[EAS] Improving IPAWS and EAS

[Sean Donelan]

On Tue, 10 Jul 2018, Randall Miller wrote:
> Sean, what does 100% compliance mean?

To explain what I mean, let me explain what I do.  Unless I need them for 
client work, I have some Sage, Dasdec and Trilithic EAS boxes sitting on a 
test bench. I leave them hooked up and tuned to a dozen different radio 
stations, 1 ATSC TV station and 1 satellite station in Washington DC. 
Every month or so, I tune the radios to different stations. That means I 
monitor all the FM stations I can receive in Washington DC for a month 
each quarter; and rotate through the TV stations over the year.  I'm not 
doing it for enforcement, and I don't report anything to the FCC.

I've given up monitoring AM stations, except occasionally WFED.

Begin by separating the stations into two major groups - Primary and 
Participating.  I don't ding the Participating stations for problems due 
to a Primary source (LP, PEP, IPAWS, etc). If the Primary transmits a 
two minute car ad during the RMT, that's not the Participating stations 
mistake  I'll talk about the Primary sources later.

So I monitor a dozen+2 Participating stations (radio and TV) every month. 
The results tend fall into an asymetrical bi-model distribution (i.e. an 
U-shape), with the stations in at either end of the curve:

Groupe 1. Almost always perfect group (usually about 85%), with at most 1 
or 2 oops a year

Group 2. Almost always a problem group (usually about 15%), with (the 
same) problems nearly every week

Almost always perfect means every week I decoded at least one 
over-the-air EAS message, and every month I decoded and could understand 
the audio of an over-the-air EAS message.  They didn't need to be 
mandatory tests, but usually were. They needed to be valid and sane 
EAS messages.

Participating stations with problems every week, don't seem to fix them 
month after month.

Participating stations without problems, almost always immediately fix an 
oops by the next week, at most the same month.

Most common problems with the over-the-air audio:

Understandable over-the-air audio is the biggest EAS issue. I have no idea 
what they write down in their EAS station logs. Their EAS logs may be FCC 
compliant, but listen to what went out over the air.

1. Some TV stations don't mute *all* the program audio during the EAS 
message. It may have squawks, but can't understand the audio over the soap 
opera.

2. Some radio stations with switching problems, clip part of the EAS 
message at the beginning or end. It sounds like an EAS message, but can't 
be decoded.

3. Understandable audio issues with the relayed audio relative to the 
source and other programming, such as bleed through audio, lots of 
static, extremely quiet or overdriven audio, stereo/mono channel or HD 
audio problems.

4. Configuration issues: no attention tone with audio, long attention 
tone, short attention tone, invalid or wildly wrong FIPS codes such as 
000000 in a weekly test, or FIPS codes hundreds of miles away. Usually 
its a FIPS code from with the first letter of the alphabet or an adjacent 
FIPS code for a county on the other side of the state. These sound like 
EAS messages, but indicate some mistake with the configuration.

Annoying, but I don't count as a mistake, double or triple attention 
tones. It could be a problem with the source, transmission, receiver, or 
with the decoder. A flaw inherent with the EAS daisy-chain distribution.

Now back to Primary sources...

Primary sources (LP, SP, PEP) have a much more difficult job compared to 
Participating stations. I put almost all Primary source mistakes as 
"Learning Opportunities." So I don't include them in the Participating 
station stats. Primary stations are almost guarranteed to make a mistake 
every year, but rarely make the same mistake twice in a row. They 
appear to try to improve the next time. There is no typical mistake 
with Primary sources. The job of a Primary just has a lot more chances to
mess up. Generally, I just say "Wow, I wonder what happened?"