AWK Programming for Packet Analysis | NS2 Basics | Lecture 12

AWK Programming

In this post, you will learn

• How to create an awk script

• How to analyse a given trace file using AWK scripts

Introduction

AWK is a text processing programming language developed at AT&T by A V Aho, P J Weinberger and B W Kernighan.  The first letters of these people make the script AWK (Aho, Weinberger, Kernighan)

There are three different versions of AWK 

1. awk

2. nawk

3. gawk

If the operating system is Linux, then it would be awk or gawk (GNU awk). This chapter tells you the very basic things about awk, and for elaborate help, refer [6]

Awk is a pattern-directed scanning and processing language. The column-wise processing is powerful as the prediction of results, analysis of data, etc are possible using this method.  

Syntax: 

awk [-Ffs] [-v var=value] [ -f file ...] [file ...]

Options: 

-Ffs option is  -F, and  fs  is field separator

-f specify an awk program

There are three different ways to execute an awk program:

Method 1:

awk '{print $1}' infile > outfile

Method 2:

awk -f awkprog   infile > outfile

Here, the awkprog is a separate file that contains

{

print $1 

}

Method 3

awkprog infile outfile 

The contents of the awkprog file as given below (it is also an executable file)          

awk '{

print $1

}' $1 >$2

Syntactically, a rule consists of a pattern followed by an action. The action is enclosed in curly braces to separate it from the pattern. Newlines usually separate rules. Therefore, an awk program looks like this

pattern { action }

pattern { action }

….

Missing action means print the line that matched the pattern.

e.g  /abc/   -print all the lines from a file where the abc string is found

missing pattern means an action to be done for all the lines

e.g { print $2 }   -print the second column for all the records

Arithmetic Operators

Similar to C programming, there are several arithmetic operators available in awk. 

Operator	Type	Meaning	Expression	Result
+	Arithmetic	Addition	9+2	11
-	Arithmetic	Subtraction	9-2	7
*	Arithmetic	Multiplication	9*2	18
/	Arithmetic	Division	9/2	4.5
%	Arithmetic	Modulo	9%2	1
<space>	String	Concatenation	9 2	92

The – and + can also be operated as an unary operator. If x is 4 then, 

print –x  will display -4 

Commands in AWK

There are only a limited number of commands in awk. They also form statements. Some of them are 

• if (expression) statement [else statement]

• if (condition)  statement1 else statement2

• while (expression) statement

• while(condition)   statement

• for (expression;expression;expression) statement

• for (exp1 ; condition ;exp2 ) statement

• for (variable in array) statement

• do statement while (expression)

• break  (** break -- to exit from a "while" or "for" loop)

• continue

• {[statement ...]}

• var=expression

• print[expression-list] [>expression]

• printf format [,expression-list ] [>expression]

• return [expression]

• next

• delete array [expression]

• exit [expression]

Built-in Variables in AWK

• FS   -field separator

• OFS  -output field separator

• NF   -no of fields 

• NR   -number record 

• FNR  -file record number

• RS   -record separator

• ORS  -output record separator

• ARGC -argument count

• ARGV -argument variables

• FILENAME -current file name

Escape characters:

• \a -  alert character

• \b - backspace character

• \f - form-feed character

• \n - new-line character

• \r - carriage-return character

• \t - tab character

• \v - vertical-tab character

Simple examples on Awk

Listing B.1 – AWK Script to print the data file as it is. Input Data  1, Pradeep,1000 2,kumar,2000 3, Krishna,3000 AWK Script  { print } Output Data: 1, Pradeep,1000 2,kumar,2000 3, Krishna,3000

Listing B.2 – AWK Script to print only the first column Input Data 1, Pradeep,1000 2,kumar,2000 3, Krishna,3000 AWK Script  BEGIN { FS="," } {   print $3 |" sort > out.dat " } Output Data: (You need to open the file out.dat) 1 2 3 4   If $3 was used, then the output will be changed to  1000 2000 3000

Listing B.3: Print the duplicate records Input Data 1 2 3 5 5 6 6 7 8 9 9 AWK Script    {     if (NR==1) {no=$1;i=0;next}     if (no==$1) print "Duplicate Entry:\t "no      else        no=$1 } Output Data: Duplicate Entry:5 Duplicate Entry:6 Duplicate Entry:9

Computing the Performance parameters using AWK

AWK scripts are much more powerful in analysing the network performance, particularly when logs or trace files are generated. Usually in networks, trace files are generated irrespective of whether the device is a simulator, emulator or even in an experimental setup. In NS2, each simulation generates an animation file and a trace file. The trace files can be processed using awk scripts to measure the following performance metrics like 

• Throughout the network

• Packet loss and packet drop calculation

• Energy calculation

• Packet delivery ratio

• End-to-end delay.

• Jitter 

• Plotting the Congestion window

Here is the input data for the following scripts 

Input Data:

r 3.885894108 _0_ AGT  --- 152 ack 60 [13a 0 1 800] ------- [1:0 0:0 32 0] [68 0] 1 0

s 3.885894108 _0_ AGT  --- 163 tcp 1040 [0 0 0 0] ------- [0:0 1:0 32 0] [88 0] 0 0

r 3.885894108 _0_ RTR  --- 163 tcp 1040 [0 0 0 0] ------- [0:0 1:0 32 0] [88 0] 0 0

s 3.885894108 _0_ RTR  --- 163 tcp 1060 [0 0 0 0] ------- [0:0 1:0 32 1] [88 0] 0 0

r 3.886183181 _1_ MAC  --- 0 ACK 38 [0 1 0 0] 

s 3.886533108 _0_ MAC  --- 0 RTS 44 [256e 1 0 0] 

r 3.886885181 _1_ MAC  --- 0 RTS 44 [256e 1 0 0] 

s 3.886895181 _1_ MAC  --- 0 CTS 38 [2434 0 0 0] 

r 3.887199254 _0_ MAC  --- 0 CTS 38 [2434 0 0 0] 

s 3.887209254 _0_ MAC  --- 130 tcp 1118 [13a 1 0 800] ------- [0:0 1:0 32 1] [73 0] 0 0

r 3.896153327 _1_ MAC  --- 130 tcp 1060 [13a 1 0 800] ------- [0:0 1:0 32 1] [73 0] 1 0

s 3.896163327 _1_ MAC  --- 0 ACK 38 [0 0 0 0] 

r 3.896178327 _1_ AGT  --- 130 tcp 1060 [13a 1 0 800] ------- [0:0 1:0 32 1] [73 0] 1 0

Here is the list of awk scripts that were helpful in measuring the above metrics. 

Listing B.4 – Throughput  AWK Script BEGIN {        recvdSize = 0        startTime = 400        stopTime = 0   }       {              event = $1              time = $2              node_id = $3              pkt_size = $8              level = $4       # Store start time   if (level == "AGT" && event == "s" && pkt_size >= 512) {     if (time < startTime) {              startTime = time              }        }       # Update total received packets' size and store packets arrival time   if (level == "AGT" && event == "r" && pkt_size >= 512) {        if (time > stopTime) {              stopTime = time              }        # Rip off the header        hdr_size = pkt_size % 512        pkt_size -= hdr_size        # Store received packet's size        recvdSize += pkt_size        }   }       END {        printf("Average Throughput[kbps] = %.2f\t\t StartTime=%.2f\tStopTime=%.2f\n",(recvdSize/(stopTime-startTime))*(8/1000),startTime,stopTime)   } Output: The output shows the throughput along with the StartTime and StopTime of the simulation

Listing B.5 – Packet Delivery Ratio AWK Scripts BEGIN {         sendLine = 0;         recvLine = 0;         fowardLine = 0; } $0 ~/^s.* AGT/ {         sendLine ++ ; } $0 ~/^r.* AGT/ {         recvLine ++ ; } $0 ~/^f.* RTR/ {         fowardLine ++ ; } END {         printf "cbr s:%d r:%d, r/s Ratio:%.4f, f:%d \n", sendLine, recvLine, (recvLine/sendLine),fowardLine; } Output:

Listing B.6 – Script for calculating Send,Received and Dropped packets and also Delivery Ratio, Average End-to-End Delay. AWK Script  BEGIN { seqno = -1;  droppedPackets = 0; receivedPackets = 0; count = 0; } { #packet delivery ratio 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<=seqno; i++) { if(end_time[i] > 0) { delay[i] = end_time[i] - start_time[i]; count++; } else { delay[i] = -1; } } for(i=0; i<count; i++) { if(delay[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"; } Output:

Conclusion:

AWK scripts is a powerful tool helpful in processing the trace files generated by the network elements. AWK processes the scripts based on pattern matching and helpful in plotting the network performance characteristics like throughput, packet delivery ratio, end to end delay, packet loss, drop calculation and other factors.