Summing Signal Power from Multiple Modes
How does VOACAP handle the calculation of the signal power as the signal can arrive in multiple modes?
George Lane: It helps to see what is happening when looking at all modes output (Method 25) of VOACAP or ICEPAC. VOACAP is attempting to compute the hourly median signal power values for the 30 days in the month at that hour and SSN. It does this for up to 21 different possible modes [ e.g. 2F2, 3F2, 4F2, 2F1, 3F1, 4F1, 4E, 5E and 6E for the high ray and the low ray modes plus the 4Es, 5Es and 6Es for a total of 21 ]. For each of the possible modes, the program computes the monthly median and the power levels which should be exceeded on 27 days of the month and on 3 days per month (lower and upper decile). Again remember, these are predictions of the hourly medians at a given hour over the days of the month and not what is happening during a single hour of one day.
It is assumed that the presence of multiple modes will increase the available power over the days of the month compared to what any individual mode might contribute by itself. The calculation process is very simple. The signal power in dBW is converted to Watts and the values are added to obtain the combined signal power. The sum is then computed as dBW. This process is done for the median signal power and the upper and lower decile signal powers to give the monthly distribution. Usually, the contribution of secondary modes to the monthly distribution of hourly median signal powers is not much but may be 4 to 6 dB. This happens usually when the antenna patterns support a secondary mode better than the primary mode (a bad design situation but very common).
In practical terms, the program is attempting to cover what really happens. I have had the opportunity to be making signal power measurements on paths where there was an oblique ionospheric sounder. Some days the 1F2 mode would be stronger but on other days when the 2F2 MUF would rise sufficiently, that mode would be stronger. When we look at the distribution of the signal powers at that hour over the 30 days of the month, the 2F2 mode occasionally provided a stronger signal than the 1F2. This means that 2F2 mode improved the performance of the circuit than if we only had the 1F2 mode. So the presence of other modes which become stronger than the most reliable mode on some days, will enhance the signal power available than if we had only the single mode by itself.
I think people get confused about the prediction process. They think that the mode contributions are being added for a single day in the month which doesn't make much sense. We all know that multiple modes at a given instant can cause harmful multipath which does NOT improve the circuit performance. What the program is doing is saying that on the days when the primary mode is weak, is there another mode which may come into play? If so, then the monthly distribution of signal powers is higher than if no other mode is available.
Another problem is that many people think in terms of field strength measurements where the antenna for field strength meter combines all of the incoming signals into one field strength reading. VOACAP is based on signal power measurements at the receiver and not field strength measurement. VOACAP computes the power of each arriving ray path using the appropriate antenna patterns for that takeoff or arrival angle. The summing process of the individual mode contributions to obtain a combined power distribution for an hour over the days of the month is forced to fit actual measurements by applying the Transmission Loss Table (Table 7) in the IONCAP theory manual.
There is no theoretical justification for adding the signal powers of the various modes to obtain the combined signal power distribution. But there is good agreement with measurement which is what this program is all about. If you look at Method 25 output, you will see that the contribution of the other modes is not very much over that of the most reliable mode by itself. But these are the small adjustments in the prediction process that has made the IONCAP family of programs better than the older or simpler models.
By the way, the IONCAP family of programs does have a multipath probability calculation which is supposed to give an estimate of the probability that the presence of other modes will cause serious multipath conditions. Sadly, this calculation is in error. However, funding ran out before I could get the corrections into VOACAP.