[EAS] Manual Relay and Immediacy

[Sean Donelan]

On Sat, 20 Aug 2016, Al Kenyon wrote:
> I'm not going to get into the minutiae of forwarding upon confirmation of
> header data  or waiting until the entire message has been received and
> forwarding upon receipt of an EOM. This is the crux of the current debate
> over "immediately".  I don't know that there is written guidance from the
> Commission on this point unless it is buried somewhere in the text of the
> 6th R&O. In the case of the NPT, with its roughly 30 second audio message,
> the overall difference in forwarding time will be less than a minute. One
> can make valid arguments for both positions. The Commission is rightly
> concerned with message delivery latency, particularly in the case of short
> notice, life threating situations. We hope to learn a lot about how EAS
> functions in the real world from some of the more detailed information that
> will be collected in ETRS form 3.

In various congressional testimony and previous statements, the FCC has 
said its role as the regulatory partner, with FEMA and NOAA, in the 
Emergency Broadcast System and now the Emergency Alert System is 
maintaining and enforcing its rules. But, the FCC said it does not
determine the requirements for EBS/EAS. The requirements came from the 
(1) White House, (2) FEMA (and FEMA's predecessor agencies), (3) NOAA and 
(4)industry. That may partially the FCC just telling Congress, don't 
blame us; but the historical record does backup the FCC's claim.  The
FCC's Part 11 rules are supposed to reflect the requirements determined
by others.

However the FCC's Oracle of Delphi approach causes confusion, even among 
the lead agency partners.

1. While everyone is trying to make the system work as well as possible,
a small part of the debate on "immediacy" is some participants saying 
they spent a lot of money interpreting the rules one way, and upset that 
other participants aren't also spending money to do it the same way. And 
vice-versa. Although it may not be valid justification, a lot of lobbying 
in Washington DC happens because of money. Nobody cares about the rules, 
until it affects their pocketbook.

2. What is the "intent" of a National Periodic Test (NPT)? Should a NPT
test the entire system, end-to-end, exactly the same as an Emergency 
Action Notification (EAN), including "immediate" activation, "pre-emption" 
of lower priority messages, "live" and "unlimited" transmission length, 
etc. Or is the intent of a NPT to test a subset of EAN's properties in 
the real-world, and other properties can be tested in a lab.

In the Sixth R&O, the FCC said the "intent" was the latter.  But the FCC 
wasn't as careful editing the rules about what subset of EAN properties 
the NPT must emulate.

 	Pre-emption		No
 	Unlimited length	2-minute maximum
 	Live transmission	Omitted
 	Maximum hold-off	Immediate

Other test codes have different rules than the system they are testing. 
For example, RMT has a maximum 60 minute delay, while actual event codes 
have a maximum 15 minute delay. For an RMT, does the 60 minute maximum 
delay mean from the beginning of the first EAS header code, or 60 minutes 
from the end of the last EOM? Does the 2-minute audio limit include or 
exclude the time used by EAS AFSK and 8-second Attention Signal?

As an engineer, I understand no amount of lab testing will ever replicate 
what happens in the real world. A simple broadcast chain at traditional 
radio station versus a multi-state, multi-hundred channel video provider
versus a cellular mobile network are very different.

IPAWS doesn't currently support "live" streaming of messages. IPAWS 
distributes a recorded audio file, at high-speed. For a 2 minute NPT, the 
difference may not matter.  For a 30 minute EAN, the difference between a 
"live" stream and a recorded audio file may make a difference.

Using the EAS analog protocol, there has always been a difference between 
manufactures.  Some manufacturers support "live" immediate relay for all 
event codes, with a small amount of buffering, via the EAS analog 
protocol. An immediate relay option wasn't limited to EAN/NPT, but was 
also available for TOR and other event codes. While other manufacturers, 
especially those focused on cable systems using SCTE-18, would record the 
entire EAS analog message and forward an audio file to set-top boxes for 
playout. EAN was handled differently, requiring force-tuning set-top 
boxes for "live" transmission. Is the intent of the NPT to test the 
force-tuning function of cable systems like an EAN, or can that be tested 
in a lab?

Which leads to more questions about buffering the beginning of an EAN/NPT
message during the force-tuning delay. Does the manufacturer "immediately"
transmit the "live" message, or buffer it during the switching delay as 
set-top boxes tune to a different channel. Unidirectional set-top boxes 
don't send a signal when they are ready. Complying with an "immediate" 
rule seems less important than a (small) delay so the public can hear the 
entire emergency message.

But that's why there is testing. And why I would be suprised by any FCC 
enforcement action for a good faith attempt to comply with reasonable 
interpretations of the rules.