Search Results for "clustering"

A community repository to develop Commodity Grid (CoG) Software for the access of Grids, through high level interfaces. The main repository is maintained at http://www.cogkit.org

Wireless sensor network (WSN) technologies have been employed in recent years for monitoring purposes in various domains from engineering industry to our home environment due to their ability to intelligently monitor remote locations. In this paper, we have developed a purely deterministic model that utilizes clustering to organize the WSN. We propose a deterministic energy-efficient clustering protocol that is dynamic, distributive, self-organizing and more energy efficient than the existing protocols. It utilizes a simplified approach which minimizes computational overhead-cost to self-organize the sensor network. Our simulation result shows a better performance with respect to energy consumption, which is reflected in the network lifetime in both homogeneous and heterogeneous settings when compared with the existing protocols. ...

While wireless sensor networks are increasingly equipped to handle more complex functions, in-network processing may require these battery powered sensors to judiciously use their constrained energy to prolong the effective network life time especially in a heterogeneous settings. Clustered techniques have since been employed to optimize energy consumption in this energy constrained wireless sensor networks. We propose an Enhanced-SEP clustering algorithm in a three-tier node scenario to prolong the effective network life-time. Simulation results shows that the Enhanced-SEP protocol achieves better performance in this respect, compared to other existing clustering algorithms such as LEACH and SEP in both heterogeneous and homogenous environments.

A MATLAB spectral clustering package to handle large data sets (200,000 RCV1 data) on a 4GB memory general machine. We implement various ways of approximating the dense similarity matrix, including nearest neighbors and the Nystrom method.