Dccp evaluation for sip signaling ict4 m
Download as PPTX, PDF0 likes295 views
This document discusses evaluating the performance of using the Datagram Congestion Control Protocol (DCCP) for SIP signaling compared to the traditional UDP. It describes developing a DCCP agent and SIP traffic for the NS-2 network simulator to simulate SIP call setup over DCCP and UDP. The simulation results show that DCCP has lower call drop rates than UDP and less variation in call setup delays, indicating DCCP may be a preferable transport for SIP signaling over UDP.
![DCCP
• Unreliable flows of datagrams
• Reliable handshakes for connection setup and teardown
• Reliable negotiation of options
• Mechanisms to avoid holding state for unacknowledged connection
attempts and already-finished connections
• Congestion control incorporating Explicit Congestion Notification (ECN)
[RFC3168] and the ECN Nonce [RFC3540]
• Acknowledgement mechanisms communicating packet loss and ECN
information.
• Optional mechanisms that tell the sending application which data packets
reached the receiver, and whether those packets were ECN marked,
corrupted, or dropped in the receive buffer
• Path Maximum Transmission Unit (PMTU) discovery [RFC1191].
• Two mechanisms of modular congestion control mechanisms : TCP-like
Congestion Control [RFC4341] and TCP-Friendly Rate Control (TFRC)
[RFC4342].](https://image.slidesharecdn.com/dccpevaluationforsipsignaling-ict4m-160126144540/85/Dccp-evaluation-for-sip-signaling-ict4-m-10-320.jpg)
Dccp evaluation for sip signaling ict4 m
- 1. DCCP Evaluation for SIP Signaling Agus Awaludin Mohamad Fathurahman Reza Primardiansyah Riri Fitri Sari Depeartment of Electrical Engineering, Faculty of Engineering University of Indonesia, Depok 16424 E-mail : awaludin@ui.ac.id, mohamad.fathurahman91@ui.ac.id, reza.primardiansyah@ui.ac.id, riri@ui.ac.id
- 2. Agenda • Motivation • Background • Development of DCCP_SIP Agent and SIP_DCCP Application Traffic • Simulation of SIP over DCCP • Result and Analysis • Conclusion
- 3. Motivation • Objective – Evaluate the performance impact on the use of DCCP for SIP Signaling • Approach – Utilize Network Simulator 2 (NS-2) for simulation of SIP based call setup over DCCP and UDP – Compare the impact of both Transport Signaling Protocol on Call Setup Delay and Call Drop Rate • Result – Adding new required Agent and Traffic for SIP over DCCP, based in previous work of independent SIP library – Challenges on further investigation on proving the simulation result that DCCP can minimize the Call Drop Rate
- 4. Background • NGN Protocol • SIP • UDP • Explicit Code Notification • Active Congestion Management • DCCP
- 5. NGN Protocol • Generally, these protocols are grouped into Signalling Transport Protocol, Bearer Control Protocol, and Call Control Protocol. – Signalling Transport Protocol serves transport layer signaling while serving higher level services. – Bearer Control Protocol controls NGN media gateways. – Call Control Protocol is used to control call setup. • Adopting the future multimedia services, the next generation of fixed and mobile network evolved from Next Generation (NGN) and IP Multimedia Subsytem (IMS) framework to Evolved Packet System (EPS) as the meeting point of NGNs MTP1 IP TCP H.248 MGCP SIP MAC SS7 H.248 MGCP SIP MTP3 MTP2 I S U P SCTP IP MAC SCTPUDP IP MAC UDP IP MAC UDP 1 2 3 4 5 6 7 OSI IP H.323 MAC H.323 TCPUDPM2UA M3UA SCCP TCAP INAP • Next Generation Network uses several protocols for interconnection between existing communication components, hardware or user alike.
- 6. SIP (RFC 3261) • Session Initiation Protocol is a Call Control Protocol that used for builds, modify and tear down one or more participant session • SIP is IETF main protocol for multimedia and data control signaling framework. • SIP defines the following entities in its protocol: – Location Service – Proxy Server – Redirect Server – Registrar – User Agent (Client and Server) • SIP has a significant role as the main signalling and call control protocol on all next generation framework
- 7. UDP (RFC 768) • UDP is a simpler message-based connectionless protocol. – Unreliable : no concept of Acknowledgment, retransmission or timeout – Not ordered : the order in which they arrive cannot be predicted – Lightweight: no ordering of messages, no tracking connections – Datagrams: Packets have definite boundaries, read operation at the receiver socket will yield an entire message • Until now UDP is agreed as an optimal IP transport protocol for SIP. • The unreliable characteristics of UDP brings an issue regarding the possibilites of call drop caused by the congestion on the network, leading the imposibility to implement reacheability of telecommunication service.
- 8. Explicit Code Notification (RFC3168 ) • Extention to the TCP/IP that notify the network congestion through marking packets • Use an ECN fields in the IP header with two bits, ECN-Capable Transport (ECT) codepoints '10', ECT(0), and '01', ECT(1), indicate that the end-points of the transport protocol are ECN-capable • Senders are free to use either the ECT(0) or the ECT(1) codepoint to indicate ECT, on a packet-by-packet basis • The not-ECT codepoint '00' indicates a packet that is not using ECN. • The CE codepoint '11' is set by a router to indicate congestion to the end nodes
- 9. AQM • AQM is a technique that consists in dropping or ECN-marking packets before a router's queue is full • AQM drops packets based on the average queue length exceeding a threshold, rather than only when the queue overflows. • AQM can set a Congestion Experienced (CE) codepoint in the packet header instead of dropping the packet, when such a field is provided in the IP header and understood by the transport protocol. • The use of the CE codepoint with ECN allows the receiver(s) to receive the packet, avoiding the potential for excessive delays due to retransmissions after packet losses.
- 10. DCCP • Unreliable flows of datagrams • Reliable handshakes for connection setup and teardown • Reliable negotiation of options • Mechanisms to avoid holding state for unacknowledged connection attempts and already-finished connections • Congestion control incorporating Explicit Congestion Notification (ECN) [RFC3168] and the ECN Nonce [RFC3540] • Acknowledgement mechanisms communicating packet loss and ECN information. • Optional mechanisms that tell the sending application which data packets reached the receiver, and whether those packets were ECN marked, corrupted, or dropped in the receive buffer • Path Maximum Transmission Unit (PMTU) discovery [RFC1191]. • Two mechanisms of modular congestion control mechanisms : TCP-like Congestion Control [RFC4341] and TCP-Friendly Rate Control (TFRC) [RFC4342].
- 11. DCCP Specifying the client and server ports, the service being requested, and any features being negotiated, including the CCID Exchange DCCP-Data packets, DCCP-Ack packets acknowledging that data, and, optionally, DCCP-DataAck packets containing data with piggybacked Acknowledgements indicating that it is willing to communicate with the client client sends a DCCP-Close packet acknowledging the close server sends a DCCP-CloseReq packet requesting a close server sends a DCCP-Reset packet with Reset Code 1, "Closed", and clears its connection state acknowledges the server's initial sequence number and returns any Init Cookies in the DCCP-Response
- 12. Effort on this Project • Development of DCCP_SIP Agent • SIP_DCCP Application Traffic • Simulation of SIP over DCCP
- 13. DccpSIPAgent Agent • Include all packet structure type in DCCP : • hdr_dccp , hdr_dccpack , hdr_dccpreset ,hdr_dccpreq , hdr_dccpresp, hdr_dccpdata , hdr_dccpdataack , hdr_dccpclose , hdr_dccpclosereq • Addtional packet type structure • hdr_SIP
- 14. NS Application • Additional application traffic sip_dccp for NS2 SIP library • DCCP_SIP_Traffic • INVITE_Timer_ • RINGING_Timer_ • OK_Timer_ • BYE_Timer_ • SIP_DCCP_Traffic • SIP_DCCP_TrafficClass
- 15. Simulation of SIP over DCCP and UDP
- 16. Simulation • System Services Overview – Consist of , 2 SIP Proxy (send and receive SIP message), 1 EXP sorce node, 1 destination node (SINK), 4 node act as a router node • Performace Metrics – Average call setup delay of SIP Signalling flow, Call drop rate of SIP signalling flow • System and Workload Parameter – Type of transport layer protocol ( UDP, DCCP ) , Provides a comparison of performance and behaviour of various transport layer protocols for SIP Signalling – Number of SIP invite packet initiation per second , Provides a comparison of signalling transport protocol impact on call drop avoidance and call setup delay • Factor and Values – Transport layer protocol: UDP, DCCP – Number of call per second: 5, 10, 20, 40 – Simulation Times : 3600
- 17. TCL Configuration Setup Node T01,T02,L01,L02,PR1,PR2,Voice1,Null2, test_1 ,test_2 Setup dccp_test_sink: Agent/Null, attach to test_2 Setup link between Nodes Setup voice_test : Application/Traffic/Exponential, attach to dccp_test Setup dccp_1 and dccp_2: Agent/DCCP/DCCPSIP, attach to PR1 and PR2 Set sip_1 and sip_2: Application/Traffic/SipDccp, attch to dccp_1 and dccp_2 Setup dccp_test : Agent/DCCP, attach to test_1 Connect link between dccp_test and dccp_test_sink Connect link between dccp_1 and dccp_2 at 0.0 generate traffic form application sip_1, sip_2 and voice_test at finishtime stop application sip_1, sip_2 and voice_test flush-trace and close all trace fileExit simulation Set New SimulationSet trace and its files
- 18. Simulation Step
- 19. Result 0.165 0.17 0.175 0.18 0.185 0.19 0.195 5 10 20 40 callsetupdelay(s) number of call(calls/sec) DCCP UDP DCCP UDP N/s CALL RETRAN S CALL RETRAN S 5 18036 0 12038 0 10 35130 0 23512 0 15 69246 0 45889 0 20 134089 0 93038 3325 Call Rate DCCP UDP Average Std Dev Average Std Dev 5 0.18424 0.01173 0.17583 0.01439 10 0.18488 0.01174 0.17664 0.01444 20 0.18642 0.01196 0.17807 0.01450 40 0.18980 0.01242 0.18064 0.01481 70 0.18908 0.01238 - - 80 0.19285 0.01779 - - Call Setup Delay Call Drop Rate
- 20. Discussion • Case of DCCP Call rate 70 and 80 call/sec simulations use 5 minute simulated time, yielding INVITE count 110969 and 125386, respectively, without retransmission. • This result may indicates that the congestion control implemented in DCCP affect this signalling transport protocol to trasnmit SIP messages in case of link congestion. • In case of UDP, the simulation shows that at the Call rate of 20 call/sec, INVITE messages retransmission occures. This indicate that there is a link congestion and UDP do not have any mechanism to detect and deal with it, and continuesly send the SIP message which leads to call drop • This result indicates that UDP has, in average, less call setup delay than DCCP. But DCCP has lower deviation giving better predictability.
- 21. Conclusion • SIP has became the signaling protocol standard on IP based telecommunication services. • In this paper we have shown that DCCP has strong characteristics that make it preferrable as transport medium for SIP compared to UDP. • The characteristics shown here are very low drop rate and lower performance variation. We can not see packet drop even in 80 call/second, twice the rate where UDP drops