Balancing Performance, Energy, and Quality in Pervasive Computing
ICDCS '02 Proceedings of the 22 nd International Conference on Distributed Computing Systems (ICDCS'02)
Tactics-based remote execution for mobile computing
Proceedings of the 1st international conference on Mobile systems, applications and services
Slingshot: deploying stateful services in wireless hotspots
Proceedings of the 3rd international conference on Mobile systems, applications, and services
Simplifying cyber foraging for mobile devices
Proceedings of the 5th international conference on Mobile systems, applications and services
Mobile computing: the next decade
ACM SIGMOBILE Mobile Computing and Communications Review
Just-in-time provisioning for cyber foraging
Proceeding of the 11th annual international conference on Mobile systems, applications, and services
Hi-index | 0.00 |
Cyber foraging helps small devices perform heavy tasks by opportunistically discovering and utilising available resources (such as computation, storage, bandwidth, etc.) held by larger, nearby peers. This offloading is done in an ad-hoc manner, as larger machines will not always be within reach. In order to facilitate cyber foraging, an application has to be split up into locally executable code (e.g., GUI-code) and remotely executable tasks. Most cyber foraging systems strive to keep these tasks as large as possible, so that the benefits of executing code on a faster machine outweighs the cost of remote execution. Working with large, atomic tasks has some limitations, e. g., with regards to mobility and efficient use of remote resources. In order to dynamically allocate available resources in a manner most suitable for a given job, a task can be broken down into sub-tasks and services, and arranged in an directed graph. Using this graph combined with information about available, nearby peers, a cyber foraging scheduler can distribute tasks across multiple peers, allowing the user to do more with less. This paper presents the Locusts cyber foraging framework, with special focus on the task description language and the associated scheduler.