Abstract
Mobile transaction technology ensures reliability, data consistency, and concurrency in mobile computing applications. Current solutions develop numerous transaction models and protocols in order to enhance the reliability and improve the performance of mobile transactions. However, they give no attention to the underlying network on which mobile transactions are carried out. This paper presents a network-centric strategy in order to improve the performance of mobile transactions. The proposed strategy is based on employing preemptive resume scheduling mechanism in mobile transactions. A cost effective analytical model for the network node with bursuty arrival traffic and prioritised scheduling mechanism has been developed using maximum entropy (ME) methodolgy. It involves rigorous analysis of a network node with finite capacity to accommodate messages, bursty arrival process to represent incoming multi-class messages, and the employment of a priority scheduling mechanism to give preferential treatment to high priority messages. This analytical solution provides closed form expressions for calculating the queuing delay per class involved in the commit processing of mobile transactions. The aim is to reduce queuing delays involved in the processing of mobile transactions at each network node. Such reduction of queuing delay significantly improves the performance of mobile transactions given the heavy network traffic and the lower network bandwidth. Several experiments have been conducted using the analytical modelling approach, which determine the impact of priority scheduling mechanism on the performance of mobile transactions. Experiments show significant performance improvements of the proposed strategy over existing approaches.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have