My wanderings around the Ham community as well as HF broadcasting, RADAR and communications have suggested to me that maybe at least a few Hams, engineers, programmers and just plain users of the ionospheric prediction program IONCAP might have little knowledge of how it goes about the business of predicting the probability of successful transmission using HF frequencies (3-30 MHz). Many prediction programs are on the market now as the Ham community becomes more sophisticated and relies upon these programs to guide their DX schedules and/or their schedules with friends. All the available programs seem to have some usefulness and will not be commented on in this little write-up. We are here to talk "IONCAP". The mathematical methods and ionospheric data from which IONCAP was built will be discussed in follow up papers if anyone finds this stuff interesting. You will, no doubt, find the paper a little redundant as some things are said more than once in different contexts. This is intentional to follow what an ex-boss told me. He said, "When writing technical stuff it's not good enough to write so someone can understand it -- you have to write so it cannot be misunderstood." There may be a big misunderstanding about IONCAP and the way it goes about its business. A prominent propagation expert from a defense think tank--not to be named-- went on a complete tirade during a talk he was giving at an ionospheric symposium recently. The reason he gave at some length was that IONCAP was a farce because there is no way the probability of success of a reliable signal (Reliability) at the receiver could exceed the "Fraction of days within the month". In other words, if the Fraction of days within the month at an operating frequency of 20 MHz is 20%, how can the Reliability (percentage of the days within the month you exceed some signal or signal-to-noise ratio) be 65% at 20 MHz. This paper will hopefully convince you that it is not only possible, but it is the point IONCAP makes and is just life as it is. I sure hope you'll read on as there are a few things I would like you to blank from your mind for this discussion. They are those things it took you all those years to learn, such as, FOT, MUF, HPF, LUF and most of all Fraction of Days. This term Fraction of days (F-days) was used in computer predictions to define how many days within the month some specified frequency would be below the "daily MUF" due to the distribution about the monthly median MUF or what is called the "classical MUF". Also forget the concept of complete blanketing by lower layers or complete loss of signal due to penetration by the signal of any or all of the reflecting layers (E, Es, F1, F2).
Now that we have scrapped a few things from our vocabulary, for the moment, let's start with a very simple idea I had in about 1970 -- yes, that's how old IONCAP is. By this time the prediction programs were becoming very large and would only run on large main frame computers -- so my thoughts were to build a simple computer program using available ionospheric data to predict received signal levels anywhere on earth from any transmitter on earth. The thought was to build a curve as a function of frequency and received signal-to-noise ratio for specific equipment parameters in use i.e., transmitted power, antennas, etc. For now let's forget how the curves are built, and from what data and concentrate on the curves themselves. The ionosphere is much simpler at night so let's start there. The night-time curve (Figure 1) shows the signal rising gradually and then falling rapidly as the frequency is increased. Please concentrate on the middle curve for now.
You ask, "What year, transmitter power, antenna gains and from where to where is this curve to be used?" For now, that's not important, the thing is we can generate it and it shows the received signal-to-noise ratio (S/N) as a function of operating frequency. Of course, as you suspected it will differ for all the questions you asked, but the shape will stay the same whether the signals are good enough for your special case or too weak to be detected by your specific system. All operating HF frequencies between 3-30 MHz are included for the percentage of time the curve was generated. Let's call the one we just looked at the "best guess" or the 50% probability curve (Reliability). The 50% curve of Figure 1 separates S/N of the highest 50% of days from the S/N for the lowest 50% of the days within the month, so it will be exceeded 50% of the time. Similarly, the 10% curve separates S/N of the highest 10% of the days from the lowest 90%. Obviously, the highest 10% is higher than the highest 50%. We would like high S/N ratios a high percentage of the time, but the ionosphere is just not that cooperative and I suppose life isn't either. At least, IONCAP has a family of curves that represent a statistical distribution of sorts, be it normal, log-normal, skewed or whatever; we have by proper evaluation a S/N ratio expected for any percentage of time within the month for any HF frequency we wish to operate by interpolation between curves. That's the philosophy of IONCAP and that's what it does -- pure and simple. The computer program is, of course, not simple and is very large including much ionospheric and received signal level data. As an aside I should tell you full-blown IONCAP will now run on PC's. Getting back to the curves, you can see once we have these curves we don't need to talk about Fraction of days, FOT and all the other values. The values are only frequencies associated with ionospheric parameters and in general are not sensitive to power levels. If you want to know what frequency will produce some required S/N ratio for a given percentage of the time (Reliability), you have it as all frequencies are related to received S/N ratio. I hope you agree the concept is "easy". If you say no, you might go over it again and try to relate to the discussion no matter how repugnant it may be to those of us brought up with and/or now using other programs. You say "OK? what happened to the Fraction of days, FOT, etc.". Well they're embedded in the curves and I guess we should discuss them and draw another graph with some labels on it at specific frequencies. In short, all our favorite terms can be identified on the curves and most went into building the curves but lost their identity after the curves were generated. You say, hold on, I use IONCAP and you show all these things on the printed predictions. Yes you're right, but we started out to build a small computer program too. The sponsor and chief payman demanded we print these things out so the output could easily be talked about and used by those people whose vocabulary contained the acronyms. Using the curves, many calculations and variations were wanted and the small program didn't stay small long. Again, I guess, life is like that, but I hope you keep the IONCAP philosophy and language separate from the more usual terminology of most other HF prediction programs. I'm sure it will improve your better health and reduce your frustration when using IONCAP and presenting results or just talking to other hams who use prediction programs. (Figure 2)
The graph, we said we would draw, showing our missing terms is Figure 2 and sure enough there they are, even "Fraction of days" -- so important in our upbringing. The portions of the curves at frequencies above the HPF, FOT and MUF had no meaning before IONCAP because the received signal was assumed not to exist due to complete penetration of ionosphere. This meant no signals were present above some percent of days for an MUF defined for a given time of day. No signals were calculated above this defined MUF(10%), MUF(50%) or whatever. No "OVER-THE-MUF" calculation were made. Signals were assumed to be present only below the MUF cited for some given percentage of time. These terms are then related to "Fraction of days" within a month when a given frequency does not penetrate a layer in question due to the distribution about the monthly median MUF (50%). Terms like fraction of days are meaningless in IONCAP because only two things can happen. Your operating frequency is below the MUF or it's above the MUF and IONCAP calculates signal levels below and "ABOVE-THE-MUF." Therefore, the "Fraction of the days" is not meaningful anymore when thinking and talking IONCAP. Let's take a specific example and wrap this discussion up in hopes we are not beating a "dead dog". OK, your operating frequency is 20 MHz and you wish to know what percentage of time your required S/N ratio of 10 dB will be available. As you look at Figure 2 you see that at a frequency of 20 MHz your required S/N ratio of 10 dB falls between the 50% and 90% curves roughly where we would draw the interpolated 65% curve. This means 20 MHz will be available 65% of the time with at least a 10dB S/N ratio. Please note the "fraction of days" at 20 MHz would have been around 20%, but can still have 65% "Reliability" when your required S/N is 10 dB. Note that the LUF is defined in the graph for 12 dB required S/N. It can be defined for any arbitrary value of required S/N and Reliability. This paper used the night-time ionosphere as an example, but the daytime only gets a few more "wrinkles" in the curve due to the regular E-layer MUF being much higher and the F1-layer being present at some times and at some latitudes in certain months of the year. These Reliability curves are generally not seen in the IONCAP output or even really completely generated at any one time -- unless needed for special cases. So if you run IONCAP don't expect to see them --just remember they describe how the computer program really works.
Yes, IONCAP is an old prediction program and maybe still not fully understood due to lack of documentation, but it has been used and is still being used by over 100 countries and 250 different agencies in the US, not counting Ham operators who are beginning to use it more and more as they become more acquainted with it.
Kind of an easy concept isn't it? Might even explain--at least partly -- the reason the higher power stations stay around longer when the band begins to leave us --- so to say.
|
|
