Fundamental challenges in mobile computing
PODC '96 Proceedings of the fifteenth annual ACM symposium on Principles of distributed computing
A Framework for Opportunistic Cluster Computing Using JavaSpaces
HPCN Europe 2001 Proceedings of the 9th International Conference on High-Performance Computing and Networking
Self-Organizing Agents for Grid Load Balancing
GRID '04 Proceedings of the 5th IEEE/ACM International Workshop on Grid Computing
The Organic Grid: Self-Organizing Computation on a Peer-to-Peer Network
ICAC '04 Proceedings of the First International Conference on Autonomic Computing
Overcoming free-riding behavior in peer-to-peer systems
ACM SIGecom Exchanges
A Robust Decentralized Job Scheduling Approach for Mobile Peers in Ad-hoc Grids
CCGRID '07 Proceedings of the Seventh IEEE International Symposium on Cluster Computing and the Grid
Energy efficient scheduling for parallel applications on mobile clusters
Cluster Computing
| Hi-index | 0.00 |
Mobile devices can be integrated into grids to access grid resources but also to provide resources, such as CPU cycles or memory, building mobile grids . To exploit the potential of mobile grids, we propose an opportunistic job scheduling approach to harness cycles among mobile devices. Mobile nodes decide autonomously and locally which job to take by matching the job's requirements against their capabilities and coordinate with one another by means of shared job queues. A prototype implementation has been presented in previous work. In this work, we introduce selfish nodes which are expected to occur among mobile devices with limited energy sources. To react to selfishness, we introduce three strategies of game theory, that are, Tit For Tat (TFT), generous TFT, and Go-By Majority (GBM), to our approach and investigate the emerging behavior of the system by means of simulation. First results show, that the TFT and GBM implementations converge fast to fully selfish systems, while generous TFT exhibits self-healing characteristics.