[GSoC 2026] 6LoWPAN mesh-under: framework + Trickle up, benchmark running

15 views

Skip to first unread message

Usham Roy

unread,

Jun 16, 2026, 5:42:45 PMJun 16

to ns-developers

Hi all,

A quick progress update on the 6LoWPAN mesh-under forwarding project, following my earlier intro.

It runs end to end now. The forwarding policy is split out into a pluggable strategy, the old flooding behaviour is kept as the default, and the Trickle based suppression strategy is in. You flip one flag in the benchmark example to run flooding vs suppression and the numbers come straight out.

Where things stand:

!2873 (the strategy framework, a pure refactor with no behaviour change) is open and in review. This is the one I would like to merge first. https://gitlab.com/nsnam/ns-3-dev/-/merge_requests/2873

!2881 (Trickle suppression, helper API, and a benchmark example with ICMP and UDP traffic) is stacked on it. https://gitlab.com/nsnam/ns-3-dev/-/merge_requests/2881

Early results are encouraging: in a dense network the suppression does about 4x fewer transmissions and more than 20x fewer collisions than flooding, with delivery staying high across 5 to 30 nodes. The single bridge topology shows a delivery trade-off on the thin path, which is what the adaptive jitter step will tackle next.