[EAS] Overlap/monitoring issue

[Sean Donelan]

On Thu, 6 Mar 2014, David Turnmire wrote:
> I must say I'm surprised by this.  It was my understanding that the SAGE
> (and other boxes) would "queue" incoming alerts, rather than have one
> trash another... or ignore the second.   I believe some decoders had a

Most CAP devices queue CAP messages.

EAS over the air is much more of a crap shoot with different vendors 
interpreting ambinguities in the EAS specifications. While all EAS 
boxes will pre-empt a current message for an incoming EAN, and stick with 
an active EAN even if another EAN is received on another channel, how they 
handle multiple non-EAN messages seems to vary by software release and 
vendor.

Sage will pre-empt an in-progress non-EAN message if it receives a 
"different" EAS header on another channel with the same or higher 
"priority."  Lower priority and duplicate messages are logged without
interrupting the current message.

So if you are monitoring multiple NWS transmitters with overlapping but 
not identical coverage areas (i.e. the EAS location code list is 
different), the EAS message received on the first NWS receiver will begin 
and then a few seconds later when the second NWS receiver hears its its 
version of the EAS message (i.e. same event code, different location 
list), the Sage will send an EOM in mid-message and begin a new EAS 
message recieved from the second NWS transmitter.  This is very evident
with "Automatic Relay" such as for Tornado (TOR) event codes.

If you program the Sage with "Timed Relay" hold time of a few of 
minutes, you will generally avoid the intermediate EAS messages being 
interrupted, and only re-transmit the last NWS warning received. How
long the hold time should be depends on how quickly NWS' computers push 
the EAS messages through its different transmitters.  Or only monitor
one NWS transmitter.

Hopefully when NWS makes warnings available through CAP/IPAWS, this will 
be resolved by queuing all the messages.