HighSpeed TCP Simulations Reports

HighSpeed TCP Simulation Reports

This page contains the results of several simulations done for the study of HighSpeed TCP . The NS scripts used are avaiable in the Scripts column, and its parameters at Behavior Scripts Parameters , at Notes column. For correct understanding of the reports is necessary to see the Traffic Types in the Notes column.

This is an ongoing research. The reports and scripts available here can be changed at any time.

Scripts	Equations Plot	Notes
behavior.tcl utils.tcl web.tcl	Loss Rate Decrease Rate Increase Rate HSTCP Response Function Parallel Flows Graph	Simulation Notes Behavior Script Parameter Traffic Types Slow Start Problem

Simulation Sets

Numbers	Description	Result
#01 - #20	Initial simulations with differents link parameters
#30 - #42	Comparison between HSTCP and REGTCP when the number of flows increase	#42
#43	Long term simulation of 1 hstcp and 1 regtcp (500 ms)	#43
#44 - #56	Comparison between HSTCP and REGTCP when the number of flows increase. Correction of slow start problem	#56
#57 - #69	Simulations with different link loss rate using Error Model	#69
#70 - #94	Simulation for loss rate of 0.00001 and 0.00005	#86
#95 - #96	Tests with max_ssthresh
#97 - #107	Simulation with 10 regtcp flows and other with 1 hstcp and 5 regtcp for differents link loss rate and max_ssthresh = 100	#107
#108 - #110	Simulation tests without loss
#111 - #114	Comparison between 1 HSTCP and N REGTCP with and without loss
#115 - #119	Simulations with DROPTAIL
#120 - #134	Tests to insert perturbation flow	#134
#135 - #138	Comparison with different max_ssthresh [0 10 100 1000]
#139	Comparison with 1 hstcp and 1/4/8/12 regtcp for multiple FTP
#140 - #143	Simulation with longer time (300sec)
#144 - #145	Simulation with UDP flow perturbation
#146 - #149	Simulation with variation in RED parameters and max_ssthresh slow-start
#150 - #159	Tests with perturbations

All Simulations

NR	Parameters	Observations	Statistics	Graphics
1	1000 25 0 100 5 5 10 12.5 12500 0.0000096 0.1 0		Report	Graphics
2	1000 25 0 100 5 5 10 12.5 12500 0.0000096 0.1 1		Report	Graphics
3	1000 100 0 100 5 5 10 31 83000 0.0000001 0.1 1		Report	Graphics
15	1000 25 0 100 5 5 10 31 83000 0.0000001 0.1 1		Report	Graphics
16	1000 25 0 100 5 5 10 31 83000 0.0000001 0.1 0		Report	Graphics
17	3000 25 0 100 5 5 10 31 83000 0.0000001 0.1 1		Report	Graphics
18	3000 25 0 100 5 5 10 31 83000 0.0000001 0.1 0		Report	Graphics
19	7000 25 0 100 5 5 10 31 83000 0.0000001 0.1 1		Report	Graphics
20	7000 25 0 100 5 5 10 31 83000 0.0000001 0.1 0		Report	Graphics
Changes: 1) correction link utilization 2) nr smalltcp = 10, windown_ = 8 3) qflag parameter introduced again
30	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1	2 flows, 1 hstcp and 1 regtcp	Report	Graphics
31	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 0 1	2 flows, 2 regtcp	Report	Graphics
32	1000 50 0 100 2 2 10 31 83000 0.0000001 0.1 1 1	4 flows, 2 hstcp and 2 regtcp	Report	Graphics
33	1000 50 0 100 2 2 10 31 83000 0.0000001 0.1 01	4 flows, 4 regtcp	Report	Graphics
34	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
35	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1	10 flows, 10 regtcp	Report	Graphics
36	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 1 1	20 flows, 10 hstcp and 10 regtcp	Report	Graphics
37	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 0 1	20 flows, 20 regtcp	Report	Graphics
38	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 1 1	40 flows, 20 hstcp and 20 regtcp	Report	Graphics
39	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 0 1	40 flows, 40 regtcp	Report	Graphics
40	1000 50 0 100 15 15 10 31 83000 0.0000001 0.1 1 1	30 flows, 15 hstcp and 15 regtcp	Report	Graphics
41	1000 50 0 100 15 15 10 31 83000 0.0000001 0.1 0 1	30 flows, 30 regtcp	Report	Graphics
42	Comparison between HSTCP and REGTCP when the number of flows increase (link utilization and link loss)		Report	Graphics
43	1000 50 0 500 1 1 10 31 83000 0.0000001 0.1 1 1	Long simulation with 1 hstcp/1 regtcp	Report	Graphics
Changes: 1) if (ssthresh_ == 0) ssthresh_ = int(wnd_); 2) Agent/TCP set ssthresh_ 1000
44	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1	2 flows, 1 hstcp and 1 regtcp	Report	Graphics
45	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 0 1	2 flows, 2 regtcp	Report	Graphics
46	1000 50 0 100 2 2 10 31 83000 0.0000001 0.1 1 1	4 flows, 2 hstcp and 2 regtcp	Report	Graphics
47	1000 50 0 100 2 2 10 31 83000 0.0000001 0.1 01	4 flows, 4 regtcp	Report	Graphics
48	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
49	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1	10 flows, 10 regtcp	Report	Graphics
50	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 1 1	20 flows, 10 hstcp and 10 regtcp	Report	Graphics
51	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 0 1	20 flows, 20 regtcp	Report	Graphics
52	1000 50 0 100 15 15 10 31 83000 0.0000001 0.1 1 1	30 flows, 15 hstcp and 15 regtcp	Report	Graphics
53	1000 50 0 100 15 15 10 31 83000 0.0000001 0.1 0 1	30 flows, 30 regtcp	Report	Graphics
54	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 1 1	40 flows, 20 hstcp and 20 regtcp	Report	Graphics
55	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 0 1	40 flows, 40 regtcp	Report	Graphics
56	Comparison between HSTCP and REGTCP when the number of flows increase (link utilization and link loss)		Report	Graphics
Changes: 1) new parameter em_rate (error model loss rate) and its code 2) simulation with em_rate = 10^-7, 10^-6, 10^-5, 10^-4, 10^-3 for hstcp+regtcp and all regtcp
57	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
58	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.0	10 flows, 10 regtcp	Report	Graphics
59	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 0.000001	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
60	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.0 00001	10 flows, 10 regtcp	Report	Graphics
61	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.00001	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
62	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.00001	10 flows, 10 regtcp	Report	Graphics
63	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0001	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
64	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.0 001	10 flows, 10 regtcp	Report	Graphics
65	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.001	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
66	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.001	10 flows, 10 regtcp	Report	Graphics
67	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0000001	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
68	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.0000001	10 flows, 10 regtcp	Report	Graphics
69	Simulation with different link loss rate using Error Model (uging 10 flows for all simulations, 5hstcp/5regtcp or 10regtcp)		Report	Graphics
Simulation for loss rate of 0.00001
70	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.00001	2 flows, 1 hstcp and 1 regtcp	Report
71	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 0 1 0.00001	2 flows, 2 regtcp	Report
72	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.00001	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
73	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.00001	10 flows, 10 regtcp	Report
74	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 1 1 0.00001	20 flows, 10 hstcp and 10 regtcp	Report
75	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 0 1 0.00001	20 flows, 20 regtcp	Report
76	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 1 1 0.00001	40 flows, 20 hstcp and 20 regtcp	Report
77	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 0 1 0.00001	40 flows, 40 regtcp	Report
Simulation for loss rate of 0.00005
78	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.00005	2 flows, 1 hstcp and 1 regtcp	Report
79	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 0 1 0.00005	2 flows, 2 regtcp	Report
80	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.00005	10 flows, 5 hstcp and 5 regtcp	Report
81	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.00005	10 flows, 10 regtcp	Report
82	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 1 1 0.00005	20 flows, 10 hstcp and 10 regtcp	Report
83	1000 50 0 100 10 10 10 31 83000 0.0000001 0.1 0 1 0.00005	20 flows, 20 regtcp	Report
84	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 1 1 0.00005	40 flows, 20 hstcp and 20 regtcp	Report
85	1000 50 0 100 20 20 10 31 83000 0.0000001 0.1 0 1 0.00005	40 flows, 40 regtcp	Report
86	Simulation results with different nr flows for loss rate 0.00001 and 0.0005. Link Utilization, Aggregate HSTCP-REGTCP ratio and drops graphs		Report	Graphics
87	1000 50 0 100 30 30 10 31 83000 0.0000001 0.1 1 1 0.00001	60 flows, 30 hstcp and 30 regtcp	Report
88	1000 50 0 100 30 30 10 31 83000 0.0000001 0.1 0 1 0.00001	60 flows, 60 regtcp	Report
89	1000 50 0 100 40 40 10 31 83000 0.0000001 0.1 1 1 0.00001	80 flows, 40 hstcp and 40 regtcp	Report
90	1000 50 0 100 40 40 10 31 83000 0.0000001 0.1 0 1 0.00001	80 flows, 80 regtcp	Report
91	1000 50 0 100 30 30 10 31 83000 0.0000001 0.1 1 1 0.00005	60 flows, 30 hstcp and 30 regtcp	Report
92	1000 50 0 100 30 30 10 31 83000 0.0000001 0.1 0 1 0.00005	60 flows, 60 regtcp	Report
93	1000 50 0 100 40 40 10 31 83000 0.0000001 0.1 1 1 0.00005	80 flows, 40 hstcp and 40 regtcp	Report
94	1000 50 0 100 40 40 10 31 83000 0.0000001 0.1 0 1 0.00005	80 flows, 80 regtcp	Report
Changes: 1) updated ns to 2.1b9 2) removed "Agent/TCP set ssthresh_ 1000" and included "Agent/TCP set max_ssthresh_ 1000" 3) correcetd problem in calculus of aggregate link utilization for each class. All previous simulations had an error in the calculus of 1st half aggregate link utilization. This also changes the calculus of aggregate link utilization of 2nd half of simulation
95	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0	Simulation using "Agent/TCP set max_ssthresh_ 1000"	Report	Graphics
96	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.0	Simulation using "Agent/TCP set max_ssthresh_ 1000"	Report	Graphics
Change: 1) max_ssthresh to "100"
97	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.0 000001	10 flows, 10 regtcp	Report	Graphics
98	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.000001	10 flows, 10 regtcp	Report	Graphics
99	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.00001	10 flows, 10 regtcp	Report	Graphics
100	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.0001	10 flows, 10 regtcp	Report	Graphics
101	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 0 1 0.001	10 flows, 10 regtcp	Report	Graphics
102	1000 50 0 100 1 5 10 31 83000 0.0000001 0.1 1 1 0.0000001	6 flows, 1 hstcp and 5 regtcp	Report	Graphics
103	1000 50 0 100 1 5 10 31 83000 0.0000001 0.1 1 1 0.000001	6 flows, 1 hstcp and 5 regtcp	Report	Graphics
104	1000 50 0 100 1 5 10 31 83000 0.0000001 0.1 1 1 0.00001	6 flows, 1 hstcp and 5 regtcp	Report	Graphics
105	1000 50 0 100 1 5 10 31 83000 0.0000001 0.1 1 1 0.0001	6 flows, 1 hstcp and 5 regtcp	Report	Graphics
106	1000 50 0 100 1 5 10 31 83000 0.0000001 0.1 1 1 0.001	6 flows, 1 hstcp and 5 regtcp	Report	Graphics
107	Simulation with 10 regtcp flows and other with 1 hstcp and 5 regtcp for differents link loss rate		Report-1 Report-2	Graphics
Simulations without loss
108	1000 50 0 100 1 5 10 31 83000 0.0000001 0.1 1 1 0.0	6 flows, 1 hstcp and 5 regtcp	Report	Graphics
109	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0	10 flows, 5 hstp and 5 regtcp	Report	Graphics
110	1000 50 0 100 2 3 10 31 83000 0.0000001 0.1 0 1 0.0	5 flows, 5 regtcp	Report	Graphics
1 HSTCP against N REGTCP with and without loss
111	1000 50 0 100 1 {1 3 5 7 10 15 20} 10 31 83000 0.0000001 0.1 1 1 0.0	1 hstcp + N regtcp without loss	Report	Graphics
112	1000 50 0 100 1 {1 3 5 7 10 15 20} 10 31 83000 0.0000001 0.1 1 1 0.00001	1 hstcp + N regtcp loss em_rate=10^-5	Report	Graphics
113	1000 50 0 100 1 {1 3 5 7 10 15 20} 10 31 83000 0.0000001 0.1 0 1 0.0	1 regtcp + N regtcp without loss	Report	Graphics
114	1000 50 0 100 1 {1 3 5 7 10 15 20} 10 31 83000 0.0000001 0.1 0 1 0.00001	1 regtcp + N regtcp loss em_rate=10^-5	Report	Graphics
Simulation with DROPTAIL
115	1000 50 1 100 1 {1 3 5 7 10 15 20} 10 31 83000 0.0000001 0.1 1 1 0.0	1 hstcp + N regtcp without loss and droptail	Report	Graphics
116	1000 50 1 100 1 {1 3 5 7 10 15 20} 10 31 83000 0.0000001 0.1 1 1 0.000001	1 hstcp + N regtcp with em_rate = 10^-6 and droptail	Report	Graphics
117	Simulation with 10 regtcp flows and other with 5 hstcp and 5 regtcp for differents link loss rate and droptail	10 flows, 5 hstcp and 5 regtcp	Report-1 Report-2	Graphics
118	1000 50 1 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0	10 flows, 5 hstcp and 5 regtcp	Report	Graphics
119	Comparison between HSTCP and REGTCP when the number of flows increase	2/10/20/30/40 flows, 1/5/10/15/20 hstcp and regtcp flows	Report-1 Report-2	Graphics
Changes: 1) correct problem in link loss rate 2) change the start time of flows to 5% of the total time 3) insert controlled small perturbations on the simualtion (class=220) 4) insert parameter to control perturbation time cycle (pertinterval)
120	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	10 flows, 5 hstcp and 5 regtcp, [pkts=4000 and max_ssthresh=0 for perturbation]	Report	Graphics
121	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	10 flows, 5 hstcp and 5 regtcp, [pkts=8000 and max_ssthresh=0 for perturbation]	Report	Graphics
122	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	10 flows, 5 hstcp and 5 regtcp, [pkts=4000 and max_ssthresh=1000 for perturbation]	Report	Graphics
123	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	10 flows, 5 hstcp and 5 regtcp, [pkts=8000 and max_ssthresh=1000 for perturbation]	Report	Graphics
124	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	10 flows, 5 hstcp and 5 regtcp, [pkts=32000 and max_ssthresh=0 for perturbation]	Report	Graphics
125	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	10 flows, 5 hstcp and 5 regtcp, [pkts=32000 and max_ssthresh=1000 for perturbation]	Report	Graphics
126	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	10 flows, 5 hstcp and 5 regtcp, without perturbation	Report	Graphics
127	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	2 flows, 1 hstcp and 1 regtcp, [pkts=4000 and max_ssthresh=0 for perturbation]	Report	Graphics
128	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	2 flows, 1 hstcp and 1 regtcp, [pkts=8000 and max_ssthresh=0 for perturbation]	Report	Graphics
129	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	2 flows, 1 hstcp and 1 regtcp, [pkts=32000 and max_ssthresh=0 for perturbation]	Report	Graphics
130	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	2 flows, 1 hstcp and 1 regtcp, [pkts=4000 and max_ssthresh=1000 for perturbation]	Report	Graphics
131	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	2 flows, 1 hstcp and 1 regtcp, [pkts=8000 and max_ssthresh=1000 for perturbation]	Report	Graphics
132	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	2 flows, 1 hstcp and 1 regtcp, [pkts=32000 and max_ssthresh=1000 for perturbation]	Report	Graphics
133	1000 50 0 100 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 5.0001	2 flows, 1 hstcp and 1 regtcp, [pkts=0 and max_ssthresh=0 for perturbation] (without perturbation)	Report	Graphics
134	Link utilization results of simulations with perturbation for 5 hstcp / 5 regtcp and 1 hstc and 1 regtcp. Perturbation each 5 seconds			Graphics
135	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; max_ssthresh = 0	Report	Graphics
136	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; max_ssthresh = 10	Report	Graphics
137	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; max_ssthresh =100	Report	Graphics
138	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; max_ssthresh =1000	Report	Graphics
Multiple FTP
139	1000 50 0 100 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 0	Simulation to campare the link utilization of 1, 2 hstcp alone and 1, 4, 8, 12 regtcp (multiple FTPs) also alone; with background noise X = [ 1 ], Y = [ 1 4 8 12 ]	Report	Graphics
Longer Simulations (300 secs)
140	1000 50 0 300 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; 300 secs	Report	Graphics
141	1000 50 0 300 1 0 10 31 83000 0.0000001 0.1 1 1 0.0 0	1 hstcp flow, 300 secs, RED	Report	Graphics
142	1000 50 1 300 1 0 10 31 83000 0.0000001 0.1 1 1 0.0 0	1 hstcp flow, 300 secs, DT	Report	Graphics
143	1000 50 0 300 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 0	2 flows, 1 hstcp and 1 regtcp; 300 secs	Report	Graphics
UDP Flows
144	1000 50 0 300 5 5 10 31 83000 0.0000001 0.1 1 Y 0.0 0 X	Simulation with Y=[0 1], X=[0 25 50 75 100 150 200] UDP flows with Pareto distribution and 5 hstcp + 5 regtcp and 10 regctp flows		Graphics
145	1000 50 0 300 2 2 10 31 83000 0.0000001 0.1 1 Y 0.0 0 X	Simulation with Y=[0 1], X=[0 25 50 75 100 150 200] UDP flows with Pareto distribution and 2 hstcp + 2 regtcp and 10 regctp flows		Graphics
RED + max_ssthresh
146	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0 0	10 flows, 5 hstcp and 5 regtcp; Queue RED set bottom_ 0.0001	Report	Graphics
147	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; new hstcp at 40s; max_sstresh=0	Report	Graphics
148	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; new hstcp at 40s; max_sstresh=10	Report	Graphics
149	1000 50 0 100 5 5 10 31 83000 0.0000001 0.1 1 1 0.0 0	10 flows, 5 hstcp and 5 regtcp; ew hstcp at 40s; max_sstresh=1000	Report	Graphics
Changes: 1) Removed code for UDP perturbations flows 2) Inserted code to initiate a random hstcp flow 3) Calculus of average flow link utilization per class 4) Inserted code of perturbation with TCP flows
150	1000 50 0 100 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 N	A) 6 flows, 3 hstcp and 3 regtcp with 0, 10, 20 30 perturb B) 6 hstcp with 0, 10, 20, 30 perturb C) 6 regtcp with 0, 10, 20, 30 perturb	Report	Graphics
151	1000 50 0 300 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 0	2 flows, 1 hstcp and 1 regtcp, 300s, no perturb	Report	Graphics
152	1000 50 0 300 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 1	2 flows, 1 hstcp and 1 regtcp, 300s, 1 perturb	Report	Graphics
153	1000 50 0 300 0 1 10 31 83000 0.0000001 0.1 1 1 0.0 1	1 flow, 1 regtcp, 300s, 1 perturb	Report	Graphics
154	1000 50 0 300 1 0 10 31 83000 0.0000001 0.1 1 1 0.0 1	1 flow, 1 hstcp, 300s, 1 perturb	Report	Graphics
155	1000 50 1 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.00001 0	2/10/20/40 flows; hstcp only, hstcp+regtcp, regtcp only; DT; Link Loss = 10^-5	Report	Graphics
156	1000 50 0 300 0 1 10 31 83000 0.0000001 0.1 1 1 0.0 4	1 regtcp flow, 300 secs, 4 perturbations, pertwin=3 secs	Report	Graphics
157	1000 50 1 300 0 1 10 31 83000 0.0000001 0.1 1 1 0.0 3	1 regtcp, 300 secs, 3 perturb, DT, pertwin = 10 secs	Report	Graphics
158	1000 50 0 300 0 1 10 31 83000 0.0000001 0.1 1 1 0.0 3	1 regtcp, 300 secs, 3 perturb, RED, pertwin = 10 secs	Report	Graphics
159	1000 50 0 300 1 0 10 31 83000 0.0000001 0.1 1 1 0.0 3	1 hstcp, 300 secs, 3 perturb, RED, pertwin = 10 secs	Report	Graphics
160	Comparison among 1 hstcp, 2 hstcp, 1 hstcp against 6 regtcp, 6 regtcp against 1 hstcp and 6 regtcp; in different link loss condition 10^-7 to 10^-3			Graphics
Changes: 1) pertwin = 10s
161	1000 50 0 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 N	A) 6 hstcp with 0, 1, 3, 10, 20 perturb B) 6 regtcp with 0, 1, 3, 10, 20 perturb C) 6 flows, 3 hstcp and 3 regtcp with 0, 1, 3, 10, 20 perturb	Report-RED	Graphics-RED
162	1000 50 1 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 N	A) 6 hstcp with 0, 1, 3, 10, 20 perturb B) 6 regtcp with 0, 1, 3, 10, 20 perturb C) 6 flows, 3 hstcp and 3 regtcp with 0, 1, 3, 10, 20 perturb D) DropTail	Report-DT	Graphics-DT
163	1000 50 0 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 N	A) 6 hstcp with 0, 1, 3, 10, 20 perturb B) 6 regtcp with 0, 1, 3, 10, 20 perturb C) 6 flows, 3 hstcp and 3 regtcp with 0, 1, 3, 10, 20 perturb D) RED, Perturbation random on the second half	Report	Graphics
164	1000 50 0 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 N	A) 6 hstcp with 0, 1, 3, 10, 20 perturb B) 6 regtcp with 0, 1, 3, 10, 20 perturb C) 6 flows, 3 hstcp and 3 regtcp with 0, 1, 3, 10, 20 perturb D) RED, perturbation random on the second half E) perturbations max_ssthresh = 100	Report	Graphics
165	1000 50 0 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 N	A) 6 hstcp with 0, 1, 3, 10, 20 perturb B) 6 regtcp with 0, 1, 3, 10, 20 perturb C) 6 flows, 3 hstcp and 3 regtcp with 0, 1, 3, 10, 20 perturb D) RED, perturbation random on the second half E) perturbations max_ssthresh = 100 F) Increase RTT to flows in ~ 40ms	Report	Graphics
166	1000 50 0 3600 1 1 10 31 83000 0.0000001 0.1 1 1 0.0 0	A) Log-term hstcp-regtcp interaction Time = 3600 secs , RED	Report-RED	Graphics-RED
167	1000 50 0 300 0 X 10 31 83000 0.0000001 0.1 1 1 0.0 0	A) X= [1 3] B) Simulation with packet size = 1500 bytes and 1000 bytes		Graphics
168	1000 50 Y 300 0 X 10 31 83000 0.0000001 0.1 1 1 0.0 0	Link Utilization and Queue Size for multiple TCP streams A) Y [RED/DT] B) X [ 1 2 3 4 5 6 ] regtcp flows		Graphics
169	1000 50 1 300 0 6 10 31 83000 0.0000001 0.1 1 1 X 0	Aggregate Link Utilization for 6 Parallel Regular TCP flows X = [10^-7, 5*10^-7, 10^-6]		Graphics
170	1000 50 1 300 0 X 10 31 83000 0.0000001 0.1 1 1 0.0000001 0	Link Utilization and Queue Size for multiple TCP streams with Link Loss = 10^-7 X [ 4 5 6 ] regtcp flows		Graphics
171		Traffic Shape A) Background noise B) Burstiness 10/20/30 TCP Flows C) 50 ON/OFF UDP flows, pareto distribution	Report-TCP Report-Pareto	Graphics-TCP Graphics-Pareto
Changes: 1) Separate UDP-Pareto flows from TCP perturbation flows parameters 2) Possible to remove background noise 3) Packetsize_ 1500 4) Save aggregate flow class utilization each 100 ms 5) Create scripts to calculate Jain's fairness index and min-max fairness
172	1000 50 0 300 Y Z 10 31 83000 0.0000001 0.1 1 1 0.0 0 20 X	X = [4 40 400] miliseconds Y, Z = 10 HSTCP, 10 regtcp, 5hstcp - 5 regtcp	Report-RED Report-DT	Graphics
173	1000 50 0 300 X Y 10 31 83000 0.0000001 0.1 1 1 Z 0 0 0 1	X = [1, 4], Y = [1, 4, 8, 12] Z = [10e-6, 10e-5, 10e-4, 10e-3] and 1 hstcp, 1 regtcp, every 100ms, with loss rate = 10e-6	Report-RED Report-Loss	Graphics
174	Comparison with real network deployment (Brian's Test) PATH - BWD (Mbits/s) - RTT (secs) 1 400 4 2 300 175 3 750 10 4 450 80 5 450 80 6 460 56 7 420 67 8 530 77 9 460 154	Duration = 30 s Queue type = Droptail 1 HSTCP alone 1 REGTCP alone		Graphics
175	1000 50 0 50 1 0 10 31 83000 0.0000001 0.1 1 1 0.00001 0 0 0 0	Bug in TCP (old ns version)	Trace Report-right Report-wrong	CongWind Graphics-right Graphics-wrong

360	A set of random simulation of 2/6/10/20/30/40 flows HSTCP+RegTCP and RegTCP+RegTCP	RED	Report-1 Report-2	Graphics
361	A set of random simulation of 2/6/10/20/30/40 flows HSTCP+RegTCP and RegTCP+RegTCP	DT	Report-1 Report-2	Graphics
362	A set of random simulation of 1 HSTCP flow running against 1/3/5/7/11/15/20/25/30 RegTCP flows	RED and DT	Report-RED Report-DT	Graphics-RED Graphics-DT
363	A set of random simulation of 1/2 HSTCP and 1/4/8/12 RegTCP	RED and DT	Report-RED	Graphics-RED
364	A set od random simulation of 5 hstcp / 5 regtcp and 10 regtcp	RED and DT	Report-RED Report-DT	Graphics-RED Graphics-DT
365	1000 50 0 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 0	RED/ MTU=[40 552 576 1000 1500 9000] 1 regtcp and 1 hstcp	Report	Graphics-RED Graphics-DT
366	1000 50 0 300 X Y 10 31 83000 0.0000001 0.1 1 1 0.0 0.00001	RED/DT, Link Loss = 10E-5 X,Y=[2 6 10 20 30 40]		Graphics-RED

Page last modified: Wednesday, 04-Dec-2002 08:52:00 PSTContact: Webmaster < webmaster@george.lbl.gov > Credits:The research and development of the Distributed Systems Department is funded by the U.S. Dept. of Energy, Office of Science, Office of Advanced Scientific Computing Research, Mathematical, Information, and Computational Sciences Division. Privacy and site security notice to Users

You are visitor [an error occurred while processing this directive] to this page since June 2002.