17. Spawning Information Agents on the Web

Presentation on theme: "17. Spawning Information Agents on the Web"— Presentation transcript:

1 17. Spawning Information Agents on the Web
전기공학부 박정혁

2 문제 제기 1. 처리능력 밖의 작업이 Agent에게 주어졌을 때, 또는 능력 안의 일이지만 Overload 되었을 때.
Task delegation to other agents delegation to similar agents But, what if all the others are overloaded as well ??? 2. 필요로 하는 정보가 Remote host 에 존재하고 양이 많을 때, 어떻게 통신부담을 줄일 것인가?

3 Predictive Agent Cloning
대안 1 Predictive Agent Cloning simple to create 단점 원래 Agent와 같은 작업만 수행 복잡도로 인한 수행능력 저하

4 대안 2 1. Single overload => pass to others 2. All agents overload => create an agent, migrate to other host 3. Capability missing => create new agent 4. Large remote information => migrate or Spawn !!!

5 Providers Requesters Middle-agents 기반 : RETSINA middle-agents

6 Why is spawning necessary ?
Cloning 기능과 더불어 data전송 최소화, 불필요한 공간의 절약 => 효율 향상 Local spawning agent overload = machine overload idle processing capacity 활용

7 Spawning 조건 Self-awareness on internal level on MAS level
자료의 사용여부와 사용 part 내부 상태 (e.g. processor 활용도, 시간…) on MAS level middle agents 통해 원거리의 자료 파악

8 Spawning Procedure Reasoning before spawning Task splitting Spawning Reasoning after spawning

9 CPU, memory and communication loads

10 Simulation : The Cloning Case

11 When to Spawn ? Remote data : Ir Information produced by task : Ip
Size of spawn agent : S raw data vs. local processing data : α(t)

12

13 Conclusion Locally or remotely create spawn agents A copy or a specifically tailored agent overload or large remote information For large numbers of tasks, spawning significantly enhances the performance of MAS