Mohammed Anowarul Hassan

In this project, we introduce a framework to enable dynamic and transparent mobile application offloading to clouds. POMAC can take intelligent offloading decision based on the dynamic nature of the environment. A prototype has been implemented on the Dalvik virtual machine and our preliminary evaluations show that POMAC can outperform existing schemes significantly and work with real-world applications seamlessly.

Designed and developed an architecture of a web based browser to access data across multiple devices owned by a single of a group of users. Click this link for the demo.

Designed and developed an architecture to outsource computation-intensive mobile applications to the cloud in order to minimize response time (approximately by 10x) by leveraging parallel processing for performance enhancement. The mobile applications were partitioned and the computation intensive part was outsourced to private/public cloud. While outsourcing, network latency and data affinity were considered and best resources were chosen accordingly. The scheduler of MMR also considers the parallelization overhead to choose the optimal degree of parallelism. Hadoop MapReduce was modified for parallel processing in Android environment. The framework MMR was evaluated with various real world applications including face detection and searching in Android environment and Amazon EC2.

Related Articles
Amazon Silk
Clone Cloud
Map Reduce

Designed and developed a Road Network for the Hand-held Devices to augment the computation with the roadside cloudlet. An incentive model was developed for resource (commodity machines) sharing among the users. This P2P based computation sharing model enables the user to find nearby resources with high network bandwidth and low latency for better performance while outsourcing the computation. This architecture was evaluated in Android environment.

Related Articles
Alcatel
Benchmark Applications