Xin Wang Thesis Defense Adviser Henning Schulzrinne Internet

Xin Wang (Thesis Defense) Adviser: Henning Schulzrinne Internet Real -Time Laboratory Columbia University http: //www. cs. columbia. edu/~xinwang

Scope of this Talk n Main work u n Resource Negotiation, Pricing, and Qo. S for Adaptive Multimedia Applications Other work u Measurements and Analysis of LDAP Performance u IP Multicast Fault Recovery in PIM over OSPF 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 2

Outline n n n n Introduction A Resource Negotiation And Pricing Resource protocol: RNAP Negotiation Framework Pricing models User adaptation models Test-bed demonstration of Resource Negotiation Framework Simulation and discussion of Resource Negotiation Framework Conclusion and future work 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 4

Is Simple Over-Provisioning Enough? n Current Internet: u u n Growth of new IP services and applications with different bandwidth and quality of service requirements Revenue from the traditional connectivity services is declining New services present opportunities and challenges u Even though average bandwidth utilization is low, congestion can happen; access links get congested frequently u Wireless bandwidth is even more scarce u Bandwidth prices are not dropping rapidly u No intrinsic upper limit on bandwidth use Option - manage the existing bandwidth better, with a service model which uses bandwidth efficiently. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 5

A More Efficient Service Model n Quality of Service (Qo. S) u Condition the network to provide predictability to an application even during high user demand u Provide multiple levels of services u Problems: signaling to facilitate service negotiation; charging scheme to support differentiated services n Application adaptation u Source rate adaptation based on network conditions congestion control and efficient bandwidth utilization u Problems: è How adaptive applications work with Qo. S-assured services? How to motivate an application to adapt? 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 6

Design Goals n n n Develop an efficient service model which combines Qo. S assurance with user rate adaptation Increase service value to the users through greater choices over price and quality, improved connectivity, and expected Qo. S Reduce network provision complexity, improve network efficiency and increase revenue to the providers; allow network operator to create different trade-offs between blocking admissions and raising congestion prices 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 7

Related Work n Signaling for Resource Reservation u RSVP, YESSIR, SIBBS (Simple Inter-domain Bandwidth Broker Signaling protocol) u Problems èNo support for service selection and negotiation èNo support for short-term resource commitment and dynamic resource negotiation èRestricted to either sender-driven or receiver-driven èNo support for pricing and billing 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 8

Related Work (cont’d) n Adaptive Internet Multimedia Applications u Sender-driven, receiver driven, or transcoder-based u Problems èFairness is one issue. TCP friendly adaptation may lead to unpleasantly abrupt changes in quality; buffering smoothes the abrupt change at the cost of high delay èNo mechanism for rate adaptation in Qo. S-enhanced environment èNo motivation for user adaptation 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 9

Related Work (cont’d) n Pricing and Billing in the Network u Total user benefit maximization based on welfare theory èproblems: rely on a centralized optimization process for total user utility maximization; assume knowledge (user’s truthful revelation ) of utility functions u Pricing for congestion control èproblems: rely on well-defined source statistical model; not consider congestion control during sessions u Pricing in multi-service environment èproblems: Qdlyzko’ 99 rely purely on pricing and user behaviors, without any service assurance; Kumaran’ 99 is restricted to special admission control algorithm 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 10

Related Work (cont’d) u Auction Mechanism èproblems: signaling bursts, set-up delay, uncertainty of connection availability, user response to fluctuations in price u Signaling Support for Pricing and Charging èVery limited work in this area èKarsten’ 98: estimated the cost for user request, no price quotation; restricted to Int. Serv 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 11

Thesis Contributions n Propose a Resource Negotiation And Pricing protocol: RNAP u u u n n n Enables user and network (or two network domains) to dynamically negotiate multiple services Supports price and charge collation, auction bids and results distribution Allows for service predictability, multi-party negotiation Designed to be scalable and reliable Can be embedded in other protocols, or implemented independently Enables differential charging for supporting differentiated services, reflecting the service cost and long-term user demand Support short-term resource commitment for better response to user demand network conditions, and more efficient resource usage; congestion pricing to motivate user adaptation Develop reference user adaptation model 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 12

Thesis Contributions (cont’d) n n Demonstrate negotiation framework on test-bed and simulation Show significant advantages relative to static resource allocation and fixed pricing using simulations 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 13

Outline n n n n Introduction A Resource Negotiation And Pricing Resource protocol: RNAP Negotiation Framework Pricing models User adaptation model Test-bed demonstration of Resource Negotiation Framework Simulation and discussion of Resource Negotiation Framework Conclusion and future work 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 14

Protocol Architectures: Centralized (RNAP-C) Host Resource Negotiator NRN HRN RNAP Messages NRN Network Resource Negotiator NRN HRN Access Domain - A Edge Router Access Domain - B Internal Router 12/11/2021 Transit Domain Intra-domain messages Xin Wang, Henning Schulzrinne, Columbia University 15

Protocol Architectures: Distributed (RNAP-D) RNAP Messages HRN LRN Local Resource Negotiator LRN LRN Access Domain - A LRN LRN LRN HRN LRN Edge Router Access Domain - B Internal Router 12/11/2021 LRN Transit Domain Xin Wang, Henning Schulzrinne, Columbia University 16

RNAP Messages Query: Inquires about available services, prices Periodic negotiation Query Quotation: Specifies service availability, accumulates service statistics, prices Commit Reserve: Requests services and resources, Modifies earlier requests Reserve Quotation Reserve specific price or denies it. Commit Close Release 12/11/2021 Commit: Confirms the service request at a Close: Tears down negotiation session Release: Releases the resources Xin Wang, Henning Schulzrinne, Columbia University 17

Message Aggregation (RNAP-D) Turn off router alert Turn on router alert Edge Routers Sink-tree-based aggregation 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 18

Message Aggregation (RNAP-D) Turn on router alert Turn off router alert Sink-tree-based aggregation 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 19

Block Negotiation (Network-Network) Bandwidth Aggregated resources are added/removed in large blocks to minimize negotiation overhead and reduce network dynamics time 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 20

Outline n n n n Introduction A Resource Negotiation And Pricing Resource protocol: RNAP Negotiation Framework Pricing models User adaptation Test-bed demonstration of Resource Negotiation Framework Simulation and discussion of Resource Negotiation Framework Conclusion and future work 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 21

Two Volume-based Pricing Strategies n Fixed-Price (FP): fixed unit volume price congestion: higher blocking rate OR higher dropping rate and delay u During n Congestion-dependent-Price (CP): FP + congestionsensitive price component u During congestion: users have options to maintain service by paying more OR reducing sending rate OR switching to lower service class u Overall reduced rate of service blocking, packet dropping and delay 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 22

Proposed Pricing Strategies n Holding price and charge: based on cost of blocking other users by holding bandwidth even without sending data u n Usage price and charge: maximize the provider’s profit, constrained by resource availability u u n phj = j (pu j - pu j-1) , chij (n) = ph j r ij (n) j max [Σl x j (pu 1 , pu 2 , …, pu. J ) puj - f(C)], s. t. r (x (pu 2 , …, pu. J )) R cuij (n) = pu j v ij (n) Congestion price and charge: drive demand to supply level (tatonnement process or auction) 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 23

Usage Price for Differentiated Service n Usage price based on cost of class bandwidth: u lower target load (higher Qo. S) -> higher per-unit bandwidth price n Parameters: u u u n n n pbasic rate for fully used bandwidth j : expected load ratio of class j xij : effective bandwidth consumption of application i Aj : constant elasticity demand parameter Price for class j: puj = pbasic / j Demand of class j: xj ( puj ) = Aj / puj Effective bandwidth consumption: xe j ( puj ) = Aj / ( puj j ) Network maximizes profit: u max [Σl (Aj / pu j ) pu j - f (C)], puj = pbasic / j , s. t. Σl Aj / ( pu j j ) C Hence: pbasic = Σl Aj / C , puj = Σl Aj /(C j) 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 24

Congestion Price: First Mechanism Tatonnement n Tatonnement process (CPA-TAT): u Congestion charge proportional to excess demand relative to target utilization u pc j (n) = min [{pcj (n-1) + j (Dj, Sj) x (Dj-Sj)/Sj, 0 }+, pmaxj ] u ccij (n) = pc j v ij (n) 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 25

Congestion Price: Second Mechanism M-bid Second-price Auction n Auction models in literature: u u u n Assume unique bandwidth/price preference, one bid Service uncertainty: does not know about high demand until rejected Other issues: setup delay, signaling burst, user response to auction results M-bid auction Model u u u Reduce uncertainty & provide complete user preferences: user bids (bandwidth, price) for a number of bandwidths, bids obtained by sampling utility function. Congestion price: charges highest rejected bid price Bandwidth allocations: higher valued bandwidth get allocated first; more users served Congestion control: auctions for period of time Reduce set-up delay: inter-auction admission 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 26

Example of M-bid Auction n Total capacity 70, congestion price is 2 Bid Bandwidth Bidder Bid Selection Bid Price 5 4 4 3. 5 3 2 10 10 15 20 25 1 2 1 3 2 30 3 Cutoff Congestion Price 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 27

Outline n n n n Introduction A Resource Negotiation And Pricing Resource protocol: RNAP Negotiation Framework Pricing models User adaptation model Test-bed demonstration of Resource Negotiation Framework Simulation and discussion of Resource Negotiation Framework Conclusion and future work 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 28

Rate Adaptation of Multimedia System n Utility/cost/bud n Gain optimal perceptual value of the system based on the network conditions and user profile Utility function: users’ preference or willingness to pay U 1 U 3 get 12/11/2021 U 2 Cost Budget Bandwidth Xin Wang, Henning Schulzrinne, Columbia University 29

Example Utility Function n Utility is a function of bandwidth at fixed Qo. S example utility function: U (x) = U 0 + log (x / xm) u U 0 : perceived (opportunity) value at minimum bandwidth u : sensitivity of the utility to bandwidth u An n Function of both bandwidth and Qo. S (x) = U 0 + log (x / xm) - kd d - kl l , for x xm u kd : sensitivity to delay u kl : sensitivity to loss u. U 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 30

Rate-Adaptation Models n Model 1: User adaptation under CPA-TAT (tatonnementbased pricing) u Gaining optimal transmission rate by optimizing perceived surplus of the multimedia system subject to budget and application requirements èU = Σi Ui (xi (Tspec, Rspec)] è max [Σl Ui (xi ) - Ci (xi) ], s. t. Σl Ci (xi) b , xmini xmaxi è Determine optimal Tspec and Rspec u n With the example utility functions, resource request of application i: è Without budget constraint: x i = i / pi è With budget constraint: x i = bi / pi, with b i = b ( i / Σl k) Model 2: User adaptation under CPA-AUC (second-price auction) u Adapt rate based on allocated bandwidth/Qo. S 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 31

Outline n n n n Introduction A Resource Negotiation And Pricing Resource protocol: RNAP Negotiation Framework Pricing models User adaptation model Test-bed demonstration of Resource Negotiation Framework Simulation and discussion of Resource Negotiation Framework Conclusion and future work 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 32

Testbed Architecture RAT VIC n Mbus n HRN RN AP NRN n n 12/11/2021 Demonstrate functionality and performance improvement: u blocking rate, loss, delay, price stability, perceived media quality Host u HRN negotiates for a system u Host processes (HRN, VIC, RAT) communicate through Mbus Network u Router: Free. BSD 3. 4 + ALTQ 2. 2, CBQ extended for Diff. Serv u NRN: (1) Process RNAP messages; (2) Admission control, monitor statistics, compute price; (3) At edge, dynamically configure the conditioners and form charge Inter-entity signaling: RNAP Xin Wang, Henning Schulzrinne, Columbia University 33

Outline n n n n Introduction A Resource Negotiation And Pricing Resource protocol: RNAP Negotiation Framework Pricing models User adaptation Test-bed demonstration of Resource Negotiation Framework Simulation and discussion of Resource Negotiation Framework Conclusion and future work 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 34

Simulation Design n Performance comparison: fixed price policy (FP) vs. congestion price based adaptive service (CPA) u n n n Three groups of experiments: effect of traffic load, admission control, and load balance between classes Weighted Round Robin (WRR) scheduler Three classes: EF, AF, BE u u u n loss, delay, blocking rate, user benefit, network revenue, stability EF: load threshold 40%, delay bound 2 ms, loss bound 10 -6 AF: load threshold 60%, delay bound 5 ms, loss bound 10 -4 BE: load threshold 90%, delay bound 100 ms, loss bound 10 -2 Sources: mix of on-off traffic and Pareto on-off traffic 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 35

Simulation Architecture Topology 1 (60 users) 12/11/2021 Topology 2 (360 users) Xin Wang, Henning Schulzrinne, Columbia University 36

Effect of Traffic Load CPA maintains the traffic load at the targeted level, meets the expected performance bounds 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 37

Effect of Admission Control Admission control is important in maintaining the expected performance of a class. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 38

Effect of Admission Control (cont’d) With admission control, the dynamics of the network price can be better controlled. Coupled with user adaptation, the blocking rate of CPA is up to 30 times smaller than that of FP. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 39

Effect of Admission Control (cont’d) CPA allows for higher network revenue and user benefit. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 40

Load Balance Between Classes Even when a small portion of users (15%) select other service classes, the performance of the over-loaded class is greatly improved. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 41

Other Results n Users with different demand elasticity share bandwidth n n n proportional to their willingness to pay Even a small proportion of adaptive users (e. g 25%) results in a significant performance improvement for the entire user population (18% improvement) Performance of CPA further improves as the network scales and more connections share the resources Both M-bid auction and tatonnement process can be used to calculate the congestion price; auction gives higher perceived user benefit and network utilization at cost of implementation complexity and setup delay 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 42

Conclusions n n Proposed a dynamic resource negotiation framework consisting of: A Resource Negotiation And Pricing protocol (RNAP) , a rate and Qo. S adaptation model, and a pricing model RNAP: supports dynamic service negotiation Pricing models: based on resources consumed by service class and long-term user demand; including congestionsensitive component to motivate user demand adaptation Performance: u u u Effectively restricts load to targeted level and meet service assurance Provide lower blocking rate, higher user satisfaction and network revenue Admission control and inter-service class adaptation give further improvements in blocking rate and price stability 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 43

Future Work n n n Interaction of short-term resource negotiation with longerterm network provision A light-weight resource management protocol Cost distribution in Qo. S-enhanced multicast network Pricing and service negotiation in the presence of alternative data paths or competing networks User valuation models for different Qo. S Resource provisioning in wireless environment 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 44

Joint work with Dilip Kandlur, and Dinesh Verma (IBM Research)

Motivation and Experiment n n n Lightweight Directory Access Protocol (LDAP): widely used, but little study of performance Related work: Mindcraft’ 98 treated LDAP server as black box, did not study the influence of system components Thesis contributions: u u n Developed a benchmark tool to analyze the performance of LDAP Provided guidelines for the configuration of LDAP client and server Suggested schemes for LDAP performance improvement Is the first effort that addressed the performance issues and configuration issues for the widely used LDAP Usage context for thesis: management for network resources, pricing policies 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 46

Experimental Setup & General Results n Experimental setup: u u n Hardware: server- dual Ultra-2 processors, 200 MHz CPUs, 256 Mb memory; Clients- Ultra 1, 170 MHz CPU, 128 MB memory; 10 Mb/s Ethernet LDAP server: Open. LDAP 1. 2, Berkeley DB 2. 4. 14 Search filter: interface address, and corresponding policy object Default parameters: 10, 000 entries, entry size 488 bytes General results: u u response latency 8 ms up to 105 requests/second Maximum throughput 140 requests/second 5 ms processing latency - 36% from backend, 64% from front end Connect time dominates at high load, and limits the throughput 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 47

Scalability of the Performance n Scaling with Directory Size - determined by back-end processing u u n In memory operation, 10, 000 -> 50, 000: processing time increases 60%, throughput reduces 21%. Out-of-memory, 50, 000 ->100, 000: processing time increases another 87%, and throughput reduces 23%. Scaling with Entry Size (488 ->4880 bytes): u u In-memory, mainly increase in front-end processing, i. e. , time for ASN. 1 encoding. Processing time increases 8 ms, 88% due to ASN. 1 encoding, and throughput reduces 30%. Out-of-memory, throughput reduces 70%, mainly due to increased data transfer time. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 48

Performance Improvement n n Disabling Nagle algorithm: reduces latency about 50 ms Entry Caching: u n CPU: u n for 10, 000 entry directory, caching all entries gives 40% improvement in processing time, 25% improvement in throughput For in-memory operation, dual processors improve performance by 40% Connection Re-use: u 60% performance gain when connection left open 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 49

Joint work with Chien-ming Yu (Microsoft) Paul Stirpe and Wei Wu (Reuters)

Motivations n n n Many IP multicast applications require high availability, especially mission-critical real-time data A lot of work on reliable multicast, but little work on multicast fault tolerance Study failure recovery in a complete architecture: IGMP + OSPF (unicast) + PIM (multicast) Focus: the interplay of underlying protocols; the interactions of failure recovery, between routers, links, WAN and LAN Method: quantitative analysis; simulation over OPNET; study failure recovery and implementation issues on test-bed using Cisco routers 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 51

Results n General observations u Channel recovery time: dominated by unicast table -construction time. u Protocol control loads: re è PIM-DM control load increases proportionally with the redundancy factor and decreases inversely with the percentage of receivers è Below certain time interval threshold, OSPF load is dominated by Hello messages and increases proportionally as OSPF Hello interval decreases è Neither PIM nor OSPF has high control traffic during failure recovery. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 52

Results (cont’d) n PIM Enhancement for Fault Recovery u Fast recovery from Dedicated Router (DR) failure: reduce Hello-Holdtime to detect neighbor failure faster; Backup DR; IGMP group information caching in all LAN routers (reloading group membership information leads to minutes of delay) u Fast recovery from last-hop router failure: DR records the last-hop router address, actively recover, instead of waiting for an IGMP report to reactivate its oif to the LAN (up to minutes of delay) u Use interrupts instead of polling to reduce delay 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 53

Some References n n n X. Wang, H. Schulzrinne, “Auction or Tatonnement - Finding Congestion Prices for Adaptive Applications”, submitted. X. Wang, H. Schulzrinne, “Pricing Network Resources for Adaptive Applications in a Differentiated Services Network, ” In Proceeding of INFOCOM'2001, April 22 -26, Anchorage, Alaska. X. Wang, H. Schulzrinne, “An Integrated Resource Negotiation, Pricing, and Qo. S Adaptation Framework for Multimedia Applications, ” IEEE JSAC, vol. 18, 2000. Special Issue on Internet Qo. S. X. Wang, H. Schulzrinne, “Comparison of Adaptive Internet Multimedia Applications, ” IEICE Transactions on Communications, Vol. E 82 -B, No. 6, pp. 806 --818, June 1999. X. Wang, H. Schulzrinne, D. Kandlur, D. Verma, “Measurement and Analysis of LDAP Performance, ” International Conference on Measurement and Modeling of Computer Systems (ACM SIGMETRICS'2000). X. Wang, H. Schulzrinne, C. Yu, P. Stirpe, W. Wu, “IP Multicast Fault Recovery in PIM over OSPF, ” In 8 th International Conference on Network Protocols (ICNP'00), 2000. Also appears at ACM SIGMETRICS’ 2000 as short paper. 12/11/2021 Xin Wang, Henning Schulzrinne, Columbia University 54

Questions and Answers Thanks !