Utilitybased Bandwidth Allocation for TriplePlay Services Changbin Liu

Utility-based Bandwidth Allocation for Triple-Play Services Changbin Liu, Lei Shi, Bin Liu Department of Computer Science and Technology, Tsinghua University Proceedings of the Fourth European Conference on Universal Multiservice Networks (ECUMN’ 07) Chen Bin Kuo (20077202) Young J. Won (20063292)

Outline �Introduction �NGN traffic classifications and their utility functions �Network utility maximization (NUM) �Numeric results and analysis �Discussion �Conclusion DPNM Lab. 3/8/2021 2

Introduction �Next generation network (NGN) must natively support triple-plays. �How to schedule traffic and allocate bandwidth at both backbone and access links. �Designing a scheduling (bandwidth allocation) algorithm is exactly the issue this paper tries to settle. DPNM Lab. 3/8/2021 3

Introduction (contd. ) �In industry designing NGN [13][14], the strictpriority scheduling is mostly adopted. Rigidly favors the voice and video traffic without flexibility �Utility-based solutions Shenker [1] discussed traffic classifications in IP network from the viewpoint of user utility Kelly et al. [5][6] applying utility-based methods to scheduling and bandwidth allocation in the objective of Network Utility Maximization (NUM) DPNM Lab. 3/8/2021 4

Introduction (contd. ) �No single work has emphasized on the practical issue of scheduling triple-play services under the background of NGN. �Translating this issue into a nonlinear maximization problem with inequality constraints. DPNM Lab. 3/8/2021 5

NGN traffic classifications and their utility functions a) b) c) d) e) Vo. IP traffic IPTV traffic TCP elastic traffic HTTP traffic Other UDP traffic

NGN traffic classifications and their utility functions �Due to remarkable distinction of Qo. S requirements in NGN Classifying NGN traffic into five categories �User utility function is introduced To measure network performance and user satisfaction degree Determined by the Qo. S metrics received in the user end Including packet delay, jitter and loss rate DPNM Lab. 3/8/2021 7

Vo. IP traffic �Sensitive to packet delay and loss caused by bandwidth insufficiency Utility function falls into the category of hard real-time kind [1][2][10], with a minimal bandwidth requirement of Bmin 1 DPNM Lab. 3/8/2021 8

IPTV traffic �Utility function is similar to Vo. IP’s but with some differences Tolerate occasional delay-bound violations and packet drops Minimal encoding rate, denoted as Bmin 2 is independent of network congestion Logistic model is used DPNM Lab. 3/8/2021 9

TCP elastic traffic �Generated by delay-tolerant TCP applications Such as file transfer and email �Utility function have been studied by Kelly et al. [6] and other researchers [11][12] DPNM Lab. 3/8/2021 10

HTTP traffic �TCP traffic which concerns packet delay Mainly contains the HTTP traffic generated by web services �Utility function is different from TCP elastic traffic, has a minimum tolerable bandwidth Bmin 4 DPNM Lab. 3/8/2021 11

Other UDP traffic �DNS packets, other streaming media traffic, and online gaming traffic [17][18] Delay-sensitive Every application type has a utility function �The shape of utility function resembles IPTV traffic DPNM Lab. 3/8/2021 12

Traffic setting Bmin Vo. IP IPTV 64 Kbps 100 Kbps Bmax 10 Mbps ɛ TCP elastic HTTP 24 Kbps 10 Mbps 0. 001 DPNM Lab. UDP 500 Kbps 0. 001 3/8/2021 13

Network utility maximization a) b) KKT method Lagrange multipliers method without KKT conditions

Network utility maximization �Based on NGN traffic’s utility functions, we can solve the congestion-phased bandwidth allocation issue while conforming to NUM. �Total utility gained on the link is: �Bandwidth allocation is restricted by: N : the number of NGN users utilizing this link pi : traffics classes C : the bandwidth of a link (set to 10 Gbps) DPNM Lab. 3/8/2021 15

KKT method �Lagrange Multiplier method with KKT (Karush-Kuhn -Tucker) conditions �Solving the nonlinear optimization problem �Accurate and comprehensive solution requires substantial complicated computations Applying simplified form which is enough to ravel NUM problem for triple-plays DPNM Lab. 3/8/2021 16

KKT method (contd. ) �Observing NGN traffic’s utility functions Vo. IP/IPTV/other UDP traffic’s utility functions are relatively smoother in some points It is not cost-effective to allocate bandwidth to Vo. IP/IPTV/other UDP traffic without booming the utility �Turning point (TP) IPTV HTTP Bandwidth DPNM Lab. 3/8/2021 17

Lagrange multiplier method �After finding the TP, we can apply the Lagrange Multipliers method without KKT conditions to solve the NUM problem in (10) �Subject to: DPNM Lab. 3/8/2021 18

Numeric results and analysis a) b) Data-dominated network IPTV-dominated network

Numeric results and analysis �Two network scenarios Current Internet, where HTTP and TCP elastic traffic still dominate the volume Prospective NGN, where the emerging services, especially the IPTV traffic, will dominate the network �For each scenario, calculate in two situations Maximal Utility Equalization (MUE) Maximal Utility In-equalization (MUI) V 1 (Vo. IP) V 2 (IPTV) V 3 (TCP elastic) V 4 (HTTP) V 5 (other UDP) MUE 1 1 1 MUI 1 9 1 1. 5 2 DPNM Lab. 3/8/2021 20

Numeric results and analysis �Data-dominated network According to recent trace observation [15] Vo. IP Traffic proportions 10% �IPTV-dominated Vo. IP Traffic proportions 10% DPNM Lab. IPTV TCP elastic 10% HTTP other UDP 50% 20% network IPTV 50% TCP elastic 10% HTTP other UDP 20% 3/8/2021 10% 21

Data-dominated network DPNM Lab. 3/8/2021 22

IPTV-dominated network DPNM Lab. 3/8/2021 23

Discussion �Previous bandwidth allocation schemes for triple-play services mostly adopt the strict-priority scheduling Highest priority to Vo. IP traffic Second highest priority to IPTV and lowest priority to others �In this paper Highest priority to Vo. IP traffic Assigning IPTV traffic with second-highest priority is not well supported from the objective of NUM Suggesting that ISP charges more about IPTV services (future work) DPNM Lab. 3/8/2021 24

Conclusion �Studied the problem of scheduling and bandwidth allocation for triple-play services in the objective of NUM. �Presenting theoretical method to compute bandwidth allocation results �Results: Vo. IP and other low-throughput UDP traffic can always be guaranteed of sufficient bandwidth As congestion becomes severer, IPTV’s bandwidth decreasing quickly TCP elastic and HTTP traffic experience exponential bandwidth degradations when congestion degree increases DPNM Lab. 3/8/2021 25

Q&A Thanks for your attention!