[EAS] FEMA to do webinar on results of national EAS test

[Harold Price]

Lets break it down:

There are two parts to EAN testing, each with two sub parts.

1) Getting the EAN message through the system.
  1a) getting it from FEMA PEP, then to SR, LP, and other high level 
infrastructure relay points, like the NPR squawk channel.
  1b) getting it from LP to local stations.

2) Having the local devices do the special things that need to be 
done with an EAN.
  2a) Having the device follow the proper procedures when it gets an EAN.
  2b) Having individual users configure the devices properly.

To a large extent, each of these can be tested separately, without 
getting the public involved (for example, needed PSAs to tell them it 
is just a test).  Many of the EAN differences can be tested in the lab.

Part 1a.

The point here is to test nations distribution via the legacy EAS 
relay system in as non-invasive a way as possible.  We should test 
the high level stations more often than regular participants, 
(because the lower down the hierarchy you go, the more likely it is 
that RMT testing is sufficient to check the local links). We can test 
the national parts of the system without using the EAN code.

This can and should be tested with one of the alternate test 
codes.  We have a set of alternate test codes already, and new event 
codes can be added to the much smaller number of very high level 
stations (PEP, for example).  That test can be used in the full-up 
all call distribution mode from FEMA to the PEP stations.  The PEP 
stations would put it on the air (run the test at 4:35am).  This 
would test the parts of the system that failed Nov 9.  Do it more 
than once a year.  Stations that monitor PEP stations directly could 
see this message in their logs.

Then, less often than the above, run a test from FEMA with a 
different test code, one that SR and LP stations would relay (using 
regular RMT relay rules, not immediate, with a large window.  Also 
run this late a night.  Almost everyone could see that message in their log.

Finally, about a year, replace one month's RMT with an NPT.  Everyone 
puts a wild card location NPT filter in their device.  Allow an large 
window to relay it, so it can fit into regular programming (like an 
RMT). Everyone should relay that alert.  The regular stations don't 
get any extra tests to run.

Then, at a slower repetition rate, run a live code EAN.

This testing checks the national distribution channels in a meangful 
way. With proper use of test codes, we don't need to use the EAN very often.

Part 1b
This is simply the regular monthly RMT in most cases.  Local areas 
should modify the test plan to have the local source of EANs generate 
the RMT from time to time.

Part 2a
This can be done in the lab.  The first part is deciding what to do 
with the types of problems that cropped up on Nov 9, and getting them 
fixed, then verifying them in a suitable equipped lab, with people 
who are looking for odd conditions.  We don't need to find these 
problems by running a live test.

Part 2b.
Much of this work was done for Nov 9.  Stations that didn't get it 
right last time, meaning they had a problem that is entirely within 
their control to fix, should know by now what that fix is, and they 
should be implementing it.  FOr the part 1 testing above, they will 
need to enter a "filter" for a NPT, but everyone has had a refresher 
course on how to adjust their device.

The plan about can be tweaked as needed.  This is a type of plan that 
will require more work at the top of the chain (where there are much 
fewer stations involved) and much less work by the majority of stations.

We need to improve and increase the rate of testing at the high 
levels in the chain, where a failure effects a large number of stations.

The actual difference to the regular EAS participant between an EAN 
and a regular alert is very small, and the problems on Nov 9 had 
nothing to do with that difference in almost every case.

Harold

At 08:54 PM 11/19/2011, Adrienne  Abbott wrote:
>  As I understand it, these event codes would
>have to be programmed for each state and it would take several activations
>to get them to every state in the country.