Abstract

The maturity of Grid Computing and the exponential growth in the popularity of mobile and wireless technologies opened the door to new horizons of computing models based on mobile grid infrastructure. This growing demand has led to the need to study and explore new techniques and approaches to adapt the grid computation concepts into the challenging, ad-hoc mobile environment. Mobile grid scheduling has been the center of attention in this research focus. The different scheduling efforts developed so far overcame the mobility challenges by enforcing a direct point-to-point (P2P) communication between the task initiator and the service provider. In this paper, we introduce a dynamic, adaptive scheduler that relies on the prediction of patterns that do not require the P2P connectivity. The proposed approach can achieve similar and in certain situations better results than traditional P2P scheduling approaches. We are proposing a real-time scheduling approach for mobile ad-hoc grid environments that can provide flexibility in dealing with different types of real-time application tasks. The proposed scheduler utilizes two optimization heuristics to address this problem. The first one applies a modified version of the maximum flow problem while the second approach utilizes a multi-dimensional minimum cost function. Both approaches combine real-time scheduling characteristics while accommodating the mobility and battery dependency challenges. This dynamic scheduler provides different parameters that can be modified to control the scheduler QoS, reliability and ability to adapt to mobility and power dependency.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call