An Introduction to Multiagent Systems michael wooldridge ch9 Working Together

1. Ch9
Working
Together
Prepared by
Saif Salah Abood
Supervisor by
Assist. Prof.
Dr.Maha A. Bayati

2. Ch9
Outlines
we have looked how multiagent
• Encounters
• Reach agreements.
• Communicate.
• New, we have see how agents
can work together.

3. Cooperation
The term 'cooperation‘ is frequently
used in the concurrent systems
literature, to describe systems that
must interact with one another in
order to carry out their assigned
tasks.

4. Difference between multiagent systems
and distributed systems
• Who designed and implemented agents?
(One or many).
• What are their goals? (Shared or private).
• Do they need a strategy to achieve these
goals?
• When to make their decisions (Runtime,
Design time)?
• Do they have the ability to coordinate
dynamically?

5. Cooperative Distributed Problem Solving CDPS
• Each agent or entities in the systems has
distinct (but interrelated) expertise that they
could help to solve the problem of system.
• Cooperation is necessary because each single
agent has not sufficient expertise, resources,
and information to solve a problem.

6. Benevolence assumption
on cooperative problem solving
• agents share a common goal.
• no conflict between them.
• that greatly simplifies the designer's
task.

7. Coherence and coordination
• Coherence. Refers to 'how well the
multiagent system behaves as a unit, it
measured by
–quality,
–efficiency of resource usage,
– how system performance in failure.

8. Coordination. The degree to which agents can
avoid activity such as synchronization and
compatibility.
• perfectly coordinated : agents will achieve a
common goal and not clobber each other's
sub-goals.
• poor coordination : conflict between agents,
(destructively interfering) which requires
time and effort to resolve.
Coherence and coordination

9. The main issue in CDPS.
• How can a problem be divided into smaller
tasks for distribution among agents?
• How can a problem solution be effectively
synthesized from sub-problem results?
• How can the problem-solving activities of the
agents be maximizes the coherence metric?
• What techniques can be used to coordinate
the activity of the agents?

10. CDPS process (three-stage activity)
1. Problem decomposition.
Each problem is decomposed into smaller
sub-problems, and so on, until that each
sub-problem can be solved by single agent.
– The decomposition (problem) itself may be
treated as a cooperative activity .

11. CDPS process (three-stage activity)
2. Sub-problem solution.
This stage typically involves sharing of
information between agents.
3. Solution synthesis.
The solutions of sub-problems are
integrated into an overall solution.

12. The three stages of CDPS

13. Task Sharing and Result Sharing

14. Task Sharing and Result Sharing
In Task Sharing, a task is decomposed into sub-
problems that are allocated to agents
• If all agents are homogeneous (benevolence
assumption) then task sharing is straightforward:
any task can be allocated to any agent.
• If the agents are autonomous (no benevolence
assumption), then task allocation will involve
agents reaching agreements with others.

15. Task Sharing and Result Sharing
Result sharing.
Proactively (one agent sends another agent some
information because it believes the other will be
interested in it).
Reactively (an agent sends another information in
response to a request that was previously sent

16. Task sharing in the Contract Net
The Contract Net (CNET) protocol is a high-level
protocol for achieving efficient cooperation through
task sharing in networks of communicating problem
solvers

17. Agents procedures
1. Task announcement processing. Agent looking at
the eligibility specification contained in the
announcement, If it is eligible, then details of the
task are stored, and the agent will subsequently
bid for the task.
2. Bid processing. Details of bids from would-be
contractors are stored by (would-be) managers
until some deadline is reached. The manager then
awards the task to a single bidder.

18. Agents procedures
3. Award processing. The fail bidder delete details of
the task. The successful bidder must attempt to
expedite the task (which may mean generating
new sub-tasks).
4. Request and inform processing. if that information
is immediately available the requester sent to the
requestor an inform message containing the
required information, Otherwise, the requester
informs the requestor that the information is
unknown.) An inform message causes its content
to be added to the recipient's database.

19. Result Sharing
In result sharing, problem solving proceeds by agents
cooperatively exchanging information as a solution is
developed.
Typically, these results will progress from being the
solution to small problems, which are progressively
refined into larger, more abstract solutions

20. improving performance in result
sharing
• Confidence: independently derived solutions can be
cross-checked, highlighting possible errors, and
increasing confidence in the overall solution.
• Completeness: agents can share their local views to
achieve a better overall global view.
• Precision: agents can share results to ensure that
the precision of the overall solution is increased.
• Timeliness: even if one agent could solve a problem
on its own, by sharing a solution, the result could
be derived more quickly.