网络仿真实验:基于NS2的组播路由协议实现及仿真

1.实验内容

 定义一个20个节点的IP网络拓扑图，仿真PIM-SM组播协议。组播源首先注册，在1.0s源节点10开始向RP节点0发送数据包，1.5节点4发送graft消息给RP节点0，2.0s节点5发送graft消息给RP节点0，2.5s节点11发送graft消息给RP节点0，3.0s节点12发送graft消息给RP节点0，3.5s节点15发送graft消息给RP节点0，4.0s节点16发送graft消息给RP节点0，5.0s整个通信仿真结束。在整个仿真过程中，链路带宽均为10Mb，链路延时10ms，队列丢包类型为DropTail，数据流包大小是210bytes，发送速率是448kbps。

运行环境：

操作系统：Red Hat Linux 9.0
网络仿真软件：NS2 2.28
数据处理工具：AWK
绘图软件：gnuplot

网络拓扑图：

2.组播tcl程序

#使节点支持组播set ns [new Simulator -multicast on]set group [Node allocaddr]#生成trace文件和nam文件set ft [open multicast.tr w]$ns trace-all $ftset fn [open multicast.nam w]$ns namtrace-all $fnproc finish {} { global ns ft fn $ns flush-trace close $ft close $fn exec nam multicast.nam & exit 0}#创建网络拓扑结构for {set i 0} {$i<20} {incr i} { set n($i) [$ns node]}$ns duplex-link $n(0) $n(1) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(2) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(7) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(11) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(14) 10Mb 10ms DropTail$ns duplex-link $n(1) $n(15) 10Mb 10ms DropTail$ns duplex-link $n(1) $n(19) 10Mb 10ms DropTail$ns duplex-link $n(2) $n(3) 10Mb 10ms DropTail$ns duplex-link $n(2) $n(4) 10Mb 10ms DropTail$ns duplex-link $n(3) $n(5) 10Mb 10ms DropTail$ns duplex-link $n(4) $n(5) 10Mb 10ms DropTail$ns duplex-link $n(5) $n(6) 10Mb 10ms DropTail$ns duplex-link $n(7) $n(8) 10Mb 10ms DropTail$ns duplex-link $n(7) $n(9) 10Mb 10ms DropTail$ns duplex-link $n(8) $n(10) 10Mb 10ms DropTail$ns duplex-link $n(9) $n(10) 10Mb 10ms DropTail$ns duplex-link $n(11) $n(12) 10Mb 10ms DropTail$ns duplex-link $n(11) $n(13) 10Mb 10ms DropTail$ns duplex-link $n(15) $n(16) 10Mb 10ms DropTail$ns duplex-link $n(15) $n(18) 10Mb 10ms DropTail$ns duplex-link $n(16) $n(17) 10Mb 10ms DropTail$ns duplex-link $n(17) $n(18) 10Mb 10ms DropTail#设置组播路由策略（BST模式）$ns mrtproto BSTBST set RP_($group) $n(0)#（或者DM模式）#DM set CacheMissMode dvmrp#set mproto DM#set mrthandle [$ns mrtproto $mproto]#创建业务流set udp1 [new Agent/UDP]$ns attach-agent $n(10) $udp1$udp1 set dst_addr_ $group$udp1 set dst_port_ 0set src1 [new Application/Traffic/CBR]$src1 set packetSize_ 210$src1 set rate_ 448k$src1 attach-agent $udp1set rcvr [new Agent/LossMonitor]#创建触发事件$ns at 1.0 "$src1 start"$ns at 1.5 "$n(4) join-group $rcvr $group"$ns at 2.0 "$n(5) join-group $rcvr $group"$ns at 2.5 "$n(11) join-group $rcvr $group"$ns at 3.0 "$n(12) join-group $rcvr $group"$ns at 3.5 "$n(15) join-group $rcvr $group"$ns at 4.0 "$n(16) join-group $rcvr $group"$ns at 5.0 "$src1 stop"$ns at 5.0 "finish"$ns run

其中，使用ST组播策略时老是提示不能识别的变量名称，可能是NS2的版本所致（早期版本应该可以）

节点10到节点5的延迟-awk程序

BEGIN { highest_packet_id=0;}{action=$1;time=$2;tonode=$4;packet_id=$12;#记录当前最大的数据包idif(packet_id>highest_packet_id) highest_packet_id=packet_id;#记录数据包的开始发送时间if(start_time[packet_id]==0) start_time[packet_id]=time;#记录数据包的接收时间if(action=="r" && tonode==5) end_time[packet_id]=time;}END {#统计每个数据包的开始时间和延迟for(packet_id=0;packet_id<=highest_packet_id;packet_id++) { packet_duration=end_time[packet_id]-start_time[packet_id]; if(packet_duration>0 ) { printf("%f %f/n",start_time[packet_id],packet_duration); }}}

组播源业务流发送量-awk程序

BEGIN {init=0;i=0;}{action=$1;time=$2;fromnode=$3;pktsize=$6;if(action=="+" && fromnode==10) { pkt_byte_sum[i+1]=pkt_byte_sum[i]+pktsize; #业务数据开始发送时间 if(init==0) { start_time=time; init=1; } #当前截至时间 end_time[i]=time; i++;}}END {#统计随时间变化的业务发送量for(j=1;j<i;j++) { sendsize=pkt_byte_sum[j]/(end_time[j]-start_time)*8/1000;#kbps printf("%f %f/n",end_time[j],sendsize);}}

节点7到节点0的链路吞吐量-awk程序

BEGIN {#业务开始时间start_time=1.0;i=0;}{action=$1;time=$2;fromnode=$3;tonode=$4;pktsize=$6;if(action=="r" && fromnode==7 && tonode==0) { #截至当前时间从节点7流向节点0的数据量 pkt_byte_sum[i+1]=pkt_byte_sum[i]+pktsize; #当前截至时间 end_time[i+1]=time; i++;}}END {#统计随时间变化的链路吞吐量for(j=1;j<=i;j++) { throughput=pkt_byte_sum[j]/(end_time[j]-start_time)*8/1000;#kbps printf("%f %f/n",end_time[j],throughput);}}

组播汇合节点0的负载改变情况-awk程序

BEGIN {init=0;i=0;}{action=$1;time=$2;fromnode=$3;pktsize=$6;#此计算组播节点0负载变化情况的方法感觉不很理想，呵呵if(action=="-" && fromnode==0) { pkt_byte_sum[i+1]=pkt_byte_sum[i]+pktsize; if(init==0) { start_time=time; init=1; } end_time[i]=time; i++;}}END {for(j=1;j<i;j++) { sendsize=pkt_byte_sum[j]/(end_time[j]-start_time)*8/1000;#kbps printf("%f %f/n",end_time[j],sendsize);}}

3.单播tcl程序

#比较简单的单播tcl程序set ns [new Simulator]set ft [open unicast.tr w]$ns trace-all $ftset fn [open unicast.nam w]$ns namtrace-all $fnproc finish {} { global ns ft fn $ns flush-trace close $ft close $fn exec nam unicast.nam & exit 0}#创建网络拓扑结构for {set i 0} {$i<20} {incr i} { set n($i) [$ns node]}$ns duplex-link $n(0) $n(1) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(2) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(7) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(11) 10Mb 10ms DropTail$ns duplex-link $n(0) $n(14) 10Mb 10ms DropTail$ns duplex-link $n(1) $n(15) 10Mb 10ms DropTail$ns duplex-link $n(1) $n(19) 10Mb 10ms DropTail$ns duplex-link $n(2) $n(3) 10Mb 10ms DropTail$ns duplex-link $n(2) $n(4) 10Mb 10ms DropTail$ns duplex-link $n(3) $n(5) 10Mb 10ms DropTail$ns duplex-link $n(4) $n(5) 10Mb 10ms DropTail$ns duplex-link $n(5) $n(6) 10Mb 10ms DropTail$ns duplex-link $n(7) $n(8) 10Mb 10ms DropTail$ns duplex-link $n(7) $n(9) 10Mb 10ms DropTail$ns duplex-link $n(8) $n(10) 10Mb 10ms DropTail$ns duplex-link $n(9) $n(10) 10Mb 10ms DropTail$ns duplex-link $n(11) $n(12) 10Mb 10ms DropTail$ns duplex-link $n(11) $n(13) 10Mb 10ms DropTail$ns duplex-link $n(15) $n(16) 10Mb 10ms DropTail$ns duplex-link $n(15) $n(18) 10Mb 10ms DropTail$ns duplex-link $n(16) $n(17) 10Mb 10ms DropTail$ns duplex-link $n(17) $n(18) 10Mb 10ms DropTail#创建业务流set udp4 [new Agent/UDP]$ns attach-agent $n(10) $udp4set src4 [new Application/Traffic/CBR]$src4 set packetSize_ 210$src4 set rate_ 448k$src4 attach-agent $udp4set rcvr4 [new Agent/Null]$ns attach-agent $n(4) $rcvr4$ns connect $udp4 $rcvr4set udp5 [new Agent/UDP]$ns attach-agent $n(10) $udp5set src5 [new Application/Traffic/CBR]$src5 set packetSize_ 210$src5 set rate_ 448k$src5 attach-agent $udp5set rcvr5 [new Agent/Null]$ns attach-agent $n(5) $rcvr5$ns connect $udp5 $rcvr5set udp11 [new Agent/UDP]$ns attach-agent $n(10) $udp11set src11 [new Application/Traffic/CBR]$src11 set packetSize_ 210$src11 set rate_ 448k$src11 attach-agent $udp11set rcvr11 [new Agent/Null]$ns attach-agent $n(11) $rcvr11$ns connect $udp11 $rcvr11set udp12 [new Agent/UDP]$ns attach-agent $n(10) $udp12set src12 [new Application/Traffic/CBR]$src12 set packetSize_ 210$src12 set rate_ 448k$src12 attach-agent $udp12set rcvr12 [new Agent/Null]$ns attach-agent $n(12) $rcvr12$ns connect $udp12 $rcvr12set udp15 [new Agent/UDP]$ns attach-agent $n(10) $udp15set src15 [new Application/Traffic/CBR]$src15 set packetSize_ 210$src15 set rate_ 448k$src15 attach-agent $udp15set rcvr15 [new Agent/Null]$ns attach-agent $n(15) $rcvr15$ns connect $udp15 $rcvr15set udp16 [new Agent/UDP]$ns attach-agent $n(10) $udp16set src16 [new Application/Traffic/CBR]$src16 set packetSize_ 210$src16 set rate_ 448k$src16 attach-agent $udp16set rcvr16 [new Agent/Null]$ns attach-agent $n(16) $rcvr16$ns connect $udp16 $rcvr16#创建触发事件$ns at 1.5 "$src4 start"$ns at 2.0 "$src5 start"$ns at 2.5 "$src11 start"$ns at 3.0 "$src12 start"$ns at 3.5 "$src15 start"$ns at 4.0 "$src16 start"$ns at 5.0 "$src4 stop"$ns at 5.0 "$src5 stop"$ns at 5.0 "$src11 stop"$ns at 5.0 "$src12 stop"$ns at 5.0 "$src15 stop"$ns at 5.0 "$src16 stop"$ns at 5.0 "finish"$ns run

对单播trace文件的数据分析程序（awk程序）和组播的类似

4.用gnuplot工具（或Xgraph）画出分析曲线图

都可以设置自动输出png格式的文件

数据源（节点10）业务流发送量对比图

节点10到节点5延迟对比图

节点7到节点0链路吞吐量对比图

组播汇合节点0负载改变情况

5.实验参考的文章

计算机仿真：基于NS2的组播路由协议实现及仿真

 

网上参考的资源：

http://hpds.ee.ncku.edu.tw/~smallko/ns2/ns2.htm  NS2 使用說明手冊-柯志楠教授
http://140.116.72.80/~smallko/ns2/nsBench.htm    nsBench简介
http://www.mnlab.cs.depaul.edu/projects/nsbench/
http://www.isi.edu/nsnam/ 

http://nsnam.isi.edu/nsnam/index.php/Main_Page