing a RAIM check for different receiver manufacturers. One method uses the pseudorange measurements from 5 satellites in order to do Fault Detection (FD). Some use 6 satellites so they can do Fault Detection and Exclusion (FDE). It takes 4 satellites (4 pseudoranges) to determine your position, one more to identify a faulty solution (FD), and two more (for FDE) to exclude it in your RAIM check. The check determines your HPL, which is your position circle with a 99.99999% certainty. Your position will be outside a circle of that radius around the computed position only 1 sec out of 10 million sec, statistically speaking. As mentioned above, this HPL is compared to the one required for your current flight phase. A GNS 430/530 receiver will give you a yellow INTEG message if the solution is out of bounds, and a red WARN if the receiver is not getting any signal usable for navigation. In either case you need to revert to your primary navigation system based on VOR receivers.
Integrity for WAAS GPS Receivers
If have a TSO C146 WAAS receiver there is no requirement for a RAIM check unless you have a WAAS failure or are out of the coverage area. Then, you must meet the above requirements. For that reason there must be a way to check RAIM in these receivers. But if you're getting WAAS corrections, your HPL is still being checked. And since you can do vertical approaches with the receiver it also checks your VPL (Vertical Protection Limit).
Further, there are added criteria for approaches as listed in the Table here. Listed there are the alert limits (HAL/VAL) for each type of approach. An alert is given when HPL > HAL, or VPL > VAL.
Several kinds of alerts are now possible from a WAAS GPS receiver. If you are in the Enroute or Terminal phase there is a requirement on HPL (2 nm and 1 nm respectively). For an LNAV approach it is 0.3 nm. If your current value of HPL rises above the requirement for your current phase you'll get a Loss of Integrity message. If you are on a vertical approach the above table gives the required values for HPL and VPL, and if you fail to meet one of them on your particular approach, but still have the basic 0.3 nm for a non precision approach, you can downgrade to an LNAV approach. These messages from a G480 are shown below.