ns2 Simulation results 

-----------------------------------------------------------------------------------------
Scripts to get averaged results over multiple runs, described in the following documents:

	*  Y. Gong, D. Rossi, C. Testa, S. Valenti and D. Taht, "Fighting the bufferbloat: on the coexistence of AQM and low priority congestion control", 
Technical Report, Dec 2012
	* Y. Gong,  BUFFERDeBLOAT, "A survey of different ways to counter Bufferbloat", Projet presentation, July 2012

This document present the structure and organization of the results that can be obtained by running the scripts.

-----------------------------------------------------------------------------------------
Content:

 cap/Makefile			#calls make.sh
 cap/make.sh			#calls simulation-comp.tcl and *-process.py
 bin/simulation-comp.tcl	#main ns2 script
 bin/data-process.py		#post-process
 bin/stat-process.py		#post-process

-----------------------------------------------------------------------------------------


Naming [FILE]:

[AQM] AQM: [A*:DropTail/RED/SFQ/DRR/CHOCKE/CoDel] (DropTail, RED, SFQ, DRR Choke, CoDel)
[TCC] CCType: [T*:ledbat/nice/tcp-lp] (Ledbat, NICE, TCP-LP vs. NewReno)
[WLD] Workload: [W*:0-9, step 1] (1=10% on time, 9=90% on time, by abuse of notation 0 means  backlogged transfers)
[BUF] Buffer: [B*:100/350] (128KB, 512KB with pktsize=1500B)
[CAP] Capacity: [C*:10/1/0.5] (10Mbps,1Mbps,512Kbps, with B -> maxdelay=[0.1s], [1s], 2s | 0.4s, 4s, [8s])

[RUN] Repeat 10 times: [R*****:seedID]
[FLO] Flow status: [05:total CC][10:total flow][*:no.flow]{[0:TCP/1:CC])}
[FLR] Flow status at router (eg.queue size): [05:total CC][10:total flow][R]

[DLY] Delay: Homogeneous / Heterogeneous
Fixed Ratio: 5CC (0,1,2,3,4) + 5TCP (5,6,7,8,9)

-----------------------------------------------------------------------------------------
Procedure:

make.sh
 - simulation-comp.tcl 	-> FOLDER: raw/AQM.TCC.WLD
	- filecontent1	-> $now $throuput $bytes $insdrop $drops $cwnd
	- filename1		-> AQM.TCC.WLD.BUF.CAP.RUN.FLO.DATA
	- filecontent2	-> $now $r_throuput $r_bytes $r_insdrop $r_drops $r_qsize $router_qsize_avg
	- filename2		-> AQM.TCC.WLD.BUF.CAP.RUN.FLR.DATA
	- filecontent3	-> $item $bytes $droprate $cwnd_avg(flow) $qsize_avg(router)
	- filename3		-> AQM.TCC.WLD.BUF.CAP.RUN.DATA
 - data-process.py 	-> FOLDER: raw/AQM.TCC.WLD
	- filecontent	-> $eta $tcp% $f_all $f_tcp $f_cc $d_all $d_tcp $d_cc $cwnd_tcp $cwnd_cc $qsize (1RUN)
	- filename		-> AQM.TCC.WLD.BUF.CAP.RUN.STAT
 - stat-process.py	-> FOLDER: sum/AQM.TCC.WLD
	- filecontent	-> $eta $tcp% $f_all $f_tcp $f_cc $d_all $d_tcp $d_cc $cwnd_tcp $cwnd_cc $qsize (avg of 10 RUN)
	- filename		-> AQM.TCC.WLD.BUF.CAP.STAT
 - stat-process.py	-> FOLDER: sum/AQM.TCC.WLD
	- filecontent	-> $eta $tcp% $f_all $f_tcp $f_cc $d_all $d_tcp $d_cc $cwnd_tcp $cwnd_cc $qsize (list of 3 BUF/CAP combination)
	- filename		-> AQM.TCC.WLD.STAT
 - stat-process.py	-> FOLDER: 
	- filecontent	-> $eta $tcp% $f_all $f_tcp $f_cc $d_all $d_tcp $d_cc $cwnd_tcp $cwnd_cc $qsize (list of 10 WLD, with 3 values each metric)
	- filename		-> AQM.TCC.STAT

-----------------------------------------------------------------------------------------
Metrics:
[time] $throuput 	-> *.RUN.FLO.DATA *.RUN.FLR.DATA
[time] $insdrop		-> *.RUN.FLO.DATA *.RUN.FLR.DATA
[time] $cwnd		-> *.RUN.FLO.DATA
[time] $qsize		-> *.RUN.FLR.DATA
[comp] $eta $tcp% $f_all $f_tcp $f_cc $d_all $d_tcp $d_cc $cwnd_tcp $cwnd_cc $qsize	-> *.STAT