1. Decentralized Resource Allocation in Application Layer Networks T. Eymann, M. Reinicke University Freiburg, Germany O. Ardaiz, P. Artigas, F. Freitag, L. Navarro Polytecnic University Catalunya, Spain

2. Outline Motivation Catallaxy Paradigm for Decentralized Resource Allocation Experiments Results Open Issues & Further Research

3. Application Layer Network Deployment S S S S S S S S S S S S S S S S S S S S S S S S S D D D D D Application Layer Network (Web Proxy Caching Hyrarchy): 6 servers each requires 1 Mbits net capacity, 200 Mbytes Storage, Less 2 hops from demand regions: A,B,C,D,E Programmable Infrastructure: 30 nodes distributed throught Internet each 10 Mbit net capacity, 2 GByte Storage Resource Allocation

4. Resource Allocation Problem Centralized RA is computationally intensive (and single point of failure). And it will get works:  Very Dynamic Infrastructures (Resource nodes come and go frequently): dial up nodes, mobile nodes,...  High Node Density Infrastructures (Many nodes with little resources): P2P systems, pervarsive computing,..

5. Solution: Economic Markets Resource Allocation works in Real World with an economic model: allocation of goods among human beings takes place in “markets”. Markets:  just distribution of utility by a central arbitrator (centralized economy)  decentralized action of utility-maximing agents using a central auctioneer  direct agreement between negotiating agents (Catalaxy)

6. The Catallaxy as a concept for market coordination Catallaxy is an alternative word for „market economy“ (Mises and Von Hayek of the Neo-austrian economic school) “Fundamentally, in a system in which the knowledge of the relevant facts is dispersed among many people, prices can act to co-ordinate the separate actions of different people in the same way as subjective values help the individual to co- ordinate the parts of his plan.” (Friedrich A. von Hayek, The Use of Knowledge in Society, 1945) “The Market” as a technically decentralized, distributed, dynamic coordination mechanism:  Adam Smith’s “invisible hand”, Hayek’s “spontaneous order”, Walras’ “non-tâtonnement process” Coordination and a stable environment are emergent features of the market  Pursuing local goals alone already stabilizes and coordinates the system.

7. How to Implement Catalaxy: Agents Environment, e.g. Market Agent Sensor, e.g. received offers Effector, e.g. sent offers (Intention: increase own utility) Reasoning, e.g. calculation of a counter-offer using heuristics (may become arbitrarily complex, e.g. AI)

8. Agent-mediated digital economy Characteristics for the agent-mediated digital economy:  Software agents act selfish, because their human owners do: Competition is the norm.  Software agents keep their utility function private: If made public, the agent can be exploited.  Software agents communicate directly: Centralized control institutions can always be bypassed. Consequences:  Cooperation is always pareto-eliciting (increases utility of all participants)  No free lunch: everyone has a utility function (business model), even centralized institutions  Information is not free or public (every participant operates on private knowledge and subjective values)

9. Negotiation Protocol - Example Buyer Seller cfp (service access) propose (service access, p S =$24) propose (service access, p B =$18) propose (service access, p S =$21) accept-offer(service access, p B =$21) commit (service access, p S =$21) time Client SC

10. Heuristic-Adaptive Reasoning: Example for a Seller (1) propose (service access, p S =$24) Update Market Price Valuation propose (service access, p B =$18)

11. Heuristic-Adaptive Reasoning: Example for a Seller (2) Should I leave the negotiation? propose (service access, p S =$24) propose (service access, p B =$18)

12. Heuristic-Adaptive Reasoning: Example for a Seller (3) Should I leave the negotiation? Should I make a concession? reject Yes No propose (service access, p S =$24) propose (service access, p B =$18)

13. Heuristic-Adaptive Reasoning: Example for a Seller (4) Should I leave the negotiation? Should I make a concession? What amount should I concede? reject propose (service access, p S =$24) Yes No Yes propose (service access, p S =$24) propose (service access, p B =$18)

14. Heuristic-Adaptive Reasoning: Example for a Seller (5) Should I leave the negotiation? Should I make a concession? reject propose (service access, p S =$24) propose (service access, p S =$21) propose (service access, p S =$24) propose (service access, p B =$18) Yes No Yes „costs of life“ (tax) will be deducted in discrete time slots

15. Application Coordination Communication Cooperation Application Services Network Services Physical Services Heuristic-Adaptive Reasoning: Parameters Negotiation Strategy: Achieving utility maximization setting e.g. concession rate, concession amount, time pressure in relation to market (and the transaction partner). Concession Probability Continuation Probability Concession Amount Market Price Learning Weight Mark-up

16. Heuristic-Adaptive Reasoning: adaptation by evolutionary learning Send „plumage“ (  profit x, Genotype x )     profit 1 Genotype 1  profit 2 Genotype 2  profit 3 Genotype 3  profit 4 Genotype 4 Create agent (Genotype  Genotype 1 ) select Genotype (  profit x )

17. Experiments Simulated Scenarios Evaluated Dimensions

18. Simulated Application Scenario How to match a network of clients and services? Clients (ADSL 1 Mbit) Acrobat Service Copy of Document MyCompanyPortfolio.pdf (6 Mbytes) Web Server with limited Resource (4 – 60 Mbits) 1 2 3

19. Catallactic Message Flow Client request_Service (MyComPortfolio.pdf) BW Negotiation Service Negotation

20. Baseline Message Flow Master Service Copy as Centralized Auctioner for BW and SC Client request_Service (MyComPortfolio.pdf)

21. Evaluation Dimensions CDN P2P GRID A few, powerful A lot, modest Fixed networks Mobile, ad - hoc, overloaded networks Stable Changing node density node dynamics lowmediumhigh medium high CDN P2P GRID It is required an “abstract” simulator

22. Simulator Scenarios: Resource Density Variations Low Density: Few nodes (5) Lots Resources per Node (60 Mbits) Middle Density: More nodes (25) Less Resources per Node (12 Mbits) High Density: More nodes (75) Less Resources per Node (4 Mbits)

23. Simulator Scenarios: Dynamic Values Dynamic: Nodes up & down with 20 % probability every 200 ms. Quasi-static: Nodes always up. Very dynamic: Nodes up & down with 40 % probability every 200 ms.

24. Simulator - Demand Clients located in every edge node. Client request_Service (1 Mbit Server Net Bandwidth, 50 sec). Random values: # of demands (among clients) # of serviceIDs (among 50 diferent videos) time betwen demands (average 2000 ms) Moving clients: Movement time (How often demand moves) Movement radius (How far demand moves) Movement percent (How much demand moves)

25. Simulator Choice The Catnet simulator is build over JavaSim [Univ. Ohio]: JavaSim is a network simulator based in autonomous components. Javasim implemented in java=> Ease of development, and efficient []. Javasim models every aspect of a real network: latency, bandwith, lost packets, routing,=> We take into account resource locality (vs. MAS simulators) Application module implement interfaces of common Inet protocols: TCP, UDP, Mcast => our components can be modified to work in real world without modification.

26. Preliminary Results Evaluation Criteria. Preliminary Results:  Comparison by Scenarios,  Adaptability Evaluation.

27. Evaluation Criteria RAE (Resource Allocation Efficiency)  The ratio of matched transactions divided by the number of all proposals: # "accepts“/ #"proposals“ REST (Response Time (Service Access Time))  How long does it take on average to fill a request: time between “cfp” and “accept” CC (Communication Costs)  How much communication is needed until the result: # messages * # hops.

28. Results by criterion – RAE (%) CatallacticBaseline Topology Dependency @ middle density RAE better @ very dynamic Scenario RAE at quasi- static, slow scenarios

29. Results by criterion – REST(ms) CatallacticBaseline REST is higher for catalactic: but not as much as expected.GOOD

30. Results by criterion: CC (# messages * #hops) CatallacticBaseline CC is similar. But it was expected to be higher because of more negatiations messages: GOOD. CC increases with density, since higher density means more nodes to send to.

31. Results by Scenario Quasi-static High node density Very dynamic / low ND Very dynamic / high ND  Green: confirmed, Red: rejected Resource Allocation Efficiency Communication cost Reaction time bbb b b b b b c cb System b

32. Adaptation: Baseline Simulation In baseline system prices keep constant => no adaptation

33. Adaptation: Catallactic Simulation In catalactic system prices adapt over time

34. Open Issues & Further Research Oscillations, Caotic behaviour. Tragedy of commons. Malevolous agents. Colaboration with agent researchers Colaboration with Complex Adaptive System researchers. Colaboration with Grid / P2P projects Scalability, dynamics. Theoretical Modelling. Implementation in grids & P2P scenarios.

35. Thank you, Questions? More info: http://research.ac.upc.es/catnet/