DSP basics

This chapter introduces the fundamentals of Digital Signal Processing (DSP) necessary for the design and construction of any interactive system that directly engages with the environment using sensors. There is a diverse range of such systems, including capturing the speech input for human-computer conversations [Cassell et al. 2000], hand gestures to input text [Jones et al. 2010], eye-gaze to facilitate more efficient interaction [Pfeuffer et al. 2015], recording biosignals such as heart rate to understand the body [Schmidt 2015], adapting digital content based on environmental conditions or locations [Rodden et al. 1998], and in each of the Case Studies described later in this book. All must capture and then process sensor data in order to provide input, or adapt output in a digital system. In many cases, the interactive systems developer must convert, filter, transform, and/or threshold raw signal data (from e.g. an accelerometer, microphone, or ECG sensor) into a form suitable for use in their application. The choices made in each Digital Signal Processing step have implications on the quality of the resulting output used to direct the decisions made by their application. To get the reader started, we aim to provide a beginners-guide to DSP. However, this introduction is far from exhaustive. Indeed, there are numerous great textbooks dedicated to advanced understanding of this topic and the reader is encouraged to consult them for more indepth theoretical knowledge. Recommended reading includes the ‘Digital Signal Processing’ titles by Proakis and Manolakis [2007] and Rawat [2015]. DSP basics begins with an introduction to the different types of signals, and continues to cover the analog-to-digital conversion topics of sampling, quantization, and coding. From there it covers digital-to-analog conversion, Discrete Fourier Transforms, autocorrelation, and Linear Time-invariant Systems. Finally, it ends with a walk-through of DSP use in Brain-Computer Interaction (BCI).