Content uploaded by Khaled Elleithy
Author content
All content in this area was uploaded by Khaled Elleithy
Content may be subject to copyright.
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 1
Deterministic Formulization of EndtoEnd Delay and
Bandwidth Efficiency for Multicast Systems
Syed S. Rizvi srizvi@bridgeport.edu
Computer Science and Engineering Department
University of Bridgeport
Bridgeport, CT 06601, USA
Aasia Riasat aasia.riasat@iobm.edu.pk
Computer Science Department
Institute of Business Management
Karachi, 78100, Pakistan
Khaled M. Elleithy elleithy@bridgeport.edu
Computer Science and Engineering Department
University of Bridgeport
Bridgeport, CT 06601, USA
Abstract
EndSystem multicasting (ESM) is a promising applicationlayer scheme that has
been recently proposed for implementing multicast routing in the application layer
as a practical alternative to the IP multicasting. Moreover, ESM is an efficient
application layer solution where all the multicast functionality is shifted to the end
users. However, the limitation in bandwidth and the fact that the message needs
to be forwarded from hosttohost using unicast connection, and consequently
incrementing the endtoend delay of the transmission process, contribute to the
price to pay for this new approach. Therefore, supporting highspeed realtime
applications such as live streaming multimedia, videoconferencing, distributed
simulations, and multiparty games require a sound understanding of these
multicasting schemes such as IP multicast and ESM and the factors that might
affect the enduser requirements. In this paper, we present both the analytical
and the mathematical models for formalizing the endtoend delay and the
bandwidth efficiency of both IP and ESM multicast system. For the sake of the
experimental verifications of the proposed models, numerical and simulation
results are presented in this paper. Finally, the proposed formulization can be
used to design and implement a more robust and efficient multicast systems for
the future networks.
Keywords:
Bandwidth Analysis, EndtoEnd Delay, End System Multicast, IP Multicast, Overlay Networks.
1. INTRODUCTION
There is an emerging class of Internet and Intranet multicast applications that are designed to
facilitate the simultaneous delivery of information from a single or multiple senders to multiple
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 2
receivers. Different approaches of multicasting have been suggested to improve the overall
performance of networks especially the Internet [1, 2, 3, 4]. These approaches are: multiple
unicast, IP multicast, and endsystem multicast. All of these methods have some advantages and
disadvantages but the last two approaches (IP multicast, and endsystem multicast) mentioned
above have had more research effort in terms of performance evaluation of networks. Multiple
unicast can be described as a service where one source sends the same copy of the message to
multiple destinations. There is a one to one connection all the way from the source to the
destination. No special configuration is required. In IP multicast, one source sends data to a
specific group of receivers. In this case, a unique and special IP address is used, a class D
address for the entire group. In addition, there is a special configuration adopted for efficiency
reasons. A tree rooted at the source is constructed and only one copy of the message is sent
since the routers along the paths to the destinations performed the necessary replication
functionalities.
Finally, the endsystem multicasting is a very promising application layer solution where all the
multicast functionality is shifted to the end users [7]. In an endsystem multicasting approach,
host participating in an application session have the responsibility to forward information to other
hosts depending on the role assigned by a central data and control server [15]. In this case, the
architecture adopted is similar to that of IP multicast with the difference that only IP unicast
service is required. Endsystem multicast uses an overlay structure, which is established on top
of the traditional unicast services. In this way, every pair of edges (sourcedestination) is a
unicast connection. The overlay has its meaning from the fact that the same link can have
multiple unicast connections for multiple pair of edges. Although, endsystem multicast seems to
have many advantages (no further changes to the network are required, user has more control of
the application layer, no need of special multicast router capability, etc), there is a penalty to pay.
In the overlay structure, hosts are able to multicast information and consequently use the same
link to redirect packets increasing the endtoend delay of the entire transmission process [3, 8].
Another problem is the number of receivers that a potential “multicast” host can support. End
users have a limited bandwidth and suffer the last mile problem [9].
While these different multicast approaches can displace some of the costs of facetoface
communications, their true potential business benefit lies in improving the accessibility and
timeliness of information, vastly increasing its value to both the organization and individual
employees. Although research on multicast dates back to the early days of the Internet, it has yet
to produce a multicast service that is ubiquitously and economically available. In spite of the
performance advantages, commercial deployment of multicast has not yet been fully realized.
One of factors that prevent the widerange deployment of multicast is the difficulty in providing
reliable multicast transport.
Fortunately, recently there has been a renewed effort to realize different approaches of
multicasting. In this paper, we are also going to analyze these different methods of multicasting
with respect to their performance differences.
2. THEORETICAL ANALYSIS OF MULTICAST SYSTEMS
The Internet consists of interconnected LANs. A LAN may or may not have native multicasting
support and the same holds for the IP layer on top of the LAN. In this section, we will theoretically
analyze the problems of different level of multicasting, which hinder their performance with
respect to the bandwidth utilization and latency.
2.1 Multiple Unicast Systems
In the unicast IP network, the host acting as the source transmits a copy of the message to each
destination host as shown in Fig. 1. No special configuration is needed either in the source or in
the core network. The intermediate routers will have to carry all these messages to the proper
destinations. If, for example, the source host transmits ten copies of the same message at the
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 3
same time, then obviously ten times the bandwidth of one message is required on source host’s
network and on the router connected to the source host’s network. The chains of protocol entities
that take care of the transmission process also use processing capacity on the host for each
transmission. In addition, the transmission time is increased ten times and it will affect the global
endtoend delay. We pay special attention to the nearest link between the first router and the
source since it is mostly there where the maximum bandwidth consumption takes place. These
are the reasons to consider a multiple unicast service an unpractical approach to implement on
the network.
2.2 IP Multicast Systems
IP multicast is a service where one source sends data to a group of receivers each of them
containing a class D address as membership identification. IP multicast has long been regarded
as the right mechanism due to its efficiency. In IP multicast, a packet is sent only once by the
source. Routers along the route take care of the duplication process.
The IPmulticast capable version of the network shown in Fig. 2 consists of network with native
multicast support. IP multicast capable routers are consider along the path. Efficiently routing
multicast protocols are implemented. The traditional process includes the construction of a
sourcerooted tree together with the members of the multicast group. Since only one copy of the
message is required, we can say that a minimum bandwidth effort is being used for the
transmission of the message to all group members connected in the network. The problem for IP
multicast is that there is no commercial support for multicast routers. Investors still think that there
is not enough multicast application demand and that multicast traffic could take their routers down
due to congestion problems.
Source 1
R 1
Source 2
Router 2
Source 3
Source 4
1
2
3
4
5
1
1 2
Source 1 wants to
send to 3 different
destinations
Source 3 wants to
send to 4 different
destinations
Source 2 wants to
send to 2 different
destinations
Source 4 wants to
send to only one
destinations
1
2
3
4
5
6
7
8
9
10
1
1
2
3
4
5
6
7
8
9
R

4
R3
D10
D9
D7
D6
D5
D4
D3
D2
D1
D8
FIGURE 1: Example of Multiple Unicast
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 4
The IPmulticast transmission takes the same bandwidth on source host's network as a single
copy, regardless of how many clients are members of the destination host group in the Internet.
The obvious difference between the multipleunicast and IP multicast is that IP multicast scales
very well. Even if not all LANs have native multicast support, the added cost of transmitting
copies will be limited to a single LAN.
Besides the advantages of IP multicast, there are also certain drawbacks of this approach. One of
them, like we mentioned before, is the deployment problem of IP multicast, which imposes
dependency on routers. The main disadvantage of IP multicast is the need of commercial routers
supporting multicast protocol. In theory, almost all routers support multicast but in practice this is
not the case [13]. This prevents multicast service fully implemented in Internet application
(experimental research has been conducted in the Internet in a special platform named The
Mbone) [10, 11].
Several approaches to multicast delivery in the network have been proposed which make some
improvements or simplifications in some aspects, but they do not improve upon traditional IP
multicast in terms of deployment hurdles. A major obstacle for deployment of multicast is the
necessity to bridge from/to the closest multicast router to/from the endsystems. Existing IP
multicast proposals [5, 6] embed an assumption of universal deployment, as all routers are
assumed to be multicast capable. The lack of ubiquitous multicast support limits the deployment
of multicast applications, which in turn reduces the incentive for network operators to enable
multicast [12]. Therefore, from the above discussion one can expect that we need another
multicast alternative in which network routers have not to do all of the work; instead each of the
host will equally contribute in the overall multicast process of the messages.
2.3 End System Multicast (ESM) Systems
Because of the limitations in IP multicast, researchers have explored an alternative architecture
named endsystem multicast, which built a system on top of the unicast services with multicast
functionalities. Endsystem multicast is a very promising application layer solution where all the
multicast functionality is shifted to the end users as shown in Fig. 3. In this approach, host
participating in an application session can have the responsibility to forward information to other
R3 R4
Source 3
Sender 4
1
2
3
4
5
1
2
3
4
5
6
7
8
9
10
D10
D9
Source 1
Router 1
Source 2
Router 2
1
2
3
1 2
1 2
1 2 3 4
1
2
3
4
5
6
7
8
9
1
2
3
1
D8
D7
D6
D5
D4
D3
D2
D1
FIGURE 2: Example of IP Multicast
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 5
hosts. Here, end users who participate in the multicast group communicate through an overlay
structure.
However, doing multicasting at endhosts incurs in some performance penalties. Generally, end
hosts do not handle routing information as routers do. In addition, the limitation in bandwidth and
the fact that the message needs to be forwarded from hosttohost using unicast connection, and
consequently incrementing the endtoend delay of the transmission process, contribute to the
price to pay for this new approach. These reasons make endsystem multicast less efficient than
IP multicast.
The structure of the endsystem multicast is an overlay in a sense that each of the paths between
the end systems corresponds to a unicast path [14]. In other words, endsystem multicast is built
on top of the unicast services provided by network on transport layer. Here the membership and
replication functionality is performed by the end receivers, which connect together over unicast
channels to form a multicast tree, rooted at one data source. The end receivers could play the
role of parent or children nodes. The parent nodes perform the membership and replication
process. The children nodes are receivers who are getting data directly from the parent nodes.
There is one central control server and one central data server residing in the same root source.
Any receiver can play the role of parent to forward data to its children. Each client has two
connections: a control connection and a data connection.
3. ENDTOEND DELAY APPROXIMATION FOR MULTICAST SYSTEMS
This section presents the mathematical model for approximating the endtoend delays for all
multicasting schemes. For the ease of simplicity, we divide our approximation for each type of
multicasting approach such as unicast, multiple unicast, IP multicast, and ESM.
3.1 Model and Assumptions
Let G is an irregular graph that represents a network with a set of N vertices and M edges such
as:
{
}
,
G N M
=
. Let L is a direct communication link between a single pair of source (s) and
FIGURE 3: Example of ESM, Solid Lines Represent 2 Way Packet Transmission, Dotted Lines Represent
One Way Packet Transmission.
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 6
destination (d) where both source and destination belong to N such as:
{
}
,
s d N
∈
. In addition,
each packet transmitted between source (s) and destination (d) must traverse one or more
communication links in order to reach the final destination.
Let the value of D(L) denotes packetdelay (we sometime refer it as link delay) that is associated
with each direct communication link. Therefore, each transmitted packet will typically experience
a delay of D(L) on a particular link. The delay includes transmission, processing, and propagation
delays such as: LinkDelay = D(L) = Transmission Delay + Propagation Delay + Processing Delay
where L
∈
M. In connection less communication such as IP network, there might be multiple
routes exist between a pair of source and destination. As a result, each packet might follow a
different route in order to reach the final destination where each route requires traversing of one
or more communication links (L). A single route between a pair of source and destination can be
defined as:
{
}
{
}
, ,
R s d where s d N
∈
. System parameters along with their definitions are
presented in Table 1 and Table 2.
3.2 Mathematical Model for a Unicast System
In unicast, a packet is sent from one point (source) to another point (destination). As mentioned
earlier, when packet transmit from one source (s) to a specified destination (d), there exist
multiple routes where each route can have multiple links. This implies that the packetdelay for
unicast is entirely dependent on the number of links a packet needs to traverse in order to reach
the final destination system. Based on the above argument, one can define the packet delay such
as:
1 2
( ) ( ) ( ) ......... ( )
n
D R D L D L D L
= + + +
where n is the maximum number of links that need to be
traversed on route R between s and d. The delay can be generalized for one particular route (R)
that exist between source (s) and destination (d) such as:
1
( ) ( )
n
i
i
D R D L
=
=
∑
(1)
{ }
1 2
1
( ) ......
n
i n
i
where L L L L M
=
= + + + ∈
∑
Equation (1) can be further expressed as:
( ) ( )
( )
s d L R s d
Delay D D L
−∈ −
= =
∑
(2)
where
(
)
L R s d
∈ −
represents the value of the total delay associated with the route R between
source s and destination d.
Based on (1) and (2), one can also simply derive a mathematical expression for estimating an
averagedelay (denoted by AD) which each packet may typically experience if it traverses one of
the available routes. The mathematical expression for an estimated AD is as follows:
( )
( )
1
y
i
s d
i
AD D R y
−
=
∑
(3)
where y represents the maximum number of possible routes between source s and destination d.
In addition to the average delay, one can also chose the optimal route with respect to the
minimum delay that each packet may experience when traverse from one particular source s to a
destination d such as:
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 7
( )
( )
1
y
i
s d
i
OD Min D R
−
=
=
∑
(4)
Equation (4) gives minimum delay that each packet experiences when transmitting between a
pair of s and d. We refer this as an optimal delay (OD) for a pair of s and d. The above derivations
can be further extended for the multiple unicast system where a single source (s) can transfer a
packet simultaneously to multiple destinations. This hypothesis leads us to the following
argument: multiple routes can be established between the source (s) and each destination
system. The following mathematical expression can be used to estimate the total delay that the
entire packet transmission will experience in a multiple unicast system:
( )
( )
1 2
( ) , ,......, 1
y
y
Total MU i
multiple
s d d d i
D D D R
→ =
= =
∑
(5)
where y in (5) is the maximum available unicast routes between a particular source (s) and
multiple destinations.
Although, in multiple unicast system, a single packet can be transmitted from one source to
multiple destinations, the transmitted packet may follow a different route in order to reach the
appropriate destination. Consequently, each packet transmission may yield a different delay
depending on the number of links the packet needs to traverse on the chosen unicast route. This
leads us to the following mathematical expression for an average delay:
Parameters Definition
D(L) Denotes packetdelay that is associated with each direct communication link
s Represents source
d Represents destination
C
n
Represents the child node
P
n
Represents the parent node
D(R) Denotes maximum number of links that need to be traversed on route R between s
and d
AD Averagedelay
OD Optimaldelay for pair of s and d
( )
Total MU
D
Estimate of the total delay that the entire packet transmission will experience in a
multiple unicast system
( )
G
Total M
D
Approximation of total delay experienced by multicast packets when transmitted
from a root node (s) to a multicast group (M
G
)
y Represents the maximum number of unicast routes between a source (s) and
multiple destinations
n Represents the maximum number of links a unicast route has
MU Stands for multiple unicast system
M
G
Denotes a multicast group that consists of one or more destination systems
Z Represents the size of the multicast group such as: Z =  M
G

T Represents spanning tree
D Represents total delay experienced by multicast packets when transmitted from a
root node (s) to a multicast group (M
G
)
L
n,Z
Represents the total number of links (i.e., Z
∈
R
T
) that a packet needs to traverse
in order to reach the specific destination
TABLE 1: System and Parameters Definition.
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 8
( )
(
)
( ) ( )
1 1
yn
i
i i
s d
i i
D R
AD where D R D L
y
−
= =
=
∑ ∑
(6)
where y represents the maximum number of unicast routes between a source (s) and multiple
destinations and n represents the maximum number of links a unicast route has.
3.3 Mathematical Model for a IP Multicast System
In IP multicast system, a single source (s) sends a packet to a group that consists of multiple
destination systems. In addition, a packet is sent only once by the source system where as the
intermediate routers along the route perform replications with respect to the number of
destinations a group has. Let M
G
denotes a multicast group that consists of one or more
destination systems whereas Z represents the size of the group such as Z =  M
G
. In an IP
multicast system, all multicast groups (M
G
) can be typically organized in a spanning tree. We
consider a spanning tree rooted at the multicast source (s) consisting of one of the multicast
groups (M
G
) that has a size of Z. The spanning tree can then be expressed as: T = (N
T
, M
T
)
where the numbers of destinations in one multicast group (M
G
) belong to the total number of
nodes present in the network such as: M
G
∈
M.
Also, Based on the above discussion, we can give the following hypothesis: The total delay (D)
experienced by multicast packets when transmitted from a root node (s) to a multicast group (M
G
)
can be defined as a sum of the total delay experienced by each link of a spanning tree from the
root nodes (s) to all destinations (d
∈
M
G
) and the delay experienced by each link of an
intermediate routers.
Thus, this leads us to the following expression for total delay (D
Total
) experienced by multicast
packets transmitted from root node (s) to a destination node (d) that exists within M
G
:
( )
( ) ( )
( )
1 1
GG
Z n
Total M i i
s M
i i
D D D L D L
−
= =
= = +
∑ ∑
(7)
where Z is the number of destination systems in one multicast group of a spanning tree (T) and n
represents the total number of links a route has.
Parameters Definition
ESM
G
Denotes an ESM group that consists of one or more endsystem destination
X Represents the size of the group such as X =  ESM
G

N An overlay network consists of a set of N endsystem nodes connecting
R
T
Represents a route between a source node (s) and nonleaf nodes that could be
parent or child nodes
T
T
Represents the transmission time
P
s
Represents the packet size
B
W
Denotes the bandwidth
L Represents particular link (L) between a pair of source (s) and destination (d)
D
Q
Denotes queuing delay
τ
Represents the propagation delay
L
D
Denotes distance for a communication link
SoL Represents speed of light
AB
W
Represents averagebandwidth between a pair of source (s) and destination (d)
OB
W
Represents optimalbandwidth between a pair of source (s) and destination (d)
TABLE 2: System and Parameters Definition.
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 9
The first term of (7) yields the total delay associated with the number of links with in a spanning
tree when a packet is transmitted from a root node (source) to all the leaf and nonleaf nodes.
The second term of (7) provides a total delay that a packet may experience when transmitted
along a certain route.
Equation (7) can be further generalized for one of the specific destinations (d) within a multicast
group such as d
∈
M
G
, if we assume that we have a route within a spanning tree (T) from
multicast source (s) to a specific destination (d) such as R
T
(s, d), then the multicast packets
transmitted from a source node to a destination experience a total delay of:
( )
(
)
( )
,
,
,
n Z T
G
n Z
L R s d
s d M
D D L
∈
→ ∈
=
∑
(8)
where L
n,Z
represents the total number of links (i.e., Z
∈
R
T
) that a packet needs to traverse in
order to reach the specific destination d along a path of R
T
with in the tree T as well as the
number of links from source s to a multicast group M
G
.
3.4 Mathematical Model for an End System Multicast (ESM)
Because of the limitations in IP multicast, researchers have explored an alternative architecture
named ESM, which is built on top of the unicast services with multicast functionalities. In ESM,
one of the endsystem nodes (s) participating in an application session can have the
responsibility to forward information to other hosts. Here, end users that participate in the ESM
group communicate through an overlay structure. An ESM group can have at most N endsystem
nodes where we focus on one of the endsystem nodes (s) that multicast information to the other
participating nodes of a multicast endsystem group. From the source host point of view, this
ESM group can be considered a group of destination systems. For the sake of mathematical
model, lets ESM
G
denotes an ESM group that consists of one or more endsystem destination
where as X represents the size of the group such as X =  ESM
G
. Based on the derived
expression of unicast in the previous sections, these unicast links can not exceed to M such as
1 2
, ,.........,
y
m m m M
∈
where one of the edges provides a unicast connection between two end
system nodes such as:
{
}
{
}
{
}
1 2
, ,
unicast link
m M n n s N
−
∈ → ⊂
(9)
An overlay network consists of a set of N endsystem nodes connecting though M number of
edges where one of the endsystem is designated as source host (s) such as:
{
}
, ,
G s N M
=
.
This also shows that an ESM is built on top of the unicast services using a multicast overlay
network that can be organized in a spanning tree such as T = (N
T
, M
T
) rooted as an ESM source
(s) where the numbers of destinations in one multicast group (ESM
G
) belong to the total number
of nodes present in the network such as: ESM
G
∈
M. The end receivers in a multicast tree could
be a parent or a child node depending on the location of the node.
In a multicast spanning tree (T), all the nonlead nodes can be both parent and child at the same
time where as all the leaf nodes are considered to be the child nodes. Based on the above
argument, one can say that a multicast packet originated from the root (s) of a spanning tree (T)
need to traverse typically two links; source to nonleaf node (P
n
, C
n
) and a nonleaf node to a leaf
node (C
n
). Lets R
T
(s, nonleaf node) represents a route between a source node (s) and nonleaf
nodes that could be parent or child nodes such as:
{
}
T n n G
R P C ESM= ∨ ∈
where
{
}
,
n n
P C s N
∈U
(10)
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 10
where, R
T
(P
n
, C
n
) in (10) represents a route from a parent node to a child node such as:
{
}
,
T n n G
R P C ESM
= ∈
.
Equations (9) and (10) lead us to the following expression for computing the total delay involve in
transmitting a multicast packet from a source node to one or more parent nodes (i.e., the delay
associated with the first link of transmission):
( )
{ }
( ) ( )
( )
, ,
1 1
n n n n
n n
yn
i
multiple unicast i s P C ii s P C
s P C s N i i
D D R where D R D L
−→ →
→ ∨ ∈ = =
=
∑ ∑
U
A
(11)
In (11), y is the maximum unicast routes between a source (s) and one or more nonleaf nodes
and n represents the maximum number of links a unicast route can have. Similarly, the total delay
experience by a multicast packet transmitted from a parent node to a child node can be
approximated as follows:
( ) ( )
( )
( )
( )
( )
, ,
1 1
n n n n
n n
yn
i
multiple unicast i P C i P C
P C i i
D D R where D R D L
−
→ = =
=
∑ ∑
A
(12)
By combining (11) and (12), the total delay experience by a multicast packet that transmitted from
a source node (s) to a child node (C
n
) can be approximated as:
( ) { }
( )
( )
( )
,
,
,1 1
+
n n
n n
n n
n n
multiple unicast i P C
i s P C
s C s P N i i
D D L D L
−→
→ ∈ = =
=
∑ ∑
U
(13)
4. FORMULIZATION OF BANDWIDTH EFFICIENCY FOR MULTICAST
SYSTEMS
This section presents the formulization of the bandwidth efficiency for all multicasting schemes.
For the ease of simplicity, we divide our proposed formulization for each type of multicasting
approach such as unicast, multiple unicast, IP multicast, and ESM.
4.1 Generic System Model
In order to approximate the bandwidth for Unicast, multicast, and ESM, we use the classical
definition of computing the transmission time. Based on this definition, the transmission time (T
T
)
can be defined as a product of the packet size (P
s
) which we transmit between a pair of source
(s) and destination (d) and the inverse of the bandwidth (B
W
). Mathematically, this can be
expressed as follows:
( )
1
T s W
T P
B
=
Let the value of (
(s d )
D
→
) denotes the total packetdelay which is associated with each direct
communication link. Therefore, each transmitted packet will typically experience a delay of
(s d )
D
→
on a particular link (L) between a pair of source (s) and destination (d). This delay is the
sum of the transmission time, the queuing and the propagation delays such as:
(s d )
D
→
=
Transmission Time (
T
T
) + Propagation Delay (
τ
) + Queuing Delay (
Q
D
). Also, it can be
expressed as:
(s d )
T Q
D T D
τ
→
= + +
. Changing the above expression for the transmission
delay, we got
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 11
(
)
(
)
(s d )
( )
T Q
T D D
τ
→
= − −
(14)
Recall our classical definition for the transmission time, (14) can be rewritten as the product of
bandwidth and packet size:
( )
( )
(s d )
1 ( )
s Q
W
P D D
B
τ
→
= − −
(15)
Solving (15) for approximating the bandwidth, we got
( )
( )
(s d )
( )
s
W
Q
P
B
D D
τ
→
=− −
(16)
It should be noted that the propagation delay (
τ
) is a ration between the distance for a
communication link (L
D
) and the speed of light (SoL). This allows us to further extend (16) as
follows.
(s d )
( )
s
W
D
Q
P
BL
D D
SoL
→
=
− −
(17)
Simplifying the above equation (17), we got
(
)
(
)
( )
(s d )
( ) – ( )( )
s
W
D Q
P SoL
B
SoL D L SoL D
→
=
−
(18)
For the sake of simplicity, we can ignore the queuing delay. Equation (18) can now be written as:
(
)
(
)
( )
(s d )
( ) –
s
W
D
P SoL
B
SoL D L
→
=
(19)
4.2 Bandwidth Efficiency Formulization for a Unicast System
In unicast, a packet is sent from one point (source) to another point (destination). As mentioned
earlier, when packet transmit from one source (s) to a specified destination (d), there exist
multiple routes where each route can have multiple links. This implies that the packetdelay for
unicast is entirely dependent on the number of links a packet needs to traverse in order to reach
the final destination system. Based on the above argument, one can define the packet delay such
as:
1 2
( ) ( ) ( ) ......... ( )
n
D R D L D L D L
= + + +
where n is the maximum number of links that need
to be traversed on route R between s and d.
We generalize the delay for one particular route (R) that exists between source (s) and
destination (d) such as:
1
( ) ( )
n
i
i
D R D L
=
=
∑
where
1 2
1
( ) ......
n
i n
L L L L M
= + + + ∈
∑
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 12
This expression is further extended as:
( ) ( )
( )
s d L R s d
Delay D D L
−∈ −
= =
∑
where
(
)
L R s d
∈ −
represents the value of the total delay associated with the route R between source s and
destination d. For a unicast system, taking the above expressions into account, the available
estimated bandwidth (B
W
) for a communication link (L) that exists between a pair of source (s)
and destination (d) can be approximated in the following equation:
(
)
(
)
( )
( )
( ) –
s
W
D
L R s d
P SoL
B
SoL D L L
∈ →
=
∑
(20)
The D(L) in (20) represents the link delay where as the
(
)
L R s d
∈ −
represents the value of the
total delay associated with the route R between source s and destination d.
The above equation (20) represents the approximated bandwidth which can be used by the
transmitted packet for each individual communication link between the source and destination. It
should be noted that the above equation is not representing the bandwidth approximation for one
particulate route between the source and destination. Instead, it represents the bandwidth
approximation for n number of links that need to be traversed on route R between source (s) and
destination (d).
Based on the above derivation, one can also simply derive a mathematical expression for an
averagebandwidth, denoted by AB
W
. The average bandwidth represents the available bandwidth
that each transmitted packet may utilize if it traverses one of the available routes. Equation (19)
can be modified for the average delay between a pair of source (s) and destination (d), denoted
by as follows:
(
)
(
)
( )
( )
{ }
SoL
–
s
W
D
s d
P
AB
SoL AD L
→
=
(21)
The mathematical expression for an estimated AB
W
can be derived as follows:
(
)
(
)
( )
( )
1
–
s
Wy
i
i
D
P SoL
AB
D R
SoL L
y
=
=
∑
(22)
Where
1
( ) ( )
n
i
i
D R D L
=
=
∑1 2
1
( ) ......
n
i n
L L L L M
= + + + ∈
∑
and y represents the maximum
number of possible routes between source s and destination d.
In addition to the average bandwidth, one can also choose the optimal route with respect to the
minimum bandwidth that each packet may require when traverses from one particular source (s)
to a destination (d). In order to derive an expression for the optimal bandwidth, we may need to
modify equation (19) for the optimal delay. This is due to the fact that we assume that for each
link that offers minimum bandwidth must have an optimal delay. This leads us to the following
modification of (19), such as:
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 13
(
)
(
)
( )
(s d )
( ) –
s
W
D
P SoL
OB
SoL OD L
→
=
(23)
where
(s d )
OD
→
represents the optimal delay with respect to the minimum delay that each
packet may experience when traverses from one particular source to a destination.
Based on (23), we can derive an expression for the optimal bandwidth, denoted by
W
OB
,
between a pair of source (s) and destination (d) such as:
(
)
(
)
( ) ( )
1
–
s
Wy
i D
i
P SoL
OB
SoL Min D R L
=
=
∑
(24)
where 1 2
( ) ( ) ( ) ......... ( )
n
D R D L D L D L
= + + +
and n is the maximum number of links that need
to be traversed on route R between s and d.
4.3 Bandwidth Efficiency Formulization for a Multiple Unicast System
In addition to unicast systems, we can derive the similar mathematical expressions for the
multiple unicast system where a single source (s) can transfer a packet simultaneously to multiple
destinations. In other words, in a multiple unicast system, there exist a unicast route between a
source (s) and one of the destinations. This hypothesis leads us to the following argument:
multiple routes can be established between the source (s) and each destination (d
1
, d
1
,
d
1
,…….,d
y
) where y represents the maximum number of unicast routes established in multiple
unicast. Based on this hypothesis, we can modify (19) that account the total delay such as:
( )
(
)
(
)
( )
( )
1 2
1 2
, ,......,
, ,......,
–
y
y
s
multiple
W s d d d
multiple D
s d d d
P SoL
B
SoL D L
→
→
=
(25)
The following mathematical expression can be used to estimate the total bandwidth that the entire
packet transmission utilizes in a multiple unicast system:
()
(
)
(
)
( ) ( )
, ,......,
1 2
1
–
s
s
WB y
multiple
W d d dy
i D
i
P SoL
B
SoL D R L
→
=
=
∑
(26)
Although, in multiple unicast system, a single packet can be transmitted from one source to
multiple destinations, the transmitted packet may follow a different route in order to reach the
appropriate destination. In particular, a bandwidth is always associated with links rather than the
complete routes between the source and destinations.
As a result, each transmitted packet may use a different amount of bandwidth with respect to the
number of links that the packet needs to traverse on the chosen unicast route. This implies that,
in order to estimate an average bandwidth that each packer might utilize, one should consider the
number of maximum links a unicast route has. This leads us to the following mathematical
expression for an average available bandwidth for the multiple unicast system:
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 14
(
)
(
)
( )
( )
{ }
SoL
–
s
W
D
s d
P
AB
SoL AD L
→
=
(27)
Further, solving (26) for average bandwidth approximation results (27) such as:
( )
(
)
(
)
( )
( )
( ) ( )
1
1
–
n
s
i i
y
W s d i
i
i
D
P SoL
AB where D R D L
D R
SoL L
y
→ =
=
=
∑
∑
A
(28)
where y represents the maximum number of unicast routes between a source (s) and multiple
destinations and n represents the maximum number of links that a unicast route has.
4.4 Bandwidth Efficiency Formulization for IP Multicast System
In IP multicast system, a single source (s) sends a packet to a group that consists of multiple
destination systems. In addition, a packet is sent only once by the source system where as the
intermediate routers along the route perform replications with respect to the number of
destinations a group has. Lets M
G
denotes a multicast group that consists of one or more
destination systems whereas Z represents the size of the group such as Z =  M
G
. In an IP
multicast system, in order to efficiently transmit a packet from a specific multicast source to all
multicast destinations, all multicast groups (M
G
) can be typically organized in a spanning tree (T).
For the ease of mathematical expression, we only consider a spanning tree rooted at the
multicast source (s) consisting of one of the multicast groups (M
G
) that has a size of Z.
Based on the above discussion, we describe the spanning tree such as: T = (N
T
, M
T
) rooted as
multicast source (s) where the numbers of destinations in one multicast group (M
G
) belong to the
total number of nodes present in the network such as: M
G
∈
M where M represents the total
edges that the network has. The terms N
T
and M
T
represent the vertices and the edges of the
spanning tree (T), respectively. It should be noted that we consider a spanning tree (T) that
includes only the multicast destinations of a multicast group (M
G
) with the exception of
intermediate routers. In other words, we assume that N
T
of the spanning tree (T) only consists of
one or more destination nodes. The reason for this assumption is to simplify the process of
estimating the total available bandwidth involves with the packet transmission in an IP multicast
system.
Based on the above proposed model, we can give the following hypothesis: The total available
bandwidth (TB
W
) utilizes by multicast packets when transmitted from a root node (s) to a multicast
group (M
G
) can be defined as a ration of packet size and the available bandwidth on each link of
a spanning tree (T) from the root nodes (s) to all destinations (d
∈
M
G
) with the bandwidth
available to one or more intermediate routers of each link.
The above hypothesis leads us to the following expression for total available bandwidth (TB
W
)
utilizes by multicast packets transmitted from root node (s) to a destination node (d):
(
)
(
)
( )
(s )
( ) –
G
s
W
M D
P SoL
TB
SoL D L
→
=
(29)
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 15
Computing the total delay for the IP multicast and using its resulting values in (29), we provide an
expression for the total available bandwidth such as:
( ) ( )
1 1
–
s
WZ n
D
i i
i i
P
TB
L
D L D L
SoL
= =
=
+
∑ ∑
(30)
where Z in the denominator of (30) represents the number of destination systems in one multicast
group of a spanning tree (T) where n represents the total number of links a route has.
The first term of the denominator of (30) (
( )
1
Z
i
i
D L
=
∑
)yields the total delay associated with the
number of links within a spanning tree when a packet is transmitted from a root node (source) to
all the leaf and nonleaf nodes (i.e., the multiple receivers with in T excluding the source node
(s)). In other words, according to spanning tree, if a message is transmitted from a source node
(s) to all the destination nodes, then the packet must traverse Z (i.e., the number of destinations
in one multicast group) number of links which consequently experience a delay on each link.
The second main term of the denominator (
( )
1
n
i
i
D L
=
∑
) in (30) provides a total delay that a packet
may experience when transmitted along a certain route (i.e., from a source node (s) to multicast
group (M
G
) via one or more routers along the route). The last term of the denominator can not be
considered as a design parameter and therefore its value completely depends on the type of
network.
The above equation can be further generalized for one of the specific destinations (d) within a
multicast group such as d
∈
M
G
, if we assume that we have a route within a spanning tree (T)
from multicast source (s) to a specific destination (d) such as R
T
(s, d), then the multicast packets
transmitted from a source node may utilize a total available bandwidth such as:
( )
( )
( )
–
G
G
s
W s d M
D
s d M
P
TB L
D
SoL
→ ∈
→ ∈
=
(31)
Determine the total delay for the multicast packets that transmit from a source node to one of
destinations within a group and using the resulting expression in (31), we got,
( )
(
)
(
)
( )
( )
( )
,
,
,
–
G
n Z T
s
W s d M
n Z D
L R s d
P SoL
TB
SoL D L L
→ ∈
∈
=
∑
(32)
where L
n,Z
in (32) represents the total number of links (i.e., Z
∈
R
T
) that a packet needs to
traverse in order to reach the specific destination d along a path of R
T
within the tree T as well as
the number of links from source s to a multicast group M
G
.
Since only one copy of the message is required in IP multicast, we can say that a minimum
bandwidth effort is being used for the transmission of the message to all group members
connected in the network. This minimum bandwidth is achieved due to the fact that a minimal
transmission time is required for the IP multicast which in turns reduces the overall endtoend
delay as can be seen in the denominator of (32).
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 16
4.5 Bandwidth Efficiency Formulization for End System Multicast (ESM) System
An ESM group can have at most N endsystem nodes where we focus on one of the endsystem
nodes (s) that multicast information to the other participating nodes of a multicast endsystem
group. From the source host point of view, this ESM multicast group can be considered a group
of destination systems. For the sake of mathematical model, lets ESM
G
denotes an ESM group
that consists of one or more endsystemdestinations where as X represents the size of the group
such as X =  ESM
G
.
In an overlay network, all participating endsystem nodes are fully connected to each other via the
unicast links. Based on the derived expression of unicast in the previous sections, these unicast
links that provide connection between endsystem nodes can not exceed to M such
as:
{
}
1 2
, ,.........,
y
m m m M
∈
where one of the edges (m) provides a unicast connection between
the two endsystem nodes such as:
{
}
{
}
{
}
1 2
, ,
unicast link
m M n n s N
−
∈ → ⊂
.
The structure of the ESM is an overlay network in a sense that each of the paths between the
endsystems corresponds to a unicast path. This implies that an overlay network consisting of a
set of N endsystem nodes connecting though M number of edges where one of the endsystem
is designated as a source host (s) can be expressed as:
{
}
, ,
G s N M
=
.
This also shows that an ESM is built on top of the unicast services using a multicast overlay
network that can be organized in a spanning tree such as T = (N
T
, M
T
) rooted as an ESM source
(s) where the numbers of destinations in one multicast group (ESM
G
) belong to the total number
of nodes present in the network such as: ESM
G
∈
M. Here, the membership and replication
functionality is performed by the ESM receivers, which connect together over unicast channels to
form a multicast tree (T), rooted at one data source (s). The end receivers (i.e., the number of
endsystems in ESM
G
) in a multicast tree could be a parent or a child node depending on the
location of the node. In a multicast spanning tree (T), all the nonleaf nodes can be both parent
and child at the same time where as all the leaf nodes are considered to be the child nodes. In
other words, the parent nodes perform the membership and replication process where as the
children nodes are receivers that receives multicast packet from the parent nodes.
Based on the above argument, one can say that a multicast packet originated from the root (s) of
a spanning tree (T) need to traverse typically two links; source to nonleaf node (P
n
, C
n
) and a
nonleaf node to a leaf node (C
n
). Lets R
T
(s, nonleaf node) represents a route between a source
node (s) and nonleaf nodes that could be parent or child nodes such as:
{
}
T n n G
R P C ESM= ∨ ∈
where
{
}
,
n n
P C s N
∈U
Similarly, R
T
(P
n
, C
n
) represents a route from a parent node to a child node such as:
{
}
,
T n n G
R P C ESM
= ∈
. The above arguments lead us to the following expression for computing
the total bandwidth available for transmitting a multicast packet from a source node to one or
more parent nodes (i.e., the bandwidth associated with the first link of transmission):
( )
{ }
( )
{ }
–
n n
n n
s
multiple unicast
W s P C s N
D
multiple unicast
s P C s N
P
TB
L
D
SoL
−
→ ∨ ∈
−
→ ∨ ∈
=
U
U
(33)
Determining the mathematical expression for the total delay involve in transmitting a multicast
packet from a source node to one or more parent nodes and using the resulting expression in
(33) yields (34) for approximating the total bandwidth such as:
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 17
( )
{ }
( )
( )
( )
,
1
,1
–
n n
n n
n n
s
multiple unicast y
W s P C s N
D
i s P C
i
n
i
ii s P C i
P
TB
L
D R
SoL
where D R D L
−
→ ∨ ∈
→
=
→ =
=
∑
∑
U
A
(34)
where y in (34) represents the maximum unicast routes between a source (s) and one or more
nonleaf nodes and n represents the maximum number of links a unicast route can have.
Based on (34), we can derive a similar mathematical expression for approximating the total
bandwidth available for a multicast packet which is transmitted from a parent node to a child node
as follows:
( )
( )
( )
( )
( )
( )
,
1
,
1
–
n n
n n
n n
n
s
i
multiple unicast i P C
y
W P C i
D
i P C
i
P
TB where D R D L
L
D R SoL
−
→ =
=
=
∑
∑
A
(35)
When comparing (34) with (35), it can be clearly evident that only the total delay component in
the denominator of both equations differ with each other for both presented scenarios which in
turns change the amount of available total bandwidth. By combining the above two equations, the
total bandwidth that a multicast packet may utilize when transmitted from a source node (s) to a
child node (C
n
) can be approximated as follows:
( ) { }
(
)
(
)
( )
( )
( )
( )
,
,
,
1 1
+ –
n n
n n
n n
s
multiple unicast n n
W s C s P N
D
i P C
i s P C
i i
P SoL
TB
D L D L SoL L
−
→ ∈
→
= =
=
∑ ∑
U
(36)
As one can see in (36) that the hosts are able to multicast information and consequently use the
same link to redirect the packets resulting in an increase in the endtoend delay of the entire
transmission process. The limitation in bandwidth and the fact that the message needs to be
forwarded from hosttohost using unicast connection, and consequently incrementing the endto
end delay of the transmission process as can be seen in the denominator of (36), contribute to
the price to pay for this new approach. For the ESM, this price will be paid in terms of the second
quantity of the denominator of (36). (i.e.,
( )
( )
,
1n n
n
i P C
i
D L
=
∑
). These reasons might make ESM
slightly less efficient than the IP multicast system.
5. PERFORMANCE ANALYSIS AND EXPERIMENTAL VERIFICATIONS
In this section, we provide the simulation results for analyzing the performance of different
multicast schemes with respect to the delay analysis and the bandwidth approximation.
5.1 System Model
Simulations are performed using OPNET to examine the performance of Multiple unicast, IP
multicast, and ESM schemes. Figure 4 shows an OPNET model for the multiple unicast, IP
multicast and ESM simulations. The OPNET simulation has run for a period of 900 seconds for all
three scenarios where we collect the simulated data typically after each 300 seconds. For all
scenarios, we have setup one sender node that transmits video conferencing data at the rate of
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 18
10 frames/s using 2,500stream packet size to one or more potential receivers via a link that
operates at 100 Mbps. In addition to these 100 Mbps licks, we use separate DS3 links for the
core network (Internet). The same traffic pattern is assumed for all scenarios.
It should be noted in Fig. 4 that we use four backbone routers that connect multiple subnets to
represent Bay Networks concentrator topology using ATM – Ethernet FDDI technology. In order
to generate consistent simulation results, we use the same topology for the first two scenarios
with some minor exceptions. For Multiple unicast, we disable the multicast capabilities of
backbone routers where as for the IP multicast this restriction does not impose. Finally, in order to
examine the behavior of the ESM, we use an OPNET Custom Application tool that generates the
overlay links and the source root.
5.2 Experimental Verifications
For the Multiple unicast scenarios, video conferencing data is being sent by the root sender at the
rate of 25 Kbytes per second. This implies that a total of three copies traveling which result in 75
Kbytes per second of total traffic. The last mile bandwidth limitation typically provides the most
important delay impact. OPNET collected all the delays for all the receivers and calculated the
average. The packet endtoend delay for multiple unicast was 0.0202 seconds. For the IP
multicast approach only one copy of the packet is generated at the root source. For this reason,
the total videoconferencing traffic sent and received is only 25,000 bytes/s. Thus, a better
performance in the average packet endtoend delay can be observed. This is approximately
0.0171 seconds. Finally, after performing ESM simulations, we obtain an average endtoend
delay packet of 0.0177 seconds.
It can be seen in Fig. 5 that ESM packets transmission provides comparatively good performance
than the Unicast but not as good as the IP multicast. The reasons are the RDP (Relative Delay
Figure 4: OPNET Model for Multiple Unicast, IP Multicast and EndSystem Multicast Video Conferencing
Transmissions.
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 19
Penalty or the ratio of the delay between the sender and the receiver) [6] and the LS (Link Stress
or the number of identical copies of a packet carried by a physical link) experienced by each
network schemes. Even though, a Unicast scheme provides comparatively low RDP, the value of
LS is not optimal.
On the other hand, IP multicast performs with a little bit higher RDP but it gets a better LS. ESM
has the inconvenience of RDP higher than IP multicast due to the fact that for a second receiver,
there is an increasing delivery–delay because of the enduser replication (the second user has to
wait for the data sent by its father node or subserver). This is the penalty that ESM has to pay.
One possible solution would be the design of a robust multicast protocol to optimize the delivery
of data for the final users. Note that the additional delay could be reduced if we optimize the
bandwidth utilization in the potential parent nodes. This is not a simple task because it requires a
smart protocol to recognize bandwidth limitations in potential parent nodes and to establish an
algorithm to limit the number of children nodes for these parent nodes.
6. CONSLUSION & FUTURE WORK
In this paper, we presented both analytical and mathematical models for all the multicast
approaches currently available for multimedia applications. We first presented a mathematical
model for multiple unicast systems in which the source host has to send a single copy of data to
every single receiver. Our proposed formulization suggested that this approach wastes a lot of
significant network bandwidth. Secondly, we presented a mathematical model for the IP
multicasting approach which is a more efficient concept where the data source only sends one
copy of data which is replicated as necessary when propagating over the network towards the
receivers. Our proposed formulization shows that the IP multicast demonstrates some good
bandwidth efficiency characteristics than the other multicast approaches. Finally, we presented a
complete formulization of bandwidth efficiency for ESM systems, which is an alternate to router
dependent multicast service that allows endsystems participating in a multicast group to replicate
and forward packets to other receivers. Our proposed formulization of bandwidth efficiency
suggests that the ESM is a feasible, especially for sparse, medium size group. The simulation
results presented in this paper fully support the proposed analytical model for the IP multicast and
ESM.
Figure 5: Average endtoend packet delays for multiple unicast, IP multicast and ESM simulations.
Syed S. Rizvi, Aasia Riasat, & Khaled M. Elleithy
International Journal of Computer Networks (IJCN), Volume (1): Issue (1) 20
7. REFERENCES
1.
Y. Ma, R. Chbeir, and K. Yetongnon. “An improved clusterbased P2P multicasting
approach”. Proceedings of the 2nd international conference on Scalable information systems,
Vol. 304, No. 59, June 2007.
2.
D. Kostić, A. Snoeren, A. Vahdat, R. Braud, C. Killian, J. Anderson, J. Albrecht, A. Rodriguez,
E. Vandekieft. “Highbandwidth data dissemination for largescale distributed systems”.
Transactions on Computer Systems (TOCS), 26(1), February 2008.
3.
S. Fahmy and M. Kwon. “Characterizing overlay multicast networks and their costs.”
IEEE/ACM Transactions on Networking (TON), 15(2):373 – 386, April 2007.
4.
S. Rizvi, A. Riasat, and M. Elleithy. “Deterministic formulization of endtoend delay for
multicast systems.”
Accepted in International journal of Computer Networks &
Communications, Vol. 2, No.1, 2010.
5.
H. Holbroo and D. Cheriton. “IP multicast channels: EXPRESS support for largescale single
source applications.” In Proceedings of SIGCOMM, 1999.
6.
B. Fenner, M. Handley, H. Holbrook and I. Kouvelas. “Protocol independent multicastsparse
mode (PIMSM): protocol specification (Revised).” IETF Internet Draft, work in progress,
March 2003, draftietfpimsmv2new07.txt
7.
D. Kosiur. “IP Multicasting: The complete guide to interactive corporate networks.” New York:
John Wiley & Sons, Inc., 1998.
8.
J. Saltzer, D. Reed, and D. Clark. “EndtoEnd arguments in system design.” ACM
Transactions on Computer Systems, 2(4):195–206, 1984.
9.
Y. Chua, S.Rao, S.Sechan and H. Zhang. “Enabling conferencing applications on the internet
using an overlay multicast architecture.” Proceedings of the ACM SIGCOMM’01, San Diego,
CA, August 2001.
10.
S. Deering. “Multicast routing in internetworks and extended LANs.” Proceedings of the ACM
SIGCOMM 88, pp. 55–64, Stanford, CA, August 1988.
11.
Y. Chu, S. Rao, and H. Zhang. “A case for end system multicast.” Proceedings of ACM
Sigmetrics, June 2000.
12.
S. Y. Shi and J.S. Turner. “Multicast routing and bandwidth dimensioning in overlay
networks.” IEEE Journal on Selected Areas in Communications (JSAC), 20(8):1444–1455,
October 2002.
13.
S. Shi. “Design of overlay networks for Internet multicast.” PhD dissertation, Washington
University in St. Louis, August 2002.
14.
M. Blumenthal and D. Clark. “Rethinking the design of the Internet: The endtoend argument
vs. the brave new world.” ACM Transactions on Internet Technology, 1(1): 70–109, August
2001.
15.
M. Kwon and S. Fahmy. “Topologyaware overlay networks for group communication.”
Proceedings of the 12th international workshop on Network and operating systems support
for digital audio and video, pp. 127 – 136, May 2002.