Noisy signal processing methods

Signal processing, in the general sense of the term, is concerned with the study, design and production of signal processing systems [1][2][3] . It is a methodological discipline, both conceptual ("Signal processing: a mathematician's passion" [4] ) and a branch of applied science ("Signal processing: an ancillary discipline?" [3] ), and indispensable to many fields of application, sources of "signal"-type problems. The boundaries of signal processing are often blurred, and interactions are numerous, whether with application fields or with neighbouring disciplines, whether upstream (formalizing the concepts used), downstream (dealing with the media used) or in competition (exploiting concepts or methods of the same nature in a different context); consider, for example, the impact of microelectronics and computer science on the development of digital signal processing (DSP).

Concepts are essentially of mathematical origin: modeling methods, probabilistic and statistical tools, numerical analysis, optimization... Realizations are based on computer science, electronics, optics, acoustics... Conceptual and methodological capital is also exploited in automatic control (identification and control), data processing, pattern recognition, operations research, artificial intelligence, robotics...

Applications include process control, non-destructive testing and measurement, as well as telecommunications, surveillance systems (radar, sonar, intrusion, etc.), guidance and navigation systems, geophysical exploration (remote sensing, mapping, oil prospecting), biological and medical engineering, the nuclear sector...