This paper evaluate the robustness of multicast routing and the performance evaluation of multicast service using the experimental network.  For the performance evaluation of multucast service, we use the the DV (Digital Video)  transmission service, as the practical application for the next generation Internet.  The evaluatoin system uses the PIM (Protocol Independent Multicast), that is an RPF (Reverse Path Forwarding) multicast routing protocol.  The packet loss observed at the end host is evaluated, compared with the
case where the system provide the multicast service using N of the multiple unicast packet transmission.   We can make sure the multicast system has better system scalability, regarding the number of receivers, than the unicast based multicast system has.  PIM operates correctly and stably, even when the network has a routing loop. Finally, we operate the DV multicast service over the nation-wide JB project IPv6 network using the PIM-SM, to make sure the correct operation.

1. Introduction

The most of the data communications over the currently operating Internet is using the unicast data transmission with TCP or
UDP.  And, many exisiting applications do transfer exactly the same data to large number of receivers.   Also, the upcomming Internet applications will includes a realtime multicast services, such as DV (Digital Video) program multicasting service or interactive multi-party multimedia conferences or games.  We have to establish a stable and cost effective multicast data transmission infrastructure for the next generation Internet applications.
    With a multicast data transmission, when a sender transfers a single packet toward the particular multicast group, the network copies the packet to transfer them only to the particularreceivers belonging to the corresponding multicast group.  Since the sender does not need to send the same data to every receivers, the network can provide an efficient multicast data transmission services.   With the multicast data transmission, the required processing power at the sender and the required bandwidth for the sender can be smaller than those with the unicast data transmission.  Since the total number of packtes transferred by the network with multicast data transmission is smaller than that with unicast data transmission, the possibility of inappropriate routing protocol operation due to the large amount of user packet transmission with multicast data transmission would be less than that with unicast data transmission. Also, the diversity of the data reception time by the receivers must be smaller that with the unicast data transmission, and the data reception delay by the revievers must be smaller than that with the unicast transmission.
    On the other hand, the multicast service may have the following technical concernings.

•  the performance of packet transmission at the router may degrade, due to the copying of multicast packets.  This is because the copying the received packet at the router is a new function for the router.
•  the network congestion due to the routing loop with multicast packet will be worse than that with unicast packet.   This is because the number of links and nodes affected by the routing loop with multicast packet transmission should be larger than that with unicast packet transmission.

• Digital Video (DV) multicasting

DV is one of the popular applications with high speed data transmission used in the next generation Internet.  Also, the DV is an application, that is sensitive to the  packet loss and the delay jitter.
• PC based router and host

For flexible and cost effective network operation, the experimental system uses the conventional PCs for routers and hosts.
• PIM-SM multicast routing protocol [8]

For effective and scalable multicast service, we used the PIM-SM (Protocol Independent Multicast - Sparse Mode) routing protocol.
• Native IPv6 network [9]

Since the IPv6 technology is the fundamental core protocol for the upcomming Internet, the experimental system operated over the IPv6 network.

2. Performance Evaluation of Packet Transmission in Multicast Network
2.1 Evaluation System

Figures 1 and 2 show the evaluation system.  Sender (Snd1) sends the packets to two receiver hosts (Rcv1 and Rcv2) through the routers (R1 and R2).  Rcv1 and Rcv2 receive the same data from Snd1, i.e., multicating from Snd1 to Rcv1 and Rcv2.  R1 and R2 are connected through the OC-3 ATM link, and the hosts (Snd1, Rcv1 and Rcv2) are connected to the routers through the 100MBase-T links.

In the configuration of figure 1, Snd1 estblish two (unicast) packet flows to each receiver host.  This is the case where the multicast service is provided by the multiple unicast connections. Routers (R1and R2) do not need to copy the received data, and the sender
host (Snd1) copies the sending data.  In the configuration of figure 2, Snd1 establish one (multicast) packet flow to the receiver hosts.  This is the case where the multicast service is provided by a single multicast connection. The router R2 copies the received data to deliver the data to receivers (Rcv1 and Rcv2). Here, with the configuration of figure 2, PIM-SM is applied to as the multicast routing.  Sender host (Snd1) can control the packet transmission rate to the network, with the rate control (i.e., shaping).
    All nodes, that are routers and hosts, are the ordinary IBM compatible PC with the following specification.
 

• Host

- CPU : Intel Pentium II 450 MHz
- Memory : 128MB
- NIC : Intel EtherExpress Pro 100
- OS : FreeBSD 3.4 with KAME 19990908-stable
 
• Router

- CPU : Intel Pentium III 450MHz
- Memory : 128MB
- NIC : Intel EtherExpress Pro 100
          ENI OC-3 256MB ATM
- OS : FreeBSD 3.4 with KAME 19990908-stable

Table 1 and figure 3 show the number of recieved packets and the number of dropped packets at the router (R1), parameterizing the packet transmission rate from the sender host (Snd1).

    As shown, the router can not relay the unicast packets, in accordance with the increase of packet transmission rate from the sender with unicast based multicast.  However, with the multiast service, the router can relay the packets and the packet dropped rate does not increase even when the packet transmissionrate from the sender host (Snd1) increases.

3. Evaluation of System Robustness of Multicast Routing Protocol
3.1 Overview of RPF Multicast Routing

    The RPF represents Reverse Path Forwarding.  RPF is commonly used in the major multicast routing protocols, such as PIM that is used in the evaluation system discussed in this paper.  In RPF system, the multiast packet transmission is executed using the unicast routing information.  The multicast session is recognized and managed by the pair of sourece node IP address ("S") and IP multicast address assigned for the multucast group ("G"), i.e., {S,G}.  When the router receives the multicast packet, the router checks the source IP address in it.

if (Recv_I/F = Nxt_Hop_R_I/F)
  {forward the received packet to the rest of interfaces}
else
  {silently discard the received packet}
endif

% Here,
%  (1) the Recv_I/F : the interface (including the virtual interface such as ATM or Frame Relay),
%      where receives the multicast.
%  (2) Nxt_Hop_R_I/F : the interface, where the packet to the source IP address shown in the received
        multicast packet should relay based on the unicast routing information

3.2 Evaluation System

Figure 4 shows the evaluation system.  Sender (Snd1) sends the multicast packets toward two receiver hosts (Rcv1 and Rcv2) through the routers (R1, R2 and R3).  All routers run the PIM-SM multicast routing protocol.  All the data links in the system are 100MBase-T.
 

All nodes, that are routers and hosts, are the ordinary IBM compatible PC with the following specification.

• Host

- CPU : Intel Pentium II 450 MHz
- Memory : 128MB
- NIC : Intel EtherExpress Pro 100
- OS : FreeBSD 3.4 with KAME 19990908-stable
 
• Router

- CPU : Intel Pentium III 450MHz
- Memory : 128MB
- NIC : Intel EtherExpress Pro 100
          ENI OC-3 256MB ATM
- OS : FreeBSD 3.4 with KAME 19990908-stable

3.3 Evaluation Results
     The multicast packets are transferred to the first hop router R1.  R1 has to discard the all the received multicast IP packet from the Snd1 node, since the interface receiving the IP multicast packet is different from the interface RPF expected.   We manually modify the routing information at R1, so that the evaluation system has a routing loop among R1, R2, and R3.  By the applicatin of the RPF mechanism, it is expected that the R1 silently drop the multicast packets.
    The experimental system actually discard the IP multicast packets at the R1, when the R1 has the wrong unicast routing information to form a routing loop.  When we use the multiple unicast connection to provide the multicast service from Snd1 to Recv1 and Rcv2, the unicast packets do loop among the R1, R2 and R3.  Moreover, the packet to be looped shall be two packet flows (both toward Recv1 and toward Rcv2).  We did make sure the routing loop has generate the congestion with unicast based multicast service.  And, also, we did make sure the avoid of packet transmission at the entry router of the (unicast) routing loop, to avoid the network congestion due to the unicast routing loop.

4. Digital Video Transmission over the Nation-wide JB Project Network
4.1 Network Configuration
    We evaluate our developped multicast system over the nation-wide JB project network.  The JB network [1] is the jointly operated among WIDE project [2], CKP[3] and ITRC[4].  Most of the high speed links are provided by the JGN (Japan Gigabit Network) operated by TAO[5], by the TTNet [6] and by the CRL[7].  It uses IPv6 as the basic internet protocol.  Figure 5 shows the network configuration of the JB network, regarding only the high speed links.  More than ten organizations are interconnected by the high speed datalinks (e.g., OC-12 ATM, Gigabit Ethernet).   All organizations are interconnected thtough the PC-based routers with KAME IPv6 protocol stack, and run the PIM-SM for multicast service.  We had two RP (Rendezvous Point) routers in the network, to create shared multicast trees.
 

4.2 Experiment of DV Multicast Service over the JB Project Network
    Figure 6 shows the verview of DV multicast service over the JB project IPv6 network.  The technical workshop held by the WIDE project is multicasted to more than ten sites over the nation-wide JB project network, using the DV video stream.  The DV stream is multicasted to the sites using the PIM-SM with IPv6.  Each site observes the same DV image, simultaneously.  And, when the participants from the remore site can interactively join to the workshop using the networ.  This means that the remote site can send the DV stream, whenever they have a question.   The DV stream from the remote site also multicasted to all the participating sites to the multicast service, through the RP routers.  The experimantal interactive milticast session for one-day workshop using the DV multicasting has been correctly operated using the PIM-SM.   
 

5.Conclusion
    This paper evaluate the robustness of multicast routing and the performance evaluation of multicast service using the experimental network.  For the performance evaluation of multucast service, we use the the DV (Digital Video)  transmission service, as the practical application for the next generation Internet.  The evaluatoin system uses the PIM (Protocol Independent Multicast), that is an RPF (Reverse Path Forwarding) multicast routing protocol.  The packet loss observed at the end host is evaluated, compared with the
case where the system provide the multicast service using N of the multiple unicast packet transmission.   We can make sure the multicast system has better system scalability, regarding the number of receivers, than the unicast based multicast system has.  PIM operates correctly and stably, even when the network has a routing loop. Finally, we operate the DV multicast service over the nation-wide JB project IPv6 network using the PIM-SM, to make sure the correct operation.
    The following items would be further study items.

• The jitter of packet transmission delay observed by each recievers with the multicast transmission might be larger than that with the unicast transmission. Some applications, such as a video transmission, are sensitive to the delay jitter. The jitter of packet transmission delay should be evaluated.  The evaluation should include the case where we use the actual delay jitter sentitive applications (e.g., DV transmission). 
• The robustness and stability of multicast routing protocol against the high offered load by the user packet flows.  The routing protocol should correctly operate even when the network is in congestion.
• We should evaluate the effectiveness of the packet scheduling mechanism to provide different quality of service among system control information (e.g., routing protocol information) and user information.