Transcript Design and implementation of a distributed file directory for mobile

Design and implementation of a distributed
file directory for mobile peer-to-peer.
Master’s thesis by
Victor Morales
Thesis supervisor
Raimo Kantola
Instructor
Nicklas Beijar
Presentation outline
• Suitability of current P2P for mobile networks
• Proposed network architecture for Mobile P2P
(MP2P)
• Requirements for successful deployment in
mobile networks
• SIP and the IMS
• Overview of super-peer
• Results obtained
Suitability of current P2P for mobile
networks
• Many protocols researched by scientific
and Internet communities.
• Few have been designed for mobile
networks
• Even fewer have been implemented and
tested in mobile networks
• None are based on standardized protocols
Proposed network architecture for
MP2P
Our MP2P network works using super-peers, they in turn form a ”pure” P2P
network that will serve to forward queries that do not find an answer in the ”local”
super-peer. Every mobile-peer has one asociated super-peer.
Requirements for successful
deployment
• Adoption of 3G
• Adoption of IP Multimedia Subsystem
(IMS)
• Architecture suitable for MP2P
• Flat rates for data connectivity
SIP and the IMS
Third generation mobile networks impose a strict layering division by function.
This facilitates the implementation of services because there is a common and
standardized API. The IMS resides in the network control layer and provides IP
connectivity ”end-to-end” all the way to the mobile terminal.
overview of super-peer
The super-peers can
be collocated with other
nodes in the mobile
network.
They communicate with
the mobile terminals
using the SIP proxy.
A TCP relay is necessary
for bypassing the
provider’s firewall.
Communication can
proceed in a very similar
manner when using 3G.
Results obtained
Results obtained (2)
47
48
49
50 5
1
2
3
4.5
46
4
5
6
4
45
7
3.5
44
8
3
43
9
2.5
42
10
2
41
11
1.5
1
40
12
0.5
39
13
0
38
14
37
15
36
16
35
17
34
18
33
19
32
20
31
21
30
29
28
27
26
25
24
23
22
case 4 Found
case 4 Not found
case 3 Found
case 3 Not found
case 2 Found
case 2 Not found
case 1 Found
case 1 Not found
Results obtained (3)
100
90
80
70
60
50
40
case 4
case 3
30
case 2
case 1
20
10
0
0.5 0.6 0.7 0.8 0.9
1
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9
2
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9
3
Conclusions
• The architectural design of P2P-over-SIP
for mobile networks works in practice.
• The use of SIP for transporting P2P
control messages gives us advantages.
• User perceived delay is very good for all
our test cases.
• It is critical to have a high performance
SIP stack.
Future work
• Need for more testing and development
time for improving performance.
• Problem solving with reSIProcate stack.
• Possibly evaluate other stacks.
• Fine tuning of database.
• Live trial possible in near future.
• Implementing security features.