Evaluating multi-view representations of a Web3D streaming server
Ayat Mohammed, Nicholas F. Polys, Vuk Marojevic, Richard Goff, Carl Dietrich · 2017
Analysis of multivariate data in space and time (spatio-temporal) has gained distinctive importance in all research domains. The development of new solutions that enhance massive 3D scientific visualizations on the Web is still growing. In this paper, we combine computing power, human visual perception and interaction and real-time Web3D rendering to develop a 3D visualization for multivariate data through CORNET-3D. CORNET-3D is a software tool, which dynamically displays live wireless signal spectra, i.e. magnitudes vs. radio frequency (RF), of cognitive radio nodes in the CORNET testbed through a browser. Real time visualization of the electromagnetic spectrum activities for wireless communications provides a visually perceivable representation of the cognitive radio nodes' power and occupancy in the frequency domain. This work compares user performance between three representations (views) of multivariate scientific data to mitigate the trade-off between proximity and occlusion in CORNET-3D. The three views are Colored-Spectrum, Colored-Bars and Colored-Spectrum-with-Bars. For both types of questions (power or bandwidth), the results show that it is easy for people to carry out visual discrimination in multi-variate data visualization using the Colored Spectrum-with-Bars view. But to get accurate judgments, the Colored spectrum view should be used.