1. Network Economics and ISP Settlement Problems Tianbai Ma Department of Electrical Engineering Columbia University

2. Outline Network economics research and problems Current research: two problems My research – past and future

3. The Internet is operated by hundreds of interconnected Internet Service Providers (ISPs). An ISP is a autonomous business entity –Provide Internet services. –Common objective: to make profit. Building blocks of the Internet: ISPs

4. Three types of ISPs Eyeball ISPs: –Provide Internet access to individual users. –E.g. PCCW Content ISPs: –Provide contents on the Internet. Transit ISPs: –Tier 1 ISPs: global connectivity of the Internet. –Provide transit services for other ISPs.

5. An motivating example Win-win: –Content provider finds customers –Customers get content –ISPs obtain revenue Volume-based charge Fixed charge

6. What happens with P2P traffic? Win-lose under P2P paradigm: –Content provider pays less –Customers get faster download –Content ISP obtains less revenue –Eyeball ISPs handle more traffic Consequence: –ISPs want to avoid P2P traffic

7. Engineering solutions and … beyond Consequence: –ISPs want to avoid P2P traffic Proposed engineering solutions: –Strategy of ISPs: Limit P2P traffic rate –Counter-strategy of users: Encryption –Counter-strategy of ISPs: Deep packet inspection A new/global angle: Network EconomicsA new/global angle: Network Economics –Goal: a win-win/fair solution –Issue: how to design algorithms/protocols to achieve? –An exciting and upcoming area: Inter-disciplinary ： CS, Economics, Operation Research, Policy … Stanford, CMU, UC Santa Cruz and etc. build new courses/department.

8. Other important problems that can be looked at 1. Network Neutrality Debate: Content-based Service Differentiation ? YesNo 2. Network Balkanization: Break-ups of Connected ISPs 15% of Internet unreachable Level 3Cogent Legal/regulatory policy for the Internet industry. Not only a technical/operation problem, but also economic issues. Some ISPs dominate or ISPs cannot survive. Suppress the development of the Internet. Threatens the global connectivity of the Internet.

9. Other important problems that can be looked at A new/global angle: Network EconomicsA new/global angle: Network Economics –Goal: a win-win/fair solution –Issues: how to design algorithms/protocols to achieve?

10. Problems we try to solve win-win/fairProblem 1: Find a win-win/fair profit-sharing solution for ISPs. Challenges –What’s the solution? ISPs don’t know, even with best intentions. –How to find it? Complex ISPs structure, computationally expensive. –How to implement? Need to be implementable for ISPs.

11. One content and one eyeball ISP Profit V = total revenue = content-side + eyeball-side Win-win/fair profit sharing: How to share profit? -- the baseline case

12. How to share profit? – two symmetric eyeball ISPs Symmetry: same profit for symmetric eyeball ISPs Efficiency: summation of individual ISP profits equals v Fairness: same mutual contribution for any pair of ISPs Unique solution (Shapley value) Win-win/fair properties:

13. History and properties of the Shapley value Shapley Value EfficiencySymmetryFairness Myerson 1977 EfficiencySymmetryDummyAdditivity Shapley 1953 EfficiencySymmetryStrong Monotonicity Young 1985 What is the Shapley value? – A measure of one’s contribution to different coalitions that it participates in.

14. How to share profit? -- n symmetric eyeball ISPs Theorem: the Shapley profit sharing solution is

15. Results and implications of profit sharing More eyeball ISPs, the content ISP gets larger profit share. –Users may choose different eyeball ISPs; however, must go through content ISP, –Multiple eyeball ISPs provide redundancy ， –The single content ISP has leverage. Content’s profit with one less eyeball: The marginal profit loss of the content ISP: If an eyeball ISP leaves –The content ISP will lose 1/n 2 of its profit. –If n=1, the content ISP will lose all its profit.

16. Profit share -- multiple eyeball and content ISPs Theorem: the Shapley profit sharing solution is

17. Results and implications of ISP profit sharing Intuition –The larger group of ISPs provides redundancy. –The smaller group of ISPs has leverage. Each ISP’s profit share is –Inversely proportional to the number of ISPs of its own type. –Proportional to the number of ISPs of the opposite type.

18. Profit share -- eyeball, transit and content ISPs Intuition –The larger group of ISPs provides redundancy. –The smaller group of ISPs has leverage. Theorem: the Shapley profit sharing solution is

19. Profit share – general topologies 1. Shapley values under sub-topologies: 2. Whether the profit can still be generated: Theorem: Dynamic Programming ！

20. ISP business practices: a macro perspective 2. Zero-Dollar Peering 1. Customer-Provider Settlement Two forms of bilateral settlements: Provider ISPs Customer ISPs $$$

21. Implications – the value chain $$$ $$ $ $ $ $ $$$ $ $ $ $ $ $ Content-side Revenue Eyeball-side Revenue

22. Implications – the value chain Shapley value suggested revenue flows: –Content-side revenue: Content  Transit  Eyeball –Eyeball-side revenue: Eyeball  Transit  Content $$$ $$ $ $$$ $$ $ Content-side Revenue Eyeball-side Revenue

23. Implications – equivalent bilateral settlements $$$ $$ $ $$$ $$ $ When CR ≈ BR, bilateral implementations: –Customer-Provider settlements (Transit ISPs as providers) –Zero-dollar Peering settlements (between Transit ISPs) –Stable structure when local ISPs are homogeneous. Providers Customers Zero-dollar Peering Content-side Revenue Eyeball-side Revenue

24. Implications – equivalent bilateral settlements $$$ $$ $ $$$ $$ $ $$$ $$ $ $$$ $$ $ If CR >> BR, bilateral implementations: –Reverse Customer-Provider –Reverse Customer-Provider (Transits compensate Eyeballs) –Paid Peering –Paid Peering (Content-side compensates eyeball-side) –New settlements will emerge to maintain a stable structure. CustomerProvider Paid Peering Eyeball-side Revenue Content-side Revenue

25. Problems we try to solve win-win/fairProblem 1: Find a win-win/fair profit-sharing solution for ISPs. Challenges –What’s the solution? ISPs don’t know, even with best intentions. –Answer: The Shapley value. –How to find it? Complex ISPs structure, computationally expensive. –Result: Closed-form solution and Dynamic Programming. –How to implement? Need to be implementable for ISPs. –Implication: New bilateral settlements.

26. Recap: ISP practices from a macro perspective 2. Zero-Dollar Peering 1. Customer-Provider Settlement Two current forms of bilateral settlements: 3. Inverse Customer- Provider Settlement 4. Paid Peering Two new forms of bilateral settlements:

27. Problems we try to solve win-win/fairProblem 1: Find a win-win/fair profit-sharing solution for ISPs. Challenges –What’s the solution? ISPs don’t know, even with best intentions. –Answer: The Shapley value. –How to find it? Complex ISPs structure, computationally expensive. –Result: Closed-form solution and Dynamic Programming. –How to implement? Need to be implementable for ISPs. –Implication: New bilateral settlements. efficient/optimalProblem 2: Encourage ISPs to operate at efficient/optimal points. Challenges –How to induce good behavior? Selfish behavior may hurt other ISPs. –What is the impact of the entire network? Equilibrium is efficient?

28. Current ISP Business Practices: A Micro Perspective Three levels of ISP decisions Interconnecting decision E Routing decisions R (via BGP) Bilateral settlements  Shortest Path Routing Hot-potato Routing Source Destination Zero-Dollar Peering Customer-Provider Settlements Provider ISP Customer ISP Settlement  affects E, R provider charges high Interconnection withdrawal Route change

29. Peering links at both coasts Locally connect to both backbone ISPs RouteTopology An ideal case of the ISP decisions Two backbone ISPs Two local ISPs End-to-end service generates revenue Routing costs on links Backbone ISP 1 Backbone ISP 2 Local ISP 1 Local ISP 2 Fixed Revenue A simple example: Well-connected topology W = 1

30. Routing cost model x 2 /4 Assumptions (based on current practices) –Routing costs on links, e.g. bandwidth capacity and maintenance. –Going across the country is more expensive. –More expensive when link is more congested. Costs increase with link loads –Standard queueing theory results. –Capital investment for upgrades.

31. x 2 2 /4 x 7 2 /4 x 4 2 /8 x 5 2 /8 x 1 2 /16 x 6 2 /16 x 3 2 /16 x 8 2 /16 1/2 RouteTopology Global Min Cost An ideal case of the ISP decisions Two backbone ISPs Two local ISPs End-to-end service generates revenue Routing costs on links Backbone ISP 1 Backbone ISP 2 Local ISP 1 Local ISP 2 Fixed RevenueCost | Profit A simple example: Well-connected topology W = 1 We normalize the total required traffic load to be 1. MIN routing cost and MAX profit

32. x 2 2 /4 x 7 2 /4 x 4 2 /8 x 5 2 /8 x 1 2 /16 x 6 2 /16 x 3 2 /16 x 8 2 /16 1/2 RouteTopology Global Min Cost Problems with the current practice Cost | Profit Behavior 1: ISPs interconnect selfishly to maximize profits! e.g. Backbone ISPs charge local ISPs. Two backbone ISPs Two local ISPs End-to-end service generates revenue Routing costs on links A simple example: MIN routing cost and MAX profit Increased routing and reduced profit Well-connected topology Topology Balkanization

33. x 2 2 /4 x 7 2 /4 x 4 2 /8 x 5 2 /8 x 1 2 /16 x 8 2 /16 1/2 RouteTopology Global Min Cost Cost | Profit Hot Potato 1 Problems with the current practice Behavior 2: ISPs route selfishly to maximize profits! e.g. upper backbone ISP wants to use hot-potato routing to reduce its routing cost. Profit reduction by routing inefficiency Behavior 1: ISPs interconnect selfishly to maximize profits! Topology Balkanization Increased routing and reduced profit Two backbone ISPs Two local ISPs End-to-end service generates revenue Routing costs on links A simple example:

34. Issues of Current ISP Settlement: A Micro Perspective Global Ideal case: cooperative ISPs RouteTopologyCost | Profit (maximized)Connectivity Global Min Cost Well- connected RouteTopologyCost | Profit (reduced)Connectivity Global Min Cost Balkanized RouteTopologyCost | Profit (further reduced)Connectivity Hot Potato Balkanized ISPs selfishly route traffic ISPs selfishly interconnect How to encourage ISPs to operate at efficient/optimal points?

35.  collects revenue from customers  distributes profits to ISPs Our solution: The Shapley Mechanism  Recall: three levels of ISP decisions Interconnecting decision E Routing decisions R Bilateral settlements  Source Destination Zero-Dollar Peering Customer-Provider Settlements Provider ISP Customer ISP Settlement  affects E, R Multilateral settlements  (E,R)(E,R) $$$ $$

36. Local decisions: E i,R i EiEi RiRi Given:   Objective: to maximize   i (E,R) Our solution: The Shapley Mechanism  Each ISP’s local interconnecting and routing decisions.

37. x 2 2 /4 x 7 2 /4 x 4 2 /8 x 5 2 /8 x 1 2 /16 x 8 2 /16 RouteTopology Results: Routing Incentive Cost | Profit Local Min Cost 1 Hot Potato  1/4 3/4 1/2 Global Min Cost Recall the inefficiency situation Shapley mechanism distributes profit ISPs route selfishly to maximize profits! E.g. the upper ISP wants to minimize local routing cost Best strategy for all ISPs: global min cost routing Profit increaseProfit maximized

38. Given any fixed interconnecting topology E, ISPs can locally decide routing strategies {R i * } to maximize their profits. Theorem (Incentive for routing): Any ISP i can maximize its profit  i by locally minimizing the global routing cost. –Implication: ISPs adapt to global min cost routes. Corollary (Nash Equilibrium): Any global min cost routing decision is a Nash equilibrium for the set of all ISPs. –Implication: global min cost routes are stable. Results: Incentives for using Optimal Routes Surprising! Local selfish behaviors coincide with global optimal solution!

39. x 2 2 /4 x 7 2 /4 x 4 2 /8 x 5 2 /8 x 1 2 /16 x 8 2 /16 RouteTopology Results: Interconnecting Incentive Cost | Profit  1/2 Global Min Cost x 6 2 /16 7/12 5/12 x 3 2 /16 ISPs interconnect selfishly to maximize profits! Recall: the best strategy for all ISPs is to use global min cost routes. E.g. the left local ISP connects to the low backbone ISP. Profit increase Further the right local ISP connects to the upper backbone ISP.

40. For any topology, a global optimal route R * is used by all ISPs. ISPs can locally decide interconnecting strategies {E i * } to maximize their profits. Theorem (Incentive for interconnecting): By interconnecting, ISPs will have non-decreasing profits. –Implication: ISPs have incentive to interconnect. –Does not mean: All pairs of ISPs should be connected. Redundant links might not reduce routing costs. Sunk cost is not considered. Results: Incentive for Interconnecting

41. Micro perspective ISP settlement: result summary Selfishly interconnect and route RouteTopologyCost | ProfitConnectivity Global Min Cost well- connected RouteTopologyCost | ProfitConnectivity Global Min Cost Balkanized RouteTopologyCost | ProfitConnectivity Hot Potato Balkanized ISPs have incentive to interconnect ISPs have incentive to use optimal routes    solves the selfish interconnecting problem  solves the selfish routing problem

42. Problems we try to solve win-win/fairProblem 1: Find a win-win/fair profit-sharing solution for ISPs. Challenges –What’s the solution? ISPs don’t know, even with best intentions. –Answer: The Shapley value –How to find it? Complex ISPs structure, computationally expensive. –Result: Closed-form sol. and Dynamic Programming –How to implement? Need to be implementable for ISPs. –Implication: New bilateral settlements efficient/optimalProblem 2: Encourage ISPs to operate at efficient/optimal points. Challenges –How to induce good behavior? Selfish behavior may hurt other ISPs. –Answer: The Shapley mechanism –Result: Incentives for optimal routes and interconnection –What is the impact of the entire network? Equilibrium is efficient? –Result: Optimal global Nash equilibrium

43. Guideline for –ISPs: negotiate stable and incentive settlements. –Governments: make regulatory policy for the industry. New Application and Properties of the Shapley Value Shapley Value: Contributions to Coalitions EfficiencySymmetryFairness Myerson 1977 EfficiencySymmetryDummyAdditivity Shapley 1953 EfficiencySymmetryStrong Monotonicity Young 1985 ISP Routing Incentive ISP Interconnecting Incentive Ma et al. 2007 ISP Profit Sharing Ma et al. 2008

44. My research: current work

45. My research: past work

46. My research: future interests Non-cooperative Game Theory Mechanism Design: e.g. Tax Policy, Auction Theory Coalition Game Theory Implementation: Algorithm, Protocol Design, Stability and Convergence Analysis Micro-economics and Game Theory Distributed Systems and Networks Algorithmic Game Theory Algorithmic Mechanism Design Modeling: Optimization, Stochastic Process Network Economics

47. Interdisciplinary research win-win/fairProblem 1: Find a win-win/fair profit-sharing solution for ISPs. Challenges –What’s the solution? ISPs don’t know, even with best intentions. –Answer: The Shapley value (Coalition Game) –How to find it? Complex ISPs structure, computationally expensive. –Result: Closed-form sol. and Dynamic Programming (Computer Sci.) –How to implement? Need to be implementable for ISPs. –Implication: New bilateral settlements (Operations) efficient/optimalProblem 2: Encourage ISPs to operate at efficient/optimal points. Challenges –How to induce good behavior? Selfish behavior may hurt other ISPs. –Answer: The Shapley mechanism (Mechanism design) –Result 1: Incentives for optimal routes and interconnection (Behaviors) –What is the impact of the entire network? Equilibrium is efficient? –Result: Optimal global Nash equilibrium (Non-cooperative Game)

48. My research: future research problems Economics-related problems: –Facilitate P2P traffic on the Internet –Network Neutrality Networking problems: –Incentive multicast protocols –Wireless mesh networks Problems in other fields –Graph theory –Distributed database system –E-commerce, social networks

50. Marginal contribution of ISP i to set of ISPs S:  i (S). The Shapley value mechanism  Profit v(S) is defined on any subset of ISPs Profit = Revenue - Cost Revenue Routing cost Profit: v(S) x 2 2 /4 x 7 2 /4 x 4 2 /8 x 5 2 /8 x 1 2 /16 x 8 2 /16 1 v( ) = 0v( ) =0.8125 x 6 2 /16 v( ) = 0.625 x 3 2 /16 1/2  ( ) = v( ) -v( ) = 0.1875 v( ) -v( ) = 0.625  ( ) =

51.   ( )=2.4/6=0.4 Empty S  Empty  (S( , )) v( )=0 v( )=0.2v( )- v( )=0.8 v( )=0 v( )- v( )=0.8 v( )=0.6v( )- N: total # of ISPs, e.g. N=3  : set of N! orderings S( ,i): set of ISPs in front of ISP i The Shapley value mechanism 

52. General Statement of Research Problem: –Individual objectives do not align with global objectives. –Economic incentives fail the engineering design goals of Internet. Objective: Reconcile individual incentives with global objectives. Approach: Design an economic mechanism –to address individual incentives, –to achieve global objectives. Global Objectives (Engineering Goals) Efficiency Fairness … High Profit Low Cost … Individual Objectives (Economic Incentives) Economic Mechanism

53. A Short Recap What have we done? – A Macro perspective –A central authority’s view to distribute profit among ISPs and make appropriate settlements. What are we going to do? – A Micro perspective –Analysis of ISP selfish behaviors and strategies and their effects on the entire network. Questions to answer –How will ISPs behave if the profits are distributed according to the Shapley value solution? –How does that affect the entire network?