Load balancing in a distributed network environment : an ant colony inspired approach

Thumbnail Image
Other Title
Veerisetty, Neeharika
Author ORCID Profiles (clickable)
Master of Computing
Unitec Institute of Technology
Jayawardena, Chandimal
Pang, Paul
Masters Thesis
Ngā Upoko Tukutuku (Māori subject headings)
distributed system
load balancing
resource utilization
job response time
ant colony optimization
multi agent
dynamic load table
Artificial Neural Network (ANN)
decision making
ANZSRC Field of Research Code (2020)
Veerisetty, N. (2013). Load balancing in a distributed network environment: An ant colony inspired approach (Unpublished document submitted in partial fulfilment of the requirements for the degree of Master of Computing). Unitec Institute of Technology, Auckland, New Zealand. Retrieved from https://hdl.handle.net/10652/2364
With the incidence of technology at each and every juncture of human life, there has been an accelerated growth in computational needs to satisfy the technological cravings. Computer networks have evolutionarily emerged and have evolved as life blood of today’s global communication challenges. To fulfil the dynamic needs of present day networks, distributed and parallel computing applications are gaining momentum rapidly. Distributed networks have apparently become a better choice favouring the processing of large scale intensive applications which was previously unimaginable. However, it is evident that the load on a network is always relative to the volume of the application being processed. Eventually if the load on the network is not fairly distributed among all the available processing elements, it might result in improper resource usage and degraded network performance. Efficient load balancing approaches are essential to achieve proportional distribution of load among the network nodes to preserve the overall system integrity. Therefore, the process of identifying an efficient method to achieve proportional distribution of load is of paramount importance. To achieve an affective balance in load, this thesis investigates into an already existing Ant Colony based prototype called Messor and establishes a new approach based on dynamic load table concept augmented with ant search using Artificial Neural Networks. The proposed approach is simulated on a software based model network and the results are presented. The performance of the approach is evaluated based on certain performance criteria.
Link to ePress publication
Copyright holder
Copyright notice
All rights reserved
Copyright license
Available online at