Policy-based QoS Framework for Multi-service IP Networks Hoon Lee

Policy-based QoS Framework for Multi-service IP Networks Hoon Lee E-mail: [email protected] Network and Service Assurance Lab. Dept. of Information & Communications Engineering Changwon National University Changwon, Korea

Service Trends: Triple Play Voice Data Video Voice from Phone Data from PC VoD/ TV Videophon e There exists no killer applications! Pack them up!! Triple Play Services: - Italy: FastWeb - Japan: NTT RENA, KDDI, SoftBank - Korea 1. KT: All up prime (Megapass VoIP Videophone Messaging Broadcasting) 2. Dacom: Internet VoIP Broadcasting 3. Hanaro Telecom: 2 Scenarios PSTN: xDSL POTS SkyLife Cable: Cable internet VoIP Broadcasting

Technologies for Internet QoS Speed up & Over-provisioning (QoS-Free) - Current BE service - Applicable to any kind of future applications * Almost zero delay if link speed is in the order of 10s of Mbps Service Differentiation for Priority Traffic SP, CBQ, Hybrid - Wired network: IETF DiffServ MPLS (Priority service Tunneling) Priority: EF AF BE - Wireless ad hoc network: SWAN (Feedback control CAC) Priority: rt traffic BE traffic Policy-based QoS Guarantee - Policies for service differentiation / BW allocation / scheduling / routing

Policy-based QoS Framework of TEQUILA Policy Manageme nt Tool Policy management Policy server Policy consumer From Custom er SLS Subscr. Traffic Estimation Network Provisioning SLS Req. SLS Inv. SLS Management Network monitoring Resource manag. Route manag Performance Manag. Traffic Engineering Static policy (Long-term) Dynamic policy (Short-term)

NTT RENA’s QoS Framework Separation of control and data transfer plane Flexible network control Centralized QoS management e2e QoS Service/Network Control Platform NIB SCS PS From NMS BB PSTN PC PC e2e Optical network Phon e RENATM: Resilient Network Architecture SCS: Session control server BB: Bandwidth broker PS: Policy-server NIB:Network information base Video server Web server Phon e

Policies for IP QoS Principle for IP QoS: Be faithful to IP’s philosophy. - Advantage of IP: Connectionless paradigm Simple & scalable IP QoS Provisioning: via Policy-based networking - Destination-based routing based on OSPF principle - Treat QoS traffic with higher priority than the BE traffic SP does not sacrifice the lower class traffic when the link speed exceeds 10s of Mbps!! - Protection of QoS traffic: Class-based CAC - Network–wide: Interoperation of Policy Server/NMS Dynamic CAC & bandwidth management

Policy-based Networking: Big Picture Phones Best effort IP network VoIP G/W PBX Current AN Route Voice buffer r PCs/Servers Access PremiumCore backbone Router network Rout er Policy server farm QoS Server (SLA) Bandwidth Broker MPLS Tunnel VPN Voice Data Access network Traffic meter Policy NIB PCs/Phones/TEs/ Base Core node: DiffServ-based CBQ PHB-based SchedulingServers MPLS-TE Edge node: SLA negotiation, UPC, Packet classification /QoS mapping, CBQ, Packet-scheduling

Packet Level SLS Servic e type Data servic e Voice servic e Attributes Conventional BE service QoS compatible to PSTN Video servic e New Bundle servic e Interactive Applicatio n services Email, ftp, low quality Internet video telephony, Interactiv e multimedia TV, Videoconfe rencing IP VPN ,www, on-line game, streaming multimedia QoS Requirements (ITU-T) None E2E delay 150ms for 99.99% of packets, PLR 10-3 E2E delay 150ms for 99.99% of packets, PLR 10-4 Minimum contracted BW, E2E delay 1 4sec PLR 10-6

Mapping between DiffServ & MPLS QoS Services Premium service Assured service Better than BE service Best-Effort service DiffServ PHB EF AF 1/2 AF 3/4 BE MPLS Label Platinum Gold Silver/ Bronze Steel ITU-T QoS Class 0/1 2 3 4 5 Typical Applications VoIP VPN Signaling, VoD WWW, telnet, streaming service e-mail

Bandwidth Allocation Alternatives Bandwidth reservation model - Absolute QoS guarantee - Low efficiency - e.g.: IntServ architecture - Application to: Videophone service Bandwidth share with priority scheme model - Statistical QoS guarantee - High utilization - e.g.: DiffServ architecture - Application to: Multi-service

Bandwidth Reservation Model : Videophone Service Architecture ISP LAN Video Phone Internet traffic Cd E-S/W IP Network (DiffServ) Router Cv ? Cv C C: Number of videophone connection (channel) : Bandwidth of a videophone connection E-S/W

Input to The System Parameters: Number of subscribers: M (Tens of thousand) Fraction of active connections at busy hour: (10% 20%) Mean session duration: 1/ ( 1,000seconds) Mean session arrival rate: (0.01 1 ) Session broking probability: (0.5 1%) Bandwidth requirement of a Videophone session: 2Mbps (For basic rate service) (8bits/pixel 250 200pixels/frame 5frames/sec 2Mbps)

Analytic System Model Assumption on the session: Session arrival: Poisson arrival Session duration: Exponential distribution System model: Infinite number of traffic sources Full availability link M/M/c/c Queuing model with C concurrent channels Erlang B-formula for GoS of videophone service C E (C , ) C! C i 0 i . i! Constraint on the Service level: E(C, ) . where (M / )/3600

Results and Discussion Typical Assumptions: M 30,000 residential subscribers 0.1 (Residential 10%, Business 20%) 1/ 1,000 seconds 0.36/0.72 sessions / Busy hour / Person (Residential / Business) 1% 2Mbps (basic rate) Result of computation: Input traffic in Erlang: 300 Erlang Computed number of channel: 323 Channels Required bandwidth: Cv C 323 2Mbps 646Mbps To provide the safety margin, we have to take into account the traffic from alternate route of the neighboring nodes: Cv Final 2 Cv 1.3Gbps Final result.

Comparison: Residential vs. Business When the subscribers are business customers - 0.2 - 0.72 (The offered load increases to 4 times that of the residential subscribers!) Total required bandwidth for a number of subscribers: Number of subscriber Required number of channel (residential / business) Total Required Bandwidth (residential / business) 30,000 323 / 1292 1.3 / 5.2 Gbps 60,000 650 / 2600 2.6 / 10.4 Gbps 90,000 928 / 3712 3.7 / 14.8 Gbps

Bandwidth Share Model : Strict Priority Scheduling Scheme System model: DiffServ-aware MPLS Service model: Strict priority (SP) to voice over data1 over data2 Router model: M/G/1 queue with non-preemptive service Objectives: Evaluation of delay for class1, 2, and 3 packets Voice packet SP Data1 packet C Data 2 packet Our concern: 1. Can we apply the SPSS in a DiffServ router for BcN? 2. How about the behavior of delay with respect to the system parameters?

System Model System parameters: - Mean arrival rate for voice/data1/data2: 1, 2 , 3 - Mean service time for voice/data1/data2 : 1/ 1, 1/ 2 , 1/ 3 - Second moment of service time: E[ k2],k 1,2,3 - Offered load for voice/data1/data2 : 1, 2, 3 - Link capacity: C Source models: - Voice: Poisson arrival, fixed packet size - Data1 & data2: Poisson arrivals, Pareto distributions

Delay Performance Mean waiting times for M/G/1 queue with SP service: 3 2 R E [ k ] k 1 k W1 2(1 1 ) W2 R , (1 1 )(1 1 2 ) W3 R , (1 1 2 )(1 1 2 3 ) Mean waiting time for M/G/1 queue with FIFO service: WFIFO 2 1 2 1 CS2 ( ). CS squared coefficient of variation for 1 2 service time of a packet

Numerical Experiments Source traffic profile: - Voice source: G.711 Voice coder, 216bytes - Data source: Ethernet frame, Pareto distribution, m F (l ) Pr{L l} 1 ( ) , (l m, 0). l Minimum packet size, m: 500 1500bytes Tail index: 3 Link capacity per output port: 1M, 10M, 100Mbps

Traffic Load Type Load Type 1 2 3 A 0.1 0.4 0.4 0.9 B 0.3 0.3 0.3 0.9 C 0.5 0.2 0.2 0.9 D 0.7 0.1 0.1 0.9 Light-voice Heavy-data Heavy-voice Light-data

Waiting Time of Voice Packets for Different Link Capacities m2 500bytes, m3 1500bytes Under SP scheduling scheme, delay of voice packet is almost negligible for high-speed links!

Waiting Time of Voice Packets for Different Service Schemes m2 m3 1,000 bytes, C 1Mbps The conventional wisdom of “SP isolates voice traffic from non- voice traffic” does not hold! This is more evident for the WFQ-families.

Delay Performance of Data Traffic Performance comparison between different classes: W2 1 . W1 1 1 2 W3 1 1 . W2 (1 1 2 3 ) 1 0.2 1 0.4 1 0.2 2 0.4 2 0.2

Summary Policy is important for QoS provisioning in future Internet. Network provisioning is dependent on the policy. Reservation model over-estimates the network resources. Shared bandwidth model will prevail. Accurate dimensioning of network resources saves cost.

References [Lee] Hoon Lee, “Strategies for the construction of Policy-based managed IP QoS”, Final Report of NCA II-RER-04041, November 30, 2004. [Lee] Hoon Lee et al., “Dimensioning NGN for QoS guaranteed voice services”, Jr. of IEEK, Vol. TC-40, No.12, December 2003. [Lee] Hoon Lee, “Delay analysis of DiffServ/MPLS network”, Industrial Mathematics Initiative 2004, August 26-28, Korea. [Lee] Hoon Lee et al., “Delay performance of non-realtime services for the strict priority scheduling scheme”, Jr. of the research institute of industrial technology, Vol.18, May 2004. [Trimintzios] P. Trimintzios et al., An architectural framework for providing QoS in IP differentiated services networks, TEQUILA Project report.