receivers in the MFT, for more detail about the joining/ leaving process related to this
protocol, these details can be found in [65].
In [65] some benefits are mentioned of this protocol as follows:
J Enhanced scalability by reducing forwarding state. Only branching tree nodes
keep information about the state of the group which allow groups to be bigger
without high storage space to keep state information about.
J Load distribution. At conventional multicast protocols, if a receiver is unable to
join the group for any reason, such as overload at the routers, the tree will become
partitioned. In REUNITE if this case happens then the router just needs to ignore
the JOIN message and lets an upstream router process this message and share the
load. This is because at REUNITE not all routers should keep information about
the multicast group members.
On the other hand, REUNITE has some drawbacks, such as:
J This protocol may face the scalability problem if the network is large, a problem that
is created because in this protocol each node whether it is a branching node or not has
to maintain a table to carry information about the network.
J The Join and Tree messages are sent periodically, which means an extra overhead is
happened.
J The protocol decides that a node leaves the group as soon as it does not receive a Join
message for a period of time. This assumption is not correct because there are other
reasons that force the receiver to stop sending Join messages other than the receiver
want to leave. This will cause loss of data for that receiver until it reconnected to the
group again.
3.2.4 Hop By Hop Multicasting routing protocol (HBH)
HBH [22] is a multicast tree management routing protocol which implements a multicast
distribution through a recursive unicast tree. It uses IP class D address for multicast group
addressing.
40