Leslie S. Liu

Cloud Service Portal for Mobile Device Management

Mobile device has become an important tool and component in many businesses and used widely as a successful platform to invent and develop new applications to increase efficiency and reduce cost. Maintaining a fleet of devices for the mobile task force in a secure and scalable fashion is critical to ensure the long-term success of enterprise mobility platforms. Many mobile device management suites have been designed and implemented lately to provide some management functions to enterprise mobile fleet, but to the best of our knowledge, none of them have leveraged the power of emerging cloud computing infrastructure and there are several areas that we see great potentials for significant improvements. In this paper, we present our design for a portal to provide remote management access to virtualized device management servers hosted in a service cloud. Our design is targeted to hide the details of the device management behind a standard-based, uniform control interface that can be viewed from a cross-platform agent that can run on multiple mobile platforms. We will also describe briefly the prototype we are developing as an internal pilot.

BlueStar: Managed Services for Enterprise Mobility

Mobile handheld devices communicating over wireless networks now represent a capable, robust, cost-effective platform for enterprise mobile solutions. These solutions must be manageable, secure, scalable, and capable of integration with other enterprise components. The scale of enterprise mobile solution deployments demands high levels of automation to keep labor costs in check and ensure compliance with enterprise standards for security and availability. This paper describes the design and deployment of BlueStar, a prototype automated management system for the delivery of enterprise mobile services. We are currently leveraging BlueStarpsilas management capabilities within the context of a mobile claims processing solution for the property and casualty insurance industry. We show how computer-based management processes and policies contribute to high levels of automation, making it possible to deploy applications and services in a property and casualty enterprise.

Proxima: a mobile augmented-image search system

Image-based search has become an increasingly active area of research. Despite the fact that many resources have been spent on creating new ideas and improving existing algorithms and designs, existing image-based search engines do not provide scalability or accuracy that is even remotely close to todays text-based search engines. The challenges in image-based search are due to a number of factors, including the computation intensity of existing algorithms and the difficulty of detecting and recognizing objects. In this paper, we present a system called Proxima, which leverages contextual information provided by a mobile device, such as time, location, and user data, to search for people in a stored image database who are similar to the person in the input image. Unlike other systems that automatically associate metadata with images, Proxima utilizes the image itself as part of the database query. We also describe a prototype that implements the Proxima algorithm to provide an image-based search service for social networking.

Context-Related Access Control for Mobile Caching

Mobile caching is a popular technique that has been widely applied in mobile applications to reduce the bandwidth usage, battery consumption, and perceived lag. To protect the confidentiality of cached data, the data with sensitive information has to be encrypted as it is cached on mobile devices. Currently, several mobile platforms provide encryption utilities which allow mobile applications to encrypt their local caches. However, existing encryption utilities are too coarse-grained and not directly applicable to dynamically enforcing fine-grained context-related access control policies in context-aware mobile applications.

Service-Enabled Automatic Framework for Testing and Tuning Multi-tier System

For a complex multi-tier system, to make the system well tuned and tested is of great challenge, since there are many configurations that can be changed in different tiers of the system. In this paper we proposed a service enabled workflow driven automatic framework for stress testing and tuning multi-tier system. In our hierarchical design, at the bottom are web service end-points following WSDM standard, in different tiers of the multi-tier system and in performance tester. On top of those endpoints are composite workflow endpoints, which together constitute the testing and tuning work-flow. The workflow is automatically iterative. The paper proposes XML schema for both performance tuning algorithms and performance best practices to allow seamless integration with the workflow. The efficiency of the proposed framework is shown in simulated experiments at the end of this paper.