Download - IIT Guwahati

I e  t  1  I e  t   Z e  t 
Mobile Agent based Systems (Artificial Being)
ROBOTICS LABORATORY
www.iitg.ernet.in/cse/robotics/
Department of Computer Science and Engineering
INDIAN INSTITUTE OF TECHNOLOGY GUWAHATI
MULTI-MOBILE AGENT MULTI-ROBOT SYSTEM
STIGMERGICALLY CONTROLLING A POPULATION OF
HETEROGENEOUS MOBILE AGENTS USING CLONING RESOURCE4
Mobile Agent Framework on a Mobile Robot 1
Mobile agents carry services (code, rules,
etc.) as their payload and populate a robotic
network. They either migrate
conscientiously or track pheromones
diffused by robots wanting a service.
STATIC AND MOBILE AGENTS
MOBILE ROBOT
A MOBILE AGENT FRAMEWORK FOR
AN INTERNET OF THINGS5
TASK
DISCOVERY
Architectures of mobile agent based devices that can
populate the Internet of Things
Architecture of the Cloning Controller
ROBOT
CONTROLLER
TASK
ALLOCATOR
TASK
EXECUTOR
Mobile Agent Cloning for Servicing Networked Robots 2
Request
Generation
Components of the off-board
Mobile Agent Framework
(MAF)
RRS: Robot
Requesting
Service
Functions:
• Support Migration and
Execution of Mobile
Agents
• Maintain neighbour list
The envisaged Internet of Things
Pheromoning
The air-conditioner, vacuum-cleaner and the robot use the
MAFs running on dedicated nodes while the printer and webcam have the same embedded within themselves.
Pheromones
diffused up to
two hop
neighbors
The Cloning Control Mechanism
Operations occurring at every step at each node:
AtNodeQueue
1. If an Agent is at the top of the Intra-node Queue
a. CloneifNecessary()
b. Compute “NextNode” using PheromoneConscientious Algorithm
c. Check if movement to the next node is possible. If
true goto (d) else goto (3)
d. Execute OnDeparture() Method
2. If an Agent is permitted into the Intra-Node Queue
a. Execute OnArrival() Method
3. Decrement the Life-time of every agent in the Queue
OnArrival()
1. Execute the service if this is the RRS that requested
for its service. If true goto (2) else goto (3)
2. Change the Cloning Resource and the Life-time
based on rewards
3. Enter into the Intra-Node Queue of the entering Node
OnDeparture()
1. Remove from the Intra-Node Queue of the existing
Node
Cloning
Id
Requested Service
with RRS Id
Parent Mobile Agent
Clone Mobile Agent
Artificial Immune System Based Learning Mechanism
3
Mobile Agent as an
Antibody
MOBILE
AGENTS WITH
INTELLIGENCE
BATTERY
SENSOR
ROBOT IN
A TRAP
NODE 4
TYPE 1
AGENT
BATTERY
SENSOR 1
NODE 2
Agent Migration
LEFT
LIGHT
SENSOR
ACTION TO BE
PERFORMED
BY THE ROBOT
RIGHT
LIGHT
SENSOR
TYPE 2
ROBOT
NODE 3
TYPE 1
ROBOT
TYPE 1
ROBOT
LEFT
LIGHT
SENSOR
1
RIGHT
LIGHT
SENSOR
1
ACTION 1
EPITOPE
ANTIGEN
PARATOPE
Simulation Result
Emulation Result
VARIABLE REGION
(RULES)
TYPE 1
AGENT
CloneifNecessary()
1. Find Resource needed for cloning
2. Find the Number of Clones
3. Decrement the Resource based on the Number of
clones
4. Create Clones
5. Recharge the Cloning Resource
TYPE 2
AGENT
CONSTANT REGION
(DESIGNATES THE
TYPE OF THE
AGENT/ROBOT)
NODE 1
TYPE 2
ROBOT
Components of the on-board
Mobile Agent Framework
(MAF)
ANTIBODY
Types of Mobile Agents in the
Internet of Things
Proof-of-Concept Experiment
1. SfR (Search for Resource)
Agents:
• These mobile agents are
capable of finding a resource in
the network as per the
specifications of the task to be
performed.
2. PaS (Provide a Service) Agents:
• These mobile agents carry the
source code (service) for a
task as its payload.
1
W. W. Godfrey. and Shivashankar B. Nair, "A Pheromone based Mobile Agent Migration Strategy for Servicing Networked Robots", Proceedings of the 5th
International ICST Conference on Bio-Inspired Models of Network, Information, and Computing Systems (BIONETICS 2010), December 1-4th, Boston, USA.
2 W.
W. Godfrey. and Shivashankar B. Nair, "Mobile Agent Cloning for Servicing Networked Robots", Proceedings of the 13th International Conference on Principles
and Practice of Multi-Agent Systems (PRIMA 2010), November 12-15th,2010, Kolkata, India.
3 W.
W. Godfrey. and Shivashankar B. Nair, "An Immune System based Multi-Robot Mobile Agent Network", Lecture Notes in Computer Science, Springer Berlin /
Heidelberg ISSN 0302-9743,Volume 5132/2008, pp. 424-433, The 7th International Conference on Artificial Immune Systems (ICARIS 2008), August 10-13th, 2008,
Phuket, Thailand.
4W.
W. Godfrey, Shashi Shekhar Jha and Shivashankar B. Nair, On Stigmergically Controlling a Population of Heterogeneous Mobile Agents Using Cloning Resource, Transactions on Computational Collective Intelligence, Springer. (Accepted)
Godfrey, Shashi Shekhar Jha and Shivashankar B. Nair, “On A Mobile Agent Framework for an Internet of Things”, Proceedings of the
International Conference on Communication System and Technologies, CSNT 2013, Gwalior, India, Published in IEEE Xplore, pp. 345-350.
DOI: 10.1109/CSNT.2013.79.
5W.W.