No abstract

No abstract

Summary Insights from geohash coding algorithms introduce significant opportunities for various spatial applications. However, these algorithms require massive storage, complex bit manipulation, and extensive code modification when scaled to higher dimensions. In this article, we have developed a two‐bit geohash coding algorithm that divides the search space into four equal partitions where each partition is assigned a two‐bit label as 00, 01, 10, and 11, which helps to uniquely identify a chosen data point and the two neighbors on its either side, taken along a particular dimension. This salient feature of the algorithm simplifies the generation of geohash code for the neighboring grid cells. In addition, it achieves efficient memory utilization by storing the geohash values of the training points as integers. Demonstrated by experiments for climate data assimilation, model‐to‐observation space mapping with a geohash code length of 24 bits for Lat‐Lon extent of India has shown favorable results with an accuracy of 85%. Performance and scalability evaluation of the proposed algorithm, optimized for multicore and many‐core processors has shown significant speedups outperforming a tree‐based approach. This algorithm provides a foundation for new spatial statistical methods that can be used for pattern discovery and detection in spatial big data.

Attempts to harness the big climate data that come from high-resolution model output and advanced sensors to provide more accurate and rapidly-updated weather prediction, call for innovations in the existing data assimilation systems. Matrix inversion is a key operation in a majority of data assimilation techniques. Hence, this article presents out-of-core CUDA implementation of an iterative method of matrix inversion. The results show significant speed up for even square matrices of size 1024 X 1024 and more, without sacrificing the accuracy of the results. In a similar test environment, the comparison of this approach with a direct method such as the Gauss-Jordan approach, modified to process large matrices that cannot be processed directly within a single kernel call shows that the former is twice as efficient as the latter. This acceleration is attributed to the division-free design and the embarrassingly parallel nature of every sub-task of the algorithm. The parallel algorithm has been designed to be highly scalable when implemented with multiple GPUs for handling large matrices.

This paper deals with analyze of a single server queueing system with immediate feedbacks and working vacation. Upon arrival if the customer sees the server to be busy then it joins the tail end of queue. Otherwise if server is idle, the customer gets into service. After completion of service, the customer is allowed to make an immediate feedback in finite number. Busy server may fail for a short interval of time. Using supplementary variable technique the steady state results are deduced. Some system performance measures are discussed

No abstract

No abstract

No abstract