Current work
Multicast Protocols
"Host Based Multicast (HBM)"
As the Internet is expected to better support multimedia applications, new services need to be deployed. An example of one of these next-generation services is multicast communication, the one-to-many/many-to-many delivery of data.
Over the last ten years, multicast research and deployment efforts have both been very important.
Multicasting is the ability of a communication network to accept a single message from an application and to deliver copies of the message to multiple recipients at different locations.
The HBM proposal distinguishes core members (CM) that are part of the core distribution topology and non-core members (nonCM) that graft on the existing topology as leaves.
Some of the nodes can turn out to be unstable (e.g. a mobile node
with a bad wireless connection).
Even if HBM includes redundancy and failure discovery mechanisms,
instability must be taken into account when creating the topology.
The idea is to have stable transit nodes while unstable ones are
moved to the leaves of the topology.
Of course the stability of a node is unknown when he first joins
a session.
A default (conservative) value is first assigned to the node_stability
variable and this latter is regularly updated as time goes.
Everything is under the control of a central rendez-vous point (RP). This RP knows CMs and nonCMs and the distance between them.
This RP is responsible of the distribution topology calculation and
its dissemination among CMs and nonCMs.
CMs periodically evaluate the distance between them and inform the RP.
Likewise nonCMs evaluate their distance with CMs and inform the RP.
Some redundancy in the topology are added. An algorithm is supposed to add a certain number of Redundant Vitual Links (RVL) until the probability of having a partitioned topology after a node failure falls below a predefined threshold. Of course, some loops are created. Yet RVL are clearly identified as such and using a simple suppression mechanism is easy.
Back to Main Page