The applet explores the performance of video-OR circuit used for combining uncorrelated noisy signals with a minimum of combining SNR loss; it evaluates the combining noise statistics and identifies the probability of largest signal selection.
The upper display shows the combined video-OR output signal+noise amplitude probability distribution for one signal (red) and two signals (blue). The shapes of the distributions at low signal levels are dependent upon the Number of Ports. For reference, the single-channel, no-signal case (green) for zero mean Gaussian noise is also shown. The distributions are updated in sympathy with the two signals, specified by their SNR. SNR scrollbars are linear power ratios, but annotated in decibels on the main display.
The lower left display indicates the threshold triggering probability performance. ETNR, the equivalent threshold-to-noise ratio defines the threshold level which is equal in level to a signal of numerically the same SNR. The probability of triggering for either signal 1 or signal 2 input separately is indicated by the blue and red spots.
The lower right display indicates the channel selection probability for each of the two signals, given their input amplitude ratio as a function of SNR level. The red and blue spots indicate the triggering probability of the two input SNRs.
For zero level signals, each channel selection probability is the reciprocal of the number of channels.
RF amplifier detector-video and vector-video receivers can be compared.

User Notes:
With signals and threshold in the default parameter situation (SNR2 = 7 dB at threshold and SNR1 is 1 dB less than SNR2), the probability of either of the signals actually triggering the threshold is Ptrig(1+2) = 0.69. The probability that signal 2 will be selected at the video-OR output is 57% and the probability that signal 1 will be selected at the video-OR output is 40%, which means that for these low SNRs, there is still about 3% probability that at any instant one of the other six noise-only channels would be selected.
It is interesting to note that with 1-dB signal difference, there is quite a high likelihood of the wrong channel being selected up to quite large SNRs. Reducing the number of ports to two increases the channel selection probabilities to 58.5% and 41.5% with no other options.
Reducing both SNRs to near zero, the upper display shows that the OR-combined noise output has a positive offset and only slightly increased noise probability at more positive levels.
Straight addition channel noise combining would retain a zero mean but the curve would be reduced in amplitude and broadened by a factor equal to the square root of the number of ports.