False-resistant POP detection - an idea

[kpatz]

I came up with an idea that, if implemented by the RD manufacturers, could possibly reduce or eliminate POP falses.

Since most POP falses are caused by the LO sweep emissions in certain detectors, and this sweep repeats several times per second, what do the experts here think of an algorithm similar to this to weed out Cobras from BEE IIIs:

1. When a POP burst is detected, instead of alerting immediately, wait for a period of time (<1 second) to see if the burst repeats.

2. If it repeats within the time period, do not alert. If it doesn't repeat, then alert.

About the only drawback is a slight delay in alerting to POP bursts, but if you're POPped and you're the front runner, you're dead anyway. With enough sensitivity, this would allow the detector to pick up POP bursts when other cars are clocked up ahead. Also, the detector could still false if it only picks up one LO sweep from a junk detector. But I figure this would reduce the occurrence of falses quite a bit.

Another possible drawback, if implemented, is MPH adding a multiple POP burst that emulates a Cobra, causing the detector not to alert.

VR has a method for detecting POP falses, but it's not perfect. Worse yet, it sometimes junks out real Ka threats (that aren't even POP at all).

Any thoughts from the peanut gallery?

[jimbonzzz]

Your general idea could work. I would have drawbacks like you mentioned, and would probably be a tradeoff between missing some POP hits vs less falses. But if it leads to more "believable" POP alerts that are taken seriously, it might be a good thing.

The only problem I see with it is:

-With the Enforcer and Bee, if the unit is in POP mode and the transmit button is simply held down, it will continuously transmit POP bursts at a rate of 3-4 times a second. Sometimes, it seems easy to accidentally send multiple bursts without even trying. In fact, in the GOL tests, sometimes the botton would be held down too long accidentally, resulting in two consecutive bursts being sent, so we would have to re-do that particular POP hit.

It stands to reason that if the junk detector is close enough to cause multiple consecutive POP alerts on the good detector, the good detector should be able to filter the false alert using one of the conventional methods at least *some* of the time. But maybe your filter idea could potentially cut down on the falses in times where the other filtering doesn't suppress it.

Jim

[kpatz]

I wasn't aware that the MPH guns produce multiple bursts when the button is held. That's probably why the RD manufacturers haven't done something along the lines of my idea...

[jimbonzzz]

The manual cautions against sending multiple busts in a short time though, saying something like "5% of detectors might be able to detect it" if used in that manner. Probably a moot point now though since most of the MFRs have added POP detection...

Jim

[dave1m]

I wonder what speed the CPUs of the detectors that can detect POP are using. Seems to me that a key to catching POP bursts and seperating them from LO emissions would be better resolution of the signal. A very fast CPU would allow closer examination of the signal in the same amount of time. I don't know what the differences are between a POP burst and a LO emission, but I would expect there to be characteristics to identify each.

just my $.02

Dave

[jimbonzzz]

4 or 8 Mhz

[dave1m]

4 or 8 Mhz

So wouldn't running a faster processor, say, 100, 200, even 500Mhz, give you the ability to check out the signal characteristics to determine if it is real or false before reporting it? I am not an enginner, so maybe I am barking up the wrong tree, but it makes sense to me. It's not like you have to report the POP event as it happens. That 67mS POP can be reported a small time later after signal interrogation. Surely the LO harmonics look different than a POP burst?

Dave

[jimbonzzz]

There ARE some things that could be done along the lines of what you mention, but currently the costs would be prohibitively high. But I think in the future as technology gets better and prices drop we might see some of this though...

Jim

[Mackid343]

if you're POPped and you're the front runner, you're dead anyway.

i thought you couldnt be pulled over unless the leo switched the gun to constant on to obtain a reading

[jimbonzzz]

if you're POPped and you're the front runner, you're dead anyway.

i thought you couldnt be pulled over unless the leo switched the gun to constant on to obtain a reading

True, but this can happen very fast without much time to slow down: once a POP speed is displayed, a second press of the transmit button puts it into constant on.

Jim

[dave1m]

I've been ticketed without picking up anything on my V1 (1.7). I believe there was a BEE III in the trooper's car. All he would tell me was that he "got me on radar". I am about 90% sure it was a POP only ticket and my lawyer advised against fighting it so I got it deferred. Don't think that they won't use POP to write tickets.

Dave

[dave1m]

BTW has anyone other than VR tested these POP units for accuracy? I was curious to see data that backed up VR's claim that they are error prone. Seems like every test I read about that uses a POP unit doesn't mention anything about the unit's accuracy.

Dave

[jimbonzzz]

BTW has anyone other than VR tested these POP units for accuracy? I was curious to see data that backed up VR's claim that they are error prone. Seems like every test I read about that uses a POP unit doesn't mention anything about the unit's accuracy.

Dave

Yes, we tested POP vs a laser gun at the GOL tests. In that situation, the speeds were very close.

Jim

[compu44]

radartest.com also said that they were very close.

[TexasRed]

What about sweeping for PoP bursts only @ ~33.800Ghz and/or 25.150Ghz?

[jimbonzzz]

What about sweeping for PoP bursts only @ ~33.800Ghz and/or 25.150Ghz?

From my understanding, they do that already.

[TexasRed]

interesting