Since the program is intended to be run continuously, METEOR can be configured to restart automatically if the PC reboots after a power interrupt. For automatic restart add the following lines to the AUTOEXEC.BAT file:
CD METEOR
METEOR
In manual and automatic restart, METEOR allows the user 20 seconds to begin changing parameters (described later in this section). If the user does not respond during that time period, METEOR begins acquiring and processing signals using the parameters from the previous run.
The METEOR program can be terminated at any time by pressing "Q" or "q".
Process control and header parameters
The user can set parameters to control the run (process) and for documentation of the hardware configuration (header) from METEOR's initial screen. The defaults displayed are taken from the METEOR.CFG file, and reflect the output of the previous run. There are six process control parameters: (1) trigger size which identifies a potential meteor echo and initiates an audio sample, (2) quiet signal which is set to the minimum dB level, (3) the increase in the autocorrelation coefficient which identifies a potential meteor as an actual meteor rather than a false trigger, (4) the minimum duration of a meteor echo that is counted as a 'long' echo, (5) the dB offset for displaying power on the following screen, and (6) the output file name which defaults to the month, day, and year in MM-DD-YY format.
The trigger size is usually set in the range of 3 to 10 dB, depending on the radio environment where AMCA is located. A lower setting results in more meteor echoes being counted and potentially better statistics, but if set too low it may lead to more false triggers. The best way to determine the trigger size is to experiment with different values. For each setting an operator observes program execution for about an hour at several different times of day, while listening to the receiver's audio output to determine whether the program is correctly identifying meteor echoes.
The quiet signal parameter may be set to zero in localities where there are few airplane overflights. Where air traffic is more prevalent it should be set to the dB level corresponding to the S-meter showing no signal. This will allow METEOR to filter out airplane interference.
The autocorrelation increase that distinguishes meteors from false triggers may be set between 0 and 100 percent. Zero normally works quite well, since a meteor echo almost invariably increases the autocorrelation above the reference level, while electrical noise decreases it. The reference autocorrelation is usually in the range of 10 to 60 percent due to weak reception of the ground wave. A meteor echo usually produces an autocorrelation of 90 to 100 percent.
Long echoes are those where the signal power remains high for a length of time set by the user. The duration is measured in units of 0.1 seconds, and a reasonable value is 10, or 1.0 seconds.
Additional header parameters document the hardware elements and settings of the system. They are recorded in the output file along with the control parameters. METEOR depends on the user to set the header parameters correctly, or to accept the defaults if no hardware items have been changed and no receiver setting have been adjusted since the last run.
The header parameters (and sample values) are tuning frequency (61.24 MHz), receiver mode (LSB), antenna type (dipole), antenna beam point (315 degrees azimuth), radio type (R8500), squelch control (off), lock switch (on), noise blanker (on), BFO frequency (+0.33 KHz), IF shift (zero), attenuator switches (off), AGC speed (fast), and audio peak filter (off).
The initial screen is shown in Figure 4.1. The user changes parameters, continues with execution, or quits, by pressing the keyboard menu letters in the left column enclosed in "( )".
FIGURE 4.1
Control: (T)rigger size 5 (dB) Quiet signal (Z) -127 (dB; 0 = off) (A)udio increase 0 (% correlation) (L)ong duration (0.1 sec) 3 (0.1 sec) (D)B offset 0 (dB) (F)ile name C:\METEOR\01-19-97.HR Header: (M)egahertz 61.24 (B)and LSB Antenna(X) Yagi Beam (P)ointing 315 (R)adio R8500 (S)quelch control Off Loc(K) switch On (N)oise blanker On BF(O) frequency (KHz) +0.33 (I)F shift 0 Att(E)nuator 0 A(G)C speed Fast A(U)dio peak filter Off Execute: (C)ontinue [else wait 20 seconds] (Q)uit
When the program begins to acquire radio data a new screen appears. The cumulative meteor, long, and false counts are listed at upper left, along with current signal strength and audio autocorrelation values. The minute refers to the minute count of the current hour of sampling. Below these values is a time dependent graph. The upper, middle, and lower bounds are identified by dotted blue lines at -100, -120, and -140 dB. The signal level is shown at 0.1 second intervals by small dots. At intervals of 10 seconds a symbol for the autocorrelation of the reference audio is plotted in green, in the range 0 to 100 percent. Whenever a signal jump causes a triggered audio sample, the result is plotted as a red symbol if it is a meteor, or as a blue symbol if it is a false alert. Figure 4.2 shows a sample data acquisition screen where a short duration meteor echo occurs toward the beginning of the interval, noise produces a false trigger in the middle, and a long duration meteor is recorded near the end. The screen is refreshed about once a minute, when time reaches the right border.
FIGURE 4.2
