Agent system and method

• US 7,263,702 B2
• Filed: 10/09/2003
• Issued: 08/28/2007
• Est. Priority Date: 03/24/1999
• Status: Expired due to Fees

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Cho, Kenta, Kase, Naoki
• Primary Examiner(s): Thomson; William
• Assistant Examiner(s): Hoang; Phuong N.
• Application Number:
  US10/681,186
• Publication Number:
  US 20040078801A1
• Time in Patent Office: 1,419 Days
• Field of Search: 719/317, 719/331, 709/202, 711/1, 711/209
• US Class Current: 719/317
• CPC Class Codes: G06F 9/4862 the task being a mobile agent, i.e. specifically designed to migrate

Abstract

An agent system processes information by agents. The agent system is connected to an agent system of different type through a network. A wrapper class memory stores component data of wrappers corresponding to each type of agent system. The wrapper specially processes information in a corresponding agent system. A wrapper generation section generates a wrapper corresponding to an agent system of a different type based on the component data of the different type in the wrapper class memory. The wrapper acts for the agent to use resources in the other agent system.

Claims

1. An agent system for processing information by an agent having agent state data, the agent system as an original agent system being connected to a different type agent system through a network, comprising:

a wrapper class memory configured to store component data of wrappers corresponding to each type of other agent systems, the wrappers being for processing information in a corresponding agent system;

a same type node data memory configured to store node data representing nodes included in agent systems of the same type as the original agent system and other agent systems of different type discriminately;

a different type node data memory configured to store access data for each node included in the other agent systems;

a correspondence table configured to store correspondence data between each category of movement commands and each type of the other agent systems;

a movement request detector configured to detect a request for the agent to move to the different type agent system from the agent state data by referring to the correspondence data and the node data; and

a wrapper generator, in response to the request, configured to generate a wrapper corresponding to a type of the different type agent system based on the component data of the type stored in said wrapper class memory;

wherein, the wrapper generates a wrapped agent based on a component of the wrapper, and sends the wrapped agent to the different type agent system, the wrapped agent having the agent state data and the access data,

the wrapped agent executes the agent state data to use resources in the different type agent system, and returns the agent state data based on an execution result to the agent in the original agent system,

the agent continually executes the agent state data based on the execution result in the original agent system.

2. The agent system according to claim 1, wherein said movement request detector detects a movement command to move to other node from a script of the agent state data, decides whether the other node is included in the other agent systems of different type by referring to said same type node data memory, and extracts a type of other agent system corresponding to the movement command by referring to said correspondence table if the other node is included in the other agent systems of different type.

3. The agent system according to claim 2, further comprising:

a retrieval unit configured to retrieve the component data of a wrapper corresponding to the type of other agent system extracted by said movement request detector from said wrapper class memory,

wherein said wrapper generator extracts the access data for the other node as a movement destination node from said different type node data memory after generating the wrapper, supplies the access data to the wrapper.

4. The agent system according to claim 3, wherein the wrapper comprises:

a different type movement memory configured to store movement data of the agent between the different type agent system and another agent system; and

a movement source return memory configured to store return data of the agent from the different type agent system to the original agent system.

5. The agent system according to claim 4,

wherein the wrapper supplies the movement data and the return data to the wrapped agent in the different type agent system through an external communication unit.

6. The agent system according to claim 5, wherein the wrapped agent comprises:

an agent state memory configured to store the agent state data including the script;

a movement unit configured to store the movement data;

a return unit configured to store the return data; and

a controller configured to execute action of the wrapped agent according to the data stored in said agent state memory, said movement unit and said return unit.

7. The agent system according to claim 6, wherein, if the action of the wrapped agent fails in the different type agent system, said return unit indicates the agent in the original agent system to execute the same action.

8. The agent system according to claim 6, wherein, if a movement command to a movement source node is detected from the script of the agent state data, said return unit returns a function of the wrapped agent to the agent of the movement source node in the original agent system.

9. A method for processing information by an agent having agent state data, the agent system as an original agent system being connected to a different type agent system through a network, comprising:

storing component data of wrappers corresponding to each type of other agent systems, the wrappers being for processing information in a corresponding agent system;

discriminately storing node data representing nodes included in agent systems of the same type as the original agent system and other agent systems of different type;

storing access data for each node included in the other agent systems;

storing correspondence data between each category of movement commands and each type of the other agent systems;

detecting a request for the agent to move to the different type agent system from the agent state data by referring to the correspondence data and the node data;

generating a wrapper corresponding to a type of the different type agent system based on the stored component data of the type in response to the request;

generating a wrapped agent based on a component of the wrapper, the wrapped agent having the agent state data and the access data;

sending the wrapped agent to the different type agent system;

in the wrapped agent,

executing the agent state data to use resources in the different type agent system, and

returning the agent state data based on an execution result to the agent in the original agent system; and

in the agent,

continually executing the agent state data in the agent based on the execution result in the original agent system.

10. The method according to claim 9, further comprising:

detecting a movement command to move to other node from a script of the agent state data;

deciding whether the other node is included in the other agent systems of different type by referring to the stored node data; and

extracting a type of other agent system corresponding to the movement command by referring to the stored correspondence data if the other node is included in the other agent systems of different type.

11. The method according to claim 10, further comprising:

retrieving component data of a wrapper corresponding to the type of other agent system from the stored component data;

extracting access data for the other node as a movement destination node from the stored access data after generating the wrapper; and

supplying the access data to the wrapper.

12. The method according to claim 11, further comprising:

storing movement data of the agent between the different type agent system and another agent system in the wrapper; and

storing return data of the agent from the different type agent system to the original agent system in the wrapper.

13. The method according to claim 12,

in the wrapper, further comprising:

supplying the movement data and the return data to the wrapped agent in the different type agent system through an external communication unit.

14. The method according to claim 13, further comprising:

storing the agent state data including the script in the wrapped agent;

storing the movement data in the wrapped agent;

storing the return data in the wrapped agent; and

executing action of the wrapped agent according to the stored data.

15. The method according to claim 14, further comprising:

indicating the agent in the original agent system to execute the same action, if the action of the wrapped agent fails in the different type agent system.

16. The method according to claim 14, further comprising:

returning a function of the wrapped agent to the agent of the movement source node in the original agent system, if a movement command to a movement source node is detected from the script of the agent state data.

17. A computer program product, comprising:

a computer readable program code embodied in said product for causing a computer to process information by an agent having agent state data, the agent system as an original agent system being connected to a different type agent system through a network, said computer readable program code comprising;

a first program code to store component data of wrappers corresponding to each type of other agent systems, the wrappers being for processing information in a corresponding agent system;

a second program code to discriminately store node data representing nodes included in agent systems of the same type as the original agent system and other agent systems of different type;

a third program code to store access data for each node included in the other agent systems;

a fourth program code to store correspondence data between each category of movement commands and each type of the other agent systems;

a fifth program code to detect a request for the agent to move to the different type agent system from the agent state data by referring to the correspondence data and the node data;

a sixth program code to generate a wrapper corresponding to a type of the different type agent system based on the stored component data of the type in response to the request;

a seventh program code to generate a wrapped agent based on a component of the wrapper, the wrapped agent having the agent state data and the access data;

an eighth program code to send the wrapped agent to the different type agent system;

a ninth program code for the wrapped agent to execute the agent state data to use resources in the different type agent system;

a tenth program code for the wrapped agent to return the agent state data based on an execution result to the agent in the original agent system; and

an eleventh program code for the agent to continually execute the agent state data based on the execution result in the original agent system.