An environment for visualizing higher dimensional measured data

Over the last few decades, the ability of instrumentation to measure large volumes of data has literally been growing exponentially. It is increasingly easy to build a measurement system with a large number of channels that come from acquisition cards and networked sensors. Analysis and statistics from raw measurements yeild derived measurements that add to the dimensionality of the space that a measurement user needs to understand to gain knowledge or solve problems from the data. It is therefore necessary to create methods that permit the exploration of such large and high dimensional sets of measured data within the limitations of the graphics rendering capabilities, screen sizes, and input devices that can reasonably be built into instrumentation and measurement systems. The problem addressed in this paper is that of vizualizing measured data of higher dimension (>; 10) and moderately large size. We present a software prototype that uses multiple complementary graphical views to display multidimensional data, with coordinated highlighting across views and flexible selection techniques. Here, we describe a tool providing a new combination of existing methods, scatterplot matrices (SPLOMs) and parallel coordinate plots (PCPs), for exploring higher dimensional data. We demonstrate the tool´s application to measured orthogonal frequency-division multiplexing (OFDM) signals including identifying the nature of a disturbance in the signal.