Bakabs: Managing load of cloud-based web applications from mobiles

Abstract

The cloud services invocation from the handset enables the next generation of mobile applications that are not limited by storage space and processing power. Bakabs is one such application for Android and iOS devices, which makes use of Google Analytics cloud services to track the traffic of websites, replicated on multiple instances in different locations. Bakabs also suggests the number and type of instances that are required to handle the loads, based on the linear programming model. The prediction is performed at the mobile cloud middleware, which facilitates invocation of multiple cloud services from mobiles. Based on the suggestions, the user can decide to turn on/off instances, thus saving costs taking advantage of the pay-as-you-go model and the elasticity of the cloud. The performance analysis of the application shows that Bakabs can utilize cloud services with significant ease and reasonable performance latencies on the devices, with the considered technological choices. © 2011 ACM.