Load balancing and adaptive scheduling for data intensive prioritised traffic in multi–radio multi–channel wireless mesh networks. (15th January 2013)

• Record Type: Journal Article
• Authors: Kumar, Neeraj; Chilamkurti, Naveen; Park, Jong Hyouk; Park, Doo Soon
• Is Part Of: International journal of ad hoc and ubiquitous computing. Volume 12:Number 1(2013)
• Page Start: 3
• Page End: 13
• Publication Date: 2013-01-15
• Subjects: load balancing -- fairness -- network traffic flow -- priority traffic -- adaptive scheduling -- multi–radio networks -- multi–channel networks -- wireless mesh networks -- WMNs -- wireless networks -- quality of service -- QoS; Ubiquitous computing -- Periodicals; Embedded computer systems -- Periodicals; Electronic data processing -- Distributed processing -- Periodicals; Wireless communication systems -- Periodicals; Computer architecture -- Periodicals; 004.2
• Languages: English
• ISSNs: 1743-8225

Abstract

In recent times, wireless technology has emerged as a key technology for many data intensive applications such as Video on Demand (VoD), Voice over IP (VoIP), high speed internet access, video conferencing, etc. All these applications require high data rate so that data can be transmitted to the end users in a prioritised manner. The underlying system that suits these applications better is the Wireless Mesh Network (WMN), because it is a special type of network which is cost effective, resilient and adaptive. To preserve the Quality of Service (QoS) for all these applications, in this paper, we propose a Load Balancing and Adaptive Scheduling (LBAS) algorithm for data intensive prioritised traffic in WMNs. To schedule the incoming traffic, the Available Capacity (AC) and Load Index (LI) of each link is calculated. A new Cost Evaluation Function (CEF) is defined based upon the values of AC and LI. Each traffic flow is also associated with Rate Index (RI) and Delay Estimation (DE). Each link has lower and upper bounds on AC, and as soon as the load on a particular link crosses these bounds, the load is transferred to another link dynamically at run time. Higher priority traffic will be assigned to the link having high value of AC to minimise DE, and simultaneously, a matrix for LI is also constructed. This matrix is used to keep track of the system load at run time, so that scheduling can be done for the incoming traffic in a prioritised manner based upon the values of AC and LI. The performance of the proposed algorithm is evaluated using simulations on ns–2 with respect to various metrics. The results obtained show that the proposed algorithm is quite effective for balancing the load of multiclass traffic flows.