An Optimized Parallel Computing Paradigm for Mobile Grids Based on DSPOM
Aghila Rajagopal and M.A. Maluk Mohamed
Software System Group, MAM College of Engineering, Affiliated to Anna University Chennai
Abstract—Parallel computing methods decrease the processing time in mobile distributed systems compared to the conventional sequential computing techniques. But as they are developed from smaller mobile clusters to extensive mobile grids, they are prone to issues like high latency/jitter, processing speed, communication overhead, and low data transfer rate. So, an efficient and optimized parallel computing paradigm known as Distributed Shared Proxy Object Model (DSPOM) is developed based on Surrogate Object Model (SOM) integrated with Distributed Shared Object (DSO) for mobile grid. SOM is chosen to enhance the resource sharing of mobile grid computing, while DSO is chosen to reduce the computational complexity. The unused computing determinant is utilized by SOM to save the processing time. The transparency of the DSO model in terms of distribution and heterogeneity reduces the computational complexity. DSO also enhances the load adaptability and fault-tolerance to parallel programs on the mobile grid. The DSPO model performs better in terms of query time, query latency, packet loss, load adaptability, and fault-tolerance.
Index Terms—control object, distributed shared proxy object (DSPO), mobile host (MH), peer-to-peer (P2P), proxy object (PO), and surrogate object model (SOM)
Cite: Aghila Rajagopal and M.A. Maluk Mohamed, "An Optimized Parallel Computing Paradigm for Mobile Grids Based on DSPOM," Journal of Emerging Technologies in Web Intelligence, Vol. 6, No. 1, pp. 119-132, February 2014. doi:10.4304/jetwi.6.1.119-132
Index Terms—control object, distributed shared proxy object (DSPO), mobile host (MH), peer-to-peer (P2P), proxy object (PO), and surrogate object model (SOM)
Cite: Aghila Rajagopal and M.A. Maluk Mohamed, "An Optimized Parallel Computing Paradigm for Mobile Grids Based on DSPOM," Journal of Emerging Technologies in Web Intelligence, Vol. 6, No. 1, pp. 119-132, February 2014. doi:10.4304/jetwi.6.1.119-132
Array