So, having perused the SmartBeacon code for a while now, I have a
couple questions...

// Check to see whether we've sped up sufficiently for the
// posit_next_time variable to be too far out.  If so, shorten
// that interval to match the current speed.

   if ( (posit_next_time - curr_sec) > sb_POSIT_rate)
       posit_next_time = curr_sec + sb_POSIT_rate;

Why is this statement in there?

It looks like you are checking to see if the interval is too long for
the speed.

This is looked after by this statement.

sb_POSIT_rate = (sb_posit_fast * sb_high_speed_limit) / speed;

The interval decreases as speed increases.

The one shortfall in the SmartBeacon routine is that it does not
report decreases in velocity. As velocity decreases, the beacon rate
increases, which means that the next beacon will be further away in
time.

Let's use:

high_speed = 60
high_rate = 120 (sec)
low_speed = 5
low_rate = 1800 (sec)

If I am travelling at 60, and send a beacon, my next beacon should
happen in 120 seconds. If I stop right away, my beacon rate will drop
to 1800 seconds. This leaves APRS clients dead reckoning me at 60 mph
until that next beacon gets sent.

If I am travelling at 10 mph, my rate is one beacon every 720 seconds
(12 minutes). If I accelerate to 20 mph, my rate increases to on
beacon every 360 seconds (6 minutes).

Increases in speed increase beacon rates, so there is no need to worry
about bumping rates up. When we slow down though, the opposite is
true.

At high speeds, SmartBeaconing is the primary reason for position
reports being sent out, with the occasional beacon forced due to
CornerPegging. At low speeds, the reverse is true. CornerPegging
becomes the primary reason for beacons being sent, with the very rare
occasion when a SmartBeacon triggered beacon is triggered.

In everyday driving, you will find that the combination of
SmartBeaconing and CornerPegging do a good job of representing your
track.

James
VE6SRV
_______________________________________________
Xastir mailing list
Xastir@xastir.org
http://lists.xastir.org/cgi-bin/mailman/listinfo/xastir

Reply via email to