On the Economic Viability of Network Architectures

1. On the Economic Viability of Network Architectures Roch Guerin, Kartik Hosanagar University of Pennsylvania & Andrew Odlyzko, Zhi-Li Zhang University of Minnesota

2. Overwhelming need for flexibility in technology and business plans: The goals of the advertising business model do not always correspond to providing quality search to users. ... we expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers. ... But we believe the issue of advertising causes enough mixed incentives that it is crucial to have a competitive search engine that is transparent and in the academic realm.

3. Overwhelming need for flexibility in technology and business plans: The goals of the advertising business model do not always correspond to providing quality search to users. ... we expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers. ... But we believe the issue of advertising causes enough mixed incentives that it is crucial to have a competitive search engine that is transparent and in the academic realm. - Sergey Brin and Larry Page, 1998

4. Questionable dogma of streaming video: Keynote speech by SIGCOMM 2004 lifetime contribution award winner Simon Lam, http://www.acm.org/sigs/sigcomm/talks/lam-sigcomm04.pdf Lam’s conclusions: 1. Overprovisioning not a solution 2. Flow-oriented service needed 3. More QoS research is needed 4. Widespread commercial deployment of QoS within 10 years

5. Likely future of data networks: fast file transfers incl. faster-than-real-time video file transfers

6. Bandwidth was not a big problem even half a dozen years ago: Most common causes of performance problems as well as outages in networks today: In roughly the order • Network Engineers (What’s this command do?) • Power failures (What’s this switch do?) • Cable cuts (Backhoes, enough said) • Hardware failures (What’s that smell?) • Congestion (More Bandwidth! Captain, I’m giving you all she’s got!) • Attacks (malicious, you know who you are) • Software bugs (Your call is very important to us....) Sean Donelan, NANOG list, July 2, 2001

7. Effect of almost flat rates : U.S. cell phone usage, minutes per day around June of each year.

8. Background and Motivation • We are embarking in far-reaching explorations of new technology alternatives for tomorrow’s Internet • But clean-slate does not mean “green field” • There is a behemoth of an incumbent to deal with, and it’s far from standing still • The eventual relevance and success of new alternatives is not just a function of technical superiority • Economic aspects are likely to be equally, if not more important • What can we do to explore the economic viability of emerging “Next Generation Internet” alternatives? 8

9. Project Scope • Three main thrust areas for assessing the economic viability of new network architectures • Investigate and quantify the potential benefits of key proposed architectural features • Virtualization, integration, diversity • Explore when and why the existence of a formidable incumbent can affect the emergence of new technologies • Develop models that account for how the openness and flexibility of an architecture can foster the adoption of new technology, and its ultimate success 9

10. 1. One versus Many Networks • New architectural features such as virtualization are opening a new range of network deployment options • One shared/integrated network or many parallel virtual networks? • The emergence of the Internet as the de facto network for all communications is a perfect example of the benefits of “integration” • But is integration always beneficial, when, and why? • One network for all, means sharing of common infrastructure costs • But it also means that the needs of one application may affect all the others 10

11. Two Initial Perspectives • One service but different levels of performance • One network with service differentiation, versus improved performance through selectively combining several base network • Many different services • One integrated network that carries all traffic, versus separate networks, one per service • Develop models to ascertain the impact of different cost factors • Capital and operational costs • Network size • User demand for different services or performance levels • Pricing models 11

12. An Illustrative Example • A two services (A and B) scenario • Random user demands (DA and DB) for given price • Cost components • Higher total (sum) capital and fixed costs for separate networks • Lower operational costs (dedicated network is simpler to operate) • Different outcomes based on price, demand distribution, and cost models • Shows need for further investigation 12

13. 2. Displacing the Internet Incumbent • Three main factors • Utility gap between technologies • How big and how long will it last • Inter-operability gap • How big and how expensive to overcome (gateways) • Effect of network externalities • Positive (network effect) and negative (impact of congestion) • Goals are to develop models that can help new architectures assess their odds of success and when and how (pricing) to target their deployment 13

14. Model Framework (1) Users’ utility function: base + network value – congestion cost - price, where value and cost depend on installed base Dynamic model tracks (diffusion process) users’ response to competition between the incumbent and a new architecture 14

15. Model Framework (2) • Given distributions for users’ utility and hazard rates* for the two networks, the diffusion process that characterizes the evolution of the respective user populations can be expressed • This leads to expressions for net profits over a given time horizon, from which optimal pricing policies can be derived • Targeted model extensions include • Incorporation of capacity adjustments by the providers • Impact of technology “switching” costs on user demand • Effect of timing of technology introduction • Smaller installed base but lower technology edge • Rate at which demand potential converts into realized demand 15

16. 3. Benefits of Architectural Flexibility • The flexibility and openness of a network architecture can be a major advantage • Greater ability to adapt to new technologies • Facilitates deployment of and experimentation with new services • But it may also translate into higher operational costs • More options mean more sources of possible problems and more complex troubleshooting • More openness can result into greater security exposures • How do we go about quantifying this trade-off? 16

17. Initial Directions • Theory of real options provides a tool for quantifying the “value” of service experiments enabled by a given network architecture • Assigning value to the number of potential service offerings as well as the number of different instances of such offerings • Expressions for the expected “profit” of different combinations can be obtained for a number of cases of interest • Extensions to this basic building block will focus on • Incorporating the effect of operational costs (less flexible can mean more manageable) • Accounting for the impact of economies of scope 17