Throughput is the ratio of successful packet received at the receiver to number of packet generated.

End-to-end delay is total delay between sender to receiver.

http://www.icir.org/models/tools.html

# AWK Script for calculating:

# => Average End-to-End Delay.

# ===================================================================

BEGIN {

seqno = -1;

# droppedPackets = 0;

# receivedPackets = 0;

count = 0;

}

{

if($4 == "AGT" && $1 == "s" && seqno < $6) {

seqno = $6;

}

# else if(($4 == "AGT") && ($1 == "r")) {

# receivedPackets++;

# } else if ($1 == "D" && $7 == "tcp" && $8 > 512){

# droppedPackets++;

# }

#end-to-end delay

if($4 == "AGT" && $1 == "s") {

start_time[$6] = $2;

} else if(($7 == "tcp") && ($1 == "r")) {

end_time[$6] = $2;

} else if($1 == "D" && $7 == "tcp") {

end_time[$6] = -1;

}

}

END {

for(i=0; i 0) {

delay[i] = end_time[i] - start_time[i];

count++;

}

else

{

delay[i] = -1;

}

}

for(i=0; i 0) {

n_to_n_delay = n_to_n_delay + delay[i];

}

}

n_to_n_delay = n_to_n_delay/count;

print "\n";

# print "GeneratedPackets = " seqno+1;

# print "ReceivedPackets = " receivedPackets;

# print "Packet Delivery Ratio = " receivedPackets/(seqno+1)*100

#"%";

# print "Total Dropped Packets = " droppedPackets;

print "Average End-to-End Delay = " n_to_n_delay * 1000 " ms";

print "\n";

}

just work out to tune to  ur trace file you will get the results

Throughput is nothing but how many packets are delivered to the receiver in good condition. Actually it is the ratio between number of packets generated at the transmitting end to the number os successful packets received at the receiver. Normally it is expressed in percentage say throughput is 50% etc.

For measuring the delay you need to know how much time it takes for a packet to move from transmitting end to receiver end. This delay will not be constant always. So, we need to note down the readings for several packets and then we can find the Average delay for the network or for that channel.

Jitter is delay variation. Ie all packet will not have the same delay. It varies with packets. This variation in delay is termed as Jitter.

Really thanks a lot Dear Mr. Gyanappa Walikar and Chandramathi Soundararajan 

In fact throughput is the measurement of number of bits received per second and it is measured in kbps/mbps where as the ratio of number packets received to number of packets transmitted is called packet delivery ratio which is measured in %.

Packet  delivery ratio is again the throughput right. How many pakts are transmitted and how many are received successfully?. Of course packet size is Kbps or Mbps.

Of course packet delivery ratio corresponds to throughput only but some authors prefers to call the % measurement as packet delivery ratio where as the bps measurement is called throughput.

Thanks all 

that is the code to calculate jitter in ns 2 using AWK

# JITTER  for cbr in UDP

# Jitter represent variance of delay

BEGIN { highest_packet_id=0; }

{

        action = $1;

        time = $2;

        from = $3;

        to = $4;

        type = $5;

        pktsize = $6;

        flow_id = $8;

        src = $9;

        dst = $10;

        seq_no = $11;

        packet_id = $12;

        if ( packet_id > highest_packet_id )

                highest_packet_id = packet_id;

        if ( start_time[packet_id] == 0 )

        {       pkt_seqno[packet_id] = seq_no;

                start_time[packet_id] = time;     }

        if ( flow_id == 2 && action != "d" )

        {       if( action == "r" )  { end_time[packet_id] = time; }

                else  { end_time[packet_id] = -1; }

        }

}

END {

        last_sequno = 0;

        last_delay = 0;

        seqno_diff = 0;

        for(packet_id=0; packet_id < highest_packet_id; packet_id++)

        {

                start = start_time[packet_id];

                end = end_time[packet_id];

                packet_duration = end-start;

                if( start < end )

                {

                        seqno_diff = pkt_seqno[packet_id]-last_seqno;

                        delay_diff = packet_duration - last_delay;

                        if ( seqno_diff == 0 )

                        { jitter = 0; }

                        else { jitter = delay_diff/seqno_diff; }

                        printf("%f %f\n",start, jitter);

                        last_seqno = pkt_seqno[packet_id];

                        last_delay = packet_duration;

                }

        }

}

Thanks a lot Imani 

If you know delay (response time) then you can easily calculate throughput. For example, to transfer a message (or file) of size X (bits) from point A to B. If D is the delay in (sec). Your throughput would be X/D (measured in bits/sec).

thanks nurul sarkar