3 Simulation results » History » Version 24

MOURA, Ninon, 03/23/2016 11:12 AM

1 1 MERIOCHAUD, Antoine
h3. 3. Simulation results
2 1 MERIOCHAUD, Antoine
3 1 MERIOCHAUD, Antoine
Now we have .pcap files with all the informations about the packets exchanged during the Iperf command. We will use Wireshark to produce curves with these files.
4 1 MERIOCHAUD, Antoine
5 2 MERIOCHAUD, Antoine
First, we will analyze the evolution of sequence numbers.
6 8 MERIOCHAUD, Antoine
On the Time/Sequence graph with Wireshark, the upper curve corresponds to the receiver window, the middle one is the TCP segments that are sent and the lowest one is the actual ACK number. 
7 3 MERIOCHAUD, Antoine
8 14 MERIOCHAUD, Antoine
h4. 3.1. Evolution of sequence numbers
9 24 MOURA, Ninon
10 7 MERIOCHAUD, Antoine
!{width:60%}delai_no_pepsal_sequence.png!
11 3 MERIOCHAUD, Antoine
+Figure 1 : Evolution of sequence numbers without PEPsal+
12 3 MERIOCHAUD, Antoine
13 8 MERIOCHAUD, Antoine
14 1 MERIOCHAUD, Antoine
!{width:60%}delai_no_pepsal_sequence_zoom.png!
15 8 MERIOCHAUD, Antoine
+Figure 2 : Evolution of sequence numbers without PEPsal (zoom)+
16 1 MERIOCHAUD, Antoine
17 11 MERIOCHAUD, Antoine
In these two figures we observe a classical evolution of sequence numbers using TCP. We clearly observe the slow start phase during the first 9 seconds and then we are in a congestion avoidance phase without errors. When we zoom on the curve, we can see the packets are sent by burst. This is because we have a huge delay so the transmission is limited by the window size (85KB). Consequently the terminal sends many packets until the window is full, then it has to wait acknowledgments before being able to send again.
18 10 MERIOCHAUD, Antoine
Now we will observe the same curves but with PEPsal activated.
19 9 MERIOCHAUD, Antoine
20 9 MERIOCHAUD, Antoine
21 7 MERIOCHAUD, Antoine
!{width:60%}delai_pepsal_sequence.png!
22 3 MERIOCHAUD, Antoine
+Figure 3 : Evolution of sequence numbers with PEPsal+
23 3 MERIOCHAUD, Antoine
24 1 MERIOCHAUD, Antoine
!{width:60%}delai_pepsal_sequence_zoom.png!
25 14 MERIOCHAUD, Antoine
+Figure 4 : Evolution of sequence numbers with PEPsal (zoom)+
26 13 MERIOCHAUD, Antoine
27 1 MERIOCHAUD, Antoine
On this curve we observe that the slow start phase is quite shorter (around 5 seconds). Consequently we can achieve higher sequence numbers faster, se we increase the throughput.
28 17 MERIOCHAUD, Antoine
When we zoom, the behaviour is also very different than without PEPsal. Indeed here we see that the 3 curves (window, TCP segment and ACK) are very close because when a packet is sent, it is acknowledged by the PEP and the delay between the remote terminal and the PEP is negligible.
29 14 MERIOCHAUD, Antoine
30 17 MERIOCHAUD, Antoine
To highlight the improvement due to PEPsal, we decided to decrease the size window to 20KB. We produced the two next curves :
31 14 MERIOCHAUD, Antoine
32 16 MERIOCHAUD, Antoine
!{width:60%}delai_no_pepsal__littlewindow_sequence.png!
33 15 MERIOCHAUD, Antoine
+Figure 5 : Evolution of sequence numbers without PEPsal and with a window size of 20KB+
34 14 MERIOCHAUD, Antoine
35 14 MERIOCHAUD, Antoine
36 1 MERIOCHAUD, Antoine
!{width:60%}delai_pepsal__littlewindow_sequence.png!
37 1 MERIOCHAUD, Antoine
+Figure 6 : Evolution of sequence numbers with PEPsal and with a window size of 20KB+
38 17 MERIOCHAUD, Antoine
39 24 MOURA, Ninon
Here the improvement thanks to PEPsal is clear. If we compare the sequence number achieved after 1 second, we get around 8,000 without PEPsal and near 300,000 with PEPsal (37 times more). 
40 17 MERIOCHAUD, Antoine
This part highlight the fact that PEPsal mostly increase the performances in the slow start phase, so it is better to use it when we have to deal with a lot of short connections. 
41 17 MERIOCHAUD, Antoine
42 18 MERIOCHAUD, Antoine
h4. 3.2. Throughput
43 17 MERIOCHAUD, Antoine
44 24 MOURA, Ninon
Now, throughput in fonction of time is plotted.
45 22 MERIOCHAUD, Antoine
46 17 MERIOCHAUD, Antoine
!{width:60%}delai_no_pepsal_throughput.png!
47 17 MERIOCHAUD, Antoine
+Figure 7 : Throughput without PEPsal+
48 17 MERIOCHAUD, Antoine
49 17 MERIOCHAUD, Antoine
!{width:60%}delai_pepsal_throughput.png!
50 19 MERIOCHAUD, Antoine
+Figure 8 : Throughput with PEPsal+
51 21 MERIOCHAUD, Antoine
52 1 MERIOCHAUD, Antoine
53 22 MERIOCHAUD, Antoine
As expected, the throughput is better when we activate PEPsal and this is observable mostly at the beginning of the transmission (slow start phase).
54 22 MERIOCHAUD, Antoine
55 21 MERIOCHAUD, Antoine
h4. 3.3. RTT
56 1 MERIOCHAUD, Antoine
57 22 MERIOCHAUD, Antoine
In this part we observe the Round Trip Time (RTT) of each packet.
58 22 MERIOCHAUD, Antoine
59 21 MERIOCHAUD, Antoine
!{width:60%}delai_no_pepsal_rtt.png!
60 21 MERIOCHAUD, Antoine
+Figure 9 : RTT without PEPsal+
61 21 MERIOCHAUD, Antoine
62 21 MERIOCHAUD, Antoine
!{width:60%}delai_pepsal_rtt.png!
63 1 MERIOCHAUD, Antoine
+Figure 10 : RTT with PEPsal+
64 22 MERIOCHAUD, Antoine
65 23 MERIOCHAUD, Antoine
Without PEPsal we have a RTT quite constant around 600ms. This is legitimate because we added delay on the link to get this RTT. But when we activate PEPsal, the sender is fooled and does not see the satellite link. Consequently the observed RTT is less than 50ms which corresponds to the transmission time along the Ethernet wire and the processing time.