Skip to document

Multicast Routing

Multicast Routing Notes
Course

Systems Programming (01:198:214)

23 Documents
Students shared 23 documents in this course
Academic year: 2021/2022
Uploaded by:
Anonymous Student
This document has been uploaded by a student, just like you, who decided to remain anonymous.
Rutgers University

Comments

Please sign in or register to post comments.

Preview text

Multicast Routing

As an alternative to the link-level multicast offered by multi-access networks, IP offers an IP- level multicast to better allow many-to-many and one-to-many communication. Since IP multicast is now being discussed, we also need a name for the conventional one-to-one service of IP that has been detailed thus far: It is known as a unicast service. With each group having its own IP multicast address, the fundamental IP multicast paradigm is a many-to-many approach based on multicast groups. Any packets transmitted to the multicast address of a group are copied to the hosts in the group. A host may be a member of many groups, and it may freely change groups by notifying its local router via a protocol that we shall cover in more detail later.

Multicast addresses are thus connected with an abstract group, whose membership varies dynamically over time, unlike unicast addresses, which are thought of as being associated with a node or an interface. Additionally, any host may send multicast traffic to a group under the original IP multicast service paradigm; the sender does not need to be a member of the group, and there may be any number of such senders for a particular group. A host delivers a single copy of the packet addressed to the group's multicast address, using IP multicast to send the exact same packet to each member of the group. As we will see, the routers in the internetwork share information of each group member's unique unicast IP address, so the sending host does not need to be aware of each one.

The routers will create copies of the packet anytime they need to forward it across more than one link, so the transmitting site doesn't need to send extra copies of the packet. IP multicast is more scalable than using unicast IP to send the same packets to numerous receivers because it avoids redundant traffic (packets) that would have been sent repeatedly over the same links, particularly those close to the sending site. Support for a type of one-to-many multicast has been added to IP's initial many-to-many multicast. A receiving host specifies both a multicast group and a particular sending host in this one-to-many multicast architecture, known as Source-Specific Multicast (SSM). Multicasts directed to the given group would then be received by the receiving host, but only if they are sent by the specified sender.

The SSM approach is applicable to many Internet multicast applications, such as radio broadcasts. IP's initial many-to-many approach is frequently referred to as Any Source Multicast to distinguish it from SSM (ASM). A host communicates with its local router using a specific protocol designed for this purpose to indicate if it wants to join or leave a multicast group. This protocol is known as the Internet Group Management Protocol (IGMP) in IPv4 and the Multicast Listener Discovery protocol in IPv6 (MLD). The task of ensuring proper multicast behavior with respect to that host falls to the router at that point. The router periodically polls the LAN to find out which multicast groups are still of interest to the attached hosts because a host could fail to quit a multicast group when it should (for instance, following a crash or other failure).

A portion of the IP address space is set aside for multicast addresses. These addresses are assigned in the class D address space in IPv4 and in the multicast group addresses section of the IPv6 address space (see Table 4). The multicast ranges contain several subranges that are designated for intradomain multicast, allowing various domains to utilize them independently.

When we exclude the prefix shared by all multicast addresses, there are 28 bits of possible multicast addresses in IPv4. When trying to use hardware multicasting on a local area network, this creates an issue (LAN). Take Ethernet as an example. When their shared prefix is disregarded, Ethernet multicast addresses are only 23 bits long. To put it another way, IP must translate 28-bit IP multicast addresses into 23-bit Ethernet multicast addresses in order to benefit from Ethernet multicasting. The low-order 23 bits of any IP multicast address are used as the Ethernet multicast address, and the high-order 5 bits are ignored. Unicast forwarding tables on a router specify the link to utilize to forward a unicast packet for any IP address.

A router must also have multicast forwarding tables that specify which links—possibly more than one—to utilize to forward the multicast packet based on the multicast address in order to implement multicast (the router duplicates the packet if it is to be forwarded over multiple links). Because of this, multicast forwarding tables collectively specify a collection of trees called multicast distribution trees, as opposed to unicast forwarding tables, which collectively specify a set of pathways. Additionally, the multicast forwarding tables must specify a set of trees in order to specify which links to use based on the combination of the multicast address and the (unicast) IP address of the source in order to support Source-Specific Multicast (and, it turns out, for some types of Any Source Multicast). Multicast routing, or more specifically, the method of creating the multicast forwarding tables, is how the multicast distribution trees are established. Similar to unicast routing, a multicast routing protocol must not only "work," but also scale reasonably well as the network expands and support the independence of various routing domains.

Was this document helpful?

Multicast Routing

Course: Systems Programming (01:198:214)

23 Documents
Students shared 23 documents in this course

University: Rutgers University

Was this document helpful?
Multicast Routing
As an alternative to the link-level multicast offered by multi-access networks, IP offers an IP-
level multicast to better allow many-to-many and one-to-many communication. Since IP
multicast is now being discussed, we also need a name for the conventional one-to-one service of
IP that has been detailed thus far: It is known as a unicast service. With each group having its
own IP multicast address, the fundamental IP multicast paradigm is a many-to-many approach
based on multicast groups. Any packets transmitted to the multicast address of a group are copied
to the hosts in the group. A host may be a member of many groups, and it may freely change
groups by notifying its local router via a protocol that we shall cover in more detail later.
Multicast addresses are thus connected with an abstract group, whose membership varies
dynamically over time, unlike unicast addresses, which are thought of as being associated with a
node or an interface. Additionally, any host may send multicast traffic to a group under the
original IP multicast service paradigm; the sender does not need to be a member of the group,
and there may be any number of such senders for a particular group. A host delivers a single
copy of the packet addressed to the group's multicast address, using IP multicast to send the exact
same packet to each member of the group. As we will see, the routers in the internetwork share
information of each group member's unique unicast IP address, so the sending host does not need
to be aware of each one.
The routers will create copies of the packet anytime they need to forward it across more than one
link, so the transmitting site doesn't need to send extra copies of the packet. IP multicast is more
scalable than using unicast IP to send the same packets to numerous receivers because it avoids
redundant traffic (packets) that would have been sent repeatedly over the same links, particularly
those close to the sending site. Support for a type of one-to-many multicast has been added to
IP's initial many-to-many multicast. A receiving host specifies both a multicast group and a
particular sending host in this one-to-many multicast architecture, known as Source-Specific
Multicast (SSM). Multicasts directed to the given group would then be received by the receiving
host, but only if they are sent by the specified sender.
The SSM approach is applicable to many Internet multicast applications, such as radio
broadcasts. IP's initial many-to-many approach is frequently referred to as Any Source Multicast
to distinguish it from SSM (ASM). A host communicates with its local router using a specific
protocol designed for this purpose to indicate if it wants to join or leave a multicast group. This
protocol is known as the Internet Group Management Protocol (IGMP) in IPv4 and the Multicast
Listener Discovery protocol in IPv6 (MLD). The task of ensuring proper multicast behavior with
respect to that host falls to the router at that point. The router periodically polls the LAN to find
out which multicast groups are still of interest to the attached hosts because a host could fail to
quit a multicast group when it should (for instance, following a crash or other failure).
A portion of the IP address space is set aside for multicast addresses. These addresses are
assigned in the class D address space in IPv4 and in the multicast group addresses section of the
IPv6 address space (see Table 4.1). The multicast ranges contain several subranges that are
designated for intradomain multicast, allowing various domains to utilize them independently.