Skip to content

Resolve "[tracipy]: Further propagate messages after initial propagation"

Closes #3 (closed)

Implementation details

An ALOHA inspired approach is taken. After sending a message (to exploit the fact that received messages are automatically being resend), a redistribute routine is started, where the message is going to be resend after a random simulation steps period [1, tau].

If in this period, another InformationPeer distributes the same message (and it was received by the peer which scheduled the routine for himself), this is interpreted as a collision.

When such a collision occurs, a new period [1, 2*tau] is used to pick a random period from.

In the first implementation, a uniform distribution is used to pick the scheduled waiting period before redistribution. This might however change.

Affected classes

In the beginning it's better to implement the functionality only within the Car class. This way, cars have the behavior average vehicles which just observe the road condition, send and redistribute messages. CrazyCars and information beacons may be interpreted as the main information distributors about e.g. an accident and can therefore constantly send messages without backing off when others are distributing the same information. This can be decided later after some experiments however...

Edited by boyanh

Merge request reports

Loading