A scalable storage system for structured data based on higher order index array
Mehnuma Tabassum Omar, K. M. Azharul Hasan · 2016
Conventional computing techniques extensively diverge over large scale computing. In order to store and operate structured data, most data scientists suggest higher dimensional arrays, especially in linearization of higher order data. However, with the developing size of datasets, the structures become prone to performance degradation for inability of maintaining expanded data velocity. Besides, reallocation of data is required. The index array is a dynamic storage scheme which allows the array to extend subjectively according to the bound of the dimensions. All the dimensions of index array scheme require placing index to maintain data velocity. In this paper, we propose a scalable storage scheme for index array that illustrates the dynamic expansion nature of array. This scheme entails only two indices for any number of dimensions. Hence, simple algorithms can be designed for array operations. The proposed scheme overtakes the conventional structure in terms of memory utilization, index cost and element access.