Text Box: Figure 2: Block diagram of physical environment prediction and agent operations * Text Box:  

Text Box:  

WLAN Resource Management Using Multi-Agent Systems

　

This research investigates cooperative resource management in WLAN (wireless local area networks) /WPAN (wireless personal area networks) interference environments. The objective of this research is to manage shared system resources fairly among multiple WLANs to optimize the overall performance. Expected results from the project will have a significant impact on next generation WLAN network management based on employing algorithms of agent interaction and coordination to facilitate resource management, predictive models for parameter estimation, and dynamic load balancing algorithms.

Approach

A multi-agent system-based approach is used to achieve information sharing and  decision distribution among multiple WLANs in a distributed manner.

　

　

　

　

　

　

　

　

　

Text Box: Figure 1: WLAN as a Multiagent system*

　





　

Architecture

This research maps the WLAN resource management problem as a distributed constraint optimization

problem (DCOP) as shown in Figure 2. It studies the effectiveness of DCOP algorithms to find the optimal resource assignment

through communications between distributed agents.

Mapping the WLAN to a DCOP

A={A1,...,An}:set of agents,each Acess Point is assigned an agent

X={X1,...,Xm}:set of variables, each Access Point has a variable Xi for Mobile Station i

D={d1,...,dm}:set of domains of variables

R={r1,...,rp}:set of relations,ri is a utility function

Text Box: A simple WLAN scenario(M:Mobile Station, A:Access Point)   PI:   Anita Raja Collaborators:Linda Xie (ECE), Ivan Howitt (ECE) RAs: Shanjun Cheng(Ph.D), Brian Rozier  

We extend Petcu's DPOP algorithm, a distributed constraint optimization algorithm to solve the

mapped WLAN resource allocation problem. DPOP is a Utility Propagation algorithm based

on dynamic programming , it requires a linear number of messages.



　

Co-PIs: Anita Raja (MAS focus), Professor Linda Xie (ECE), Professor Ivan Howitt (ECE)
RAs: Shanjun Cheng, James Rozier
　

Links:

This project is supported by NSF. The program webpage can be found here.

The presentation of this project can be found here.

Publications:

Jiang Xie, Ivan Howitt, and Anita Raja, "Cognitive Radio Resource Management Using Multi-Agent Systems," Proceedings of First IEEE Workshop on Cognitive Radio Networks (CRN 2007), in conjunction with IEEE Consumer Communications and Networking Conference (CCNC 2007), 2007.

*Jiang Xie, Ivan Howitt, and Anita Raja, "Framework for Decentralized Wireless LAN Resource Management" To appear in Emerging Wireless LANs, Wireless PANs, and Wireless MANs, Edited by Dr. Xiao and Dr. Pan, to be published by Wiley , 2007.c