Upload
umapathy-manchala
View
220
Download
2
Embed Size (px)
Citation preview
8/13/2019 VoIP Call Processing in the MPLS Network
1/18
1/18
Architecture of End-to-End QoSfor VoIP Call Processingin the MPLS Network
2004. 9. 29
National Computerization Agency
(Republic of Korea)ChinChol Kim
[mail to] : [email protected]
Fifth International Workshop on Quality of future Internet Service(QofIS04)
8/13/2019 VoIP Call Processing in the MPLS Network
2/18
2/18
Contents
Abstract1. Introduction2. QoS Architecture for VolP in the NGN3. Architecture of End-to-End QoS for SIP
Call Signaling
4. Implementation5. Performance Analysis6. Conclusion
8/13/2019 VoIP Call Processing in the MPLS Network
3/18
3/18
Abstractq
This paper proposes the architecture of end-to-end QoS for VoIP callprocessing in the MPLS-based NGN supporting the IPv6
The architecture include QoS resource management and
differentiated call processing by extending SIPq Also, we analyze the performance of call processing
in the MPLS networkq
A performance analysis shows that there is a considerable difference in end-to-end call setup delay depending on service priority , insetting up SIP calls in the MPLS network
8/13/2019 VoIP Call Processing in the MPLS Network
4/18
4/18
1. Introduction (1)q
With the fast development of network technology, VoIP is one of the core technology of NGN and VoIPs
core protocols are H.323 and SIP signaling protocols At present, the text-based SIP has been adopted as the
standard for NGNq SIP must provide
a service quality better than the quality provided by thePSTN for call setup in NGN
and offer priority-based call processing , depending onthe traffic properties of application services
8/13/2019 VoIP Call Processing in the MPLS Network
5/18 5/18
1. Introduction (2)q
The VoIP service in NGN must guarantee call quality for voice data transfer and call setup quality forcall setup , in which the latter must precede the former
q On the other hand, call setup quality does notguarantee end-to-end QoS , since it goes through MPLSrouter and multiple SIP servers
q In order to resolve this problem, This paper proposes the architecture of end-to-end QoS for
VoIP call processing in an MPLS-based NGN QoS resource management coordinates service priority in call
processing Differentiated call processing technology processes calls,
applying the service priority negotiated in the SIP server
8/13/2019 VoIP Call Processing in the MPLS Network
6/18 6/18
2. QoS Architecture for VolP in the NG
q The call quality refers to speech coding distortion, terminal noise and
overall delay caused by packetization, buffering, codecs can be guaranteed by QoS technology, including MPSL,
Diffserv and IntServq On the other hand, call setup quality
refers to guaranteeing call setup, which is classified intocall setup delay in the network and VoIP server
Call setup delay in the network can be guaranteed throug
QoS technology, such as MPLS and DiffServ However, there is no standard technology that ensurescall setup delay in the VoIP server
q Therefore, we need various forms of QoS mechanisms in the VoIP server to
guarantee end-to-end QoS for call setup quality
8/13/2019 VoIP Call Processing in the MPLS Network
7/18 7/18
. rc ec ure o n - o- n ofor SIP Call Signaling
MPLS NetworkSIP Server SIP Server
Local Network A Local Network B
3. Priority Marking &Differentiated Call Processing
Router
LER
End-to-end QoS basedCall Processing
SIP UA SIP UA
R e s e r v e d L
S P
LSR
SIP Server
SIP UA
LER
Router
LSR
LSR
LSR
LER
Router
. LSP Setup
LER : Label Edge RouterLSR : Label Switching Router
2. Service Priority Negotiation
LERLER
8/13/2019 VoIP Call Processing in the MPLS Network
8/18 8/18
UDP
IPv6
UDP
IPv6
3.1 System structureUser Database
SIP6d Daemon
QRM
DQoSPSMP
Receiver
SMP : SIP Message ParserQRM : QoS Resource ManagementQRR : QoS Resource Requester
DQoSP : Differentiated QoS ProcessorSLS : SIP Location Server
UAC : User Agent ClientUAS : User Agent ServerSPC : SIP Proxy Server
SRS : SIP Redirect Server
SPS SRSSLS
SIP Ex-tension
UAS
QoSMarker
UAC
QRR
SIP6 UA
8/13/2019 VoIP Call Processing in the MPLS Network
9/18 9/18
3.2 SIP Message Extension and Flow (1)
QOSREQUESTQOSREQUEST method is used by SIP UA to negotiate servicepriority with the SIP server. It sets up the service priority requestedby the user through the Qosinfo header field.
Method Type Description
QOSWITHDRAW QOSWITHDRAW method is used by SIP UA to nullify negotiatedservice priority. It specifies the service priority to be nullified throughQosinfo header field.
Table 1. SIP Method extension
Qosinfo
Syntax Formalism : Qosinfo:desired/release=ServiceLevelDesired header option is used to set up the service priorityrequested by the user in QOSREQUEST method. It is also used to setup the negotiated service priority in 200 OK response messages.Release header option is used to release the service prioritynegotiated by the user in QOSWITHDRAW method and 200 OKresponse messages.
Header Type Description
Qosmark
Syntax formalism : Qosmark : ok / noThis is included in request/response message transferred to SIPmessage pass. If the header option in Qosmark header field is Ok,
the SIP server applies differentiated call processing technology.If it is No, the SIP server doesnt.
Table 2. SIP Header and Header Option extension
8/13/2019 VoIP Call Processing in the MPLS Network
10/18 10/18
IP UA Client SIP Server SIP Server SIP UA Ser
3.2 SIP Message Extension and Flow (2)
Q O S REQ U ES T 2 0 0 O K
I N V I T E
1 0 0 T r y i ng I N V I T E
I N V I T E 1 0 0 T r y i ng
1 8 0 R i ng i ng
2 0 0 O K
2 0 0 O K
Q O S W I T H DRAW
2 0 0 O K
...
1. Resource negot i at i on fl ow
2. Di fferent i at ed call processi ng fl ow
3. Resour ce rel ease f l ow
Qosmark header insert
1 8 0 R i ng i ng 1 8 0 R i ng i ng 2 0 0 O K
A C K
A C K
A C K
8/13/2019 VoIP Call Processing in the MPLS Network
11/18 11/18
3.3 Differentiated Call Processing AlgorithmMessage Processing Buffer
Insert Qosmark headerincluded No header option
to SIP Message
Premium Buffer Assured Buffer Normal Buffer
Scheduling
InsertExecution Buffer Message Parsing
Qosmark HeaderExit?
IPv6 Flow LabelField Value?
Premium Normal
Assured
Admission Controller
Classifier
Scheduler
Qosmark HeaderOption Value?
Yes Ok
No
Priority Authentication? No
Priority value set to IPv6 flow label field,Insert Qosmark header included Ok
header option to SIP Message
Yes No
l
8/13/2019 VoIP Call Processing in the MPLS Network
12/18 12/18
4. Implementation
qWe referred to the SIP source code fromColumbia University SIP6d is implemented using C language in Linu
system that supports IPv6 The SIP UA is implemented using Tcl/Tk and C++ MySQL is used to manage user information
The major modules are implemented using POSIXthread technology
5 P f A l i (1)
8/13/2019 VoIP Call Processing in the MPLS Network
13/18 13/18
5. Performance Analysis (1)
IPv6 DNSServer(Linux)
IPv6 MPLSRouter(Linux)
Test ProgramClient
SIP6d(Linux) SIP6d(Linux)
Test ProgramServer
SIP Signaling Data
IPv6 MPLSRouter(Linux)
IPv6 MPLSRouter(Linux)
It is composed of two Linux servers with SIP6d, two PCs with test programs, and
three Linux servers, which are used as routers with MPLSmodules based on the software provided by Sourceforge.net
8/13/2019 VoIP Call Processing in the MPLS Network
14/18
14/18
5. Performance Analysis (2)The experiment has adopted the following procedure
Firstly , the test client program generate equal number ofthree different INVITE messages and simultaneouslytransfers to SIP6d in the number of 50, 100, 150, 200, 250,and 300 messages of each priority.
Secondly , the two SIP6ds process the received messages,using differentiated call processing technology and transfethem to the test client server. Each message is moved alonthe LSP path predetermined in the three routers
Next , the test server program transfers 200OK responsemessages. Then, the test client program receives 200OKresponse messages coming through two SIP6ds, andmeasures session setup time, i.e. the average end-to-endcall setup delay time. For the comparison , INVITEmessages without priority are generated and sent to theSIP6d without a differentiated call processing function ithe number of 150, 300, 450, 600, 750, and 900.
5 P f A l i (3)
8/13/2019 VoIP Call Processing in the MPLS Network
15/18
15/18
5. Performance Analysis (3)
00.050.1
0.150.2
0.250.3
0.350.4
0.450.5
0.550.6
0.650.7
0.750.8
0.850.9
0.951
50x3 100x3 150x3 200x3 250x3 300x3
Premium PriorityAssured Priority
Normal Priority
Time(sec)
MessageNumber
00.050.1
0.150.2
0.25
0.30.350.4
0.450.5
0.550.6
0.650.7
0.750.8
0.850.9
0.951
150 300 450 600 750 900
Time(sec)
MessageNumber
Fig. 6. End-to-End Call Setup Delay in SIP6d over MPLS Network
(a) End-to-end Call Setup Delay in SIP6dsupporting Differentiated Call Processing (b) End-to-end Call Setup Delay in SIP6d
As one can see from the graph , the SIP6d that supports differentiated call processingshows a difference in call setup delay when processing messages. In particular, INVITEmessages with premium priority have very short call setup delay. Therefore, we can seethat INVITE messages with higher service priority have far shorter call setup delay than
those with lower service priority. However, SIP6d that does not support differentiatedcall processing has no difference in call setup delay.
5 P f A l i (4)
8/13/2019 VoIP Call Processing in the MPLS Network
16/18
16/18
5. Performance Analysis (4)
Fig. 7. End-to-End Call Setup Delay in SIP6d over Non-MPLS Network
(a) End-to-end Call Setup Delay in SIP6dsupporting Differentiated Call Processing
(b) End-to-end Call Setup Delay in SIP6d
00.050.1
0.150.2
0.250.3
0.350.4
0.450.5
0.550.6
0.650.7
0.750.8
0.850.9
0.951
50x3 100x3 150x3 200x3 250x3 300x3
Premium PriorityAssured PriorityNormal Priority
Time(sec)
MessageNumber
00.050.1
0.150.2
0.250.3
0.350.4
0.450.5
0.550.6
0.650.7
0.750.8
0.850.9
0.951
150 300 450 600 750 900
Time(sec)
MessageNumber
Also, End-to-end call setup delay over MPLS Network show better performancethan end-to-end call setup delay over Non-MPLS Network
6 C l i
8/13/2019 VoIP Call Processing in the MPLS Network
17/18
17/18
6. ConclusionThis paper proposes an architecture of end-to-enQoS for VoIP call processing in the MPLS network
The differentiated call processing technology reserveresources by extending SIP , and minimizes end-to-encall setup delay for specific calls by using prioritscheduling technology in the application level
It also has an advantage of setting up the service prioritthrough the flow label field of IPv6 header , considerinfuture MPLS label mapping
A performance analysis has showed that SIP6d provides a very fast processing rate fomessages with high service priority These results prove that we can provide excellen
performance for call setups that require real-time oservice priority when providing future voice service ithe NGN based on MPLS
8/13/2019 VoIP Call Processing in the MPLS Network
18/18
18/18
Q & AI f y ou hav e a quest i on, y ou send mai l t o me.
I w i l l send response mai l t o y ou.