Send the GeoNetworking lifetime as 1 s, not 3200 s

geonet_wrap_shb wrote lifetime 0x83, commented as about 60 s. The field
holds the multiplier in its upper six bits and the base in the lower two
(50 ms, 1 s, 10 s, 100 s), so 0x83 is 32 x 100 s = 3200 s. That is over
the 600 s itsGnMaxPacketLifetime a sender may use at all; vanetza
refuses to send such a packet. The byte arrived with the Phase 03 commit
as a placeholder and was never checked against the encoding.

For a single-hop CAM this is non-compliance rather than a functional
fault: nothing stores or forwards an SHB packet, so no receiver acts on
the value, and no dropped CAM was ever traced to it.

0x05 (1 x 1 s) is what every other station in
its-g5-receiver-firmware/recordings sends its CAMs with; the recorded
GeoBroadcast DENMs use 0x79 (30 s). Changed in obu-cam-transmistter's
copy as well. Both firmwares build (IDF 6.1 and 5.5.4), and the
disassembled geonet_wrap_shb of each stores 0x05. Not yet seen on air:
the production OBU still runs the 2026-09-10 build, and the on-air
check is listed in TODO.md.

obu-cam-transmistter/main/geonet.c

@@ -19,7 +19,10 @@ int geonet_wrap_shb(const uint8_t *its_payload, int its_len,
     // ---- GN Basic Header (4 bytes) ---- (EN 302 636-4-1 clause 9.6)
     *p++ = (uint8_t)((1 << 4) | 1); // version=1, NextHeader=1 (Common Header, unsecured)
     *p++ = 0x00;                    // reserved
-    *p++ = 0x83;                    // lifetime (~60s in the base/multiplier encoding) - tune if needed
+    // Lifetime: multiplier in the upper 6 bits, base in the lower 2 (0 = 50 ms, 1 = 1 s, 2 = 10 s,
+    // 3 = 100 s). 0x05 = 1 x 1 s, what real stations send their CAMs with. Was 0x83, commented as
+    // ~60 s but decoding to 32 x 100 s = 3200 s. See obu-firmware's geonet.c.
+    *p++ = 0x05;
     *p++ = 1;                       // remaining hop limit = 1 (SHB single-hop; matches CAM in the Rust reference)
 
     // ---- GN Common Header (8 bytes) ---- (clause 9.7)