Signal-space, the key to signal processing

The `signal-space¿ approach is put forward as providing a unifying perceptual framework for, and new insights into, the diverse and pervasive field of signal processing. In particular, it provides a convenient means of identifying a number of limiting conditions which must be satisfied by any signal-processing transformation, and it leads to the important new conclusion that the limiting processing gain achievable in any operation is equal to the quantity of known or postulated information embodied in the processing algorithm. The method leads to theoretically optimum solutions in some areas where these may not have been recognized before, and in others it indicates the full scope and/or limitations of optima previously developed or discovered in a more specific context. It is also effective for assessing the penalties and `trade-offs¿ of practical sub-optimal solutions. The approach dopted is explanatory rather than rigorous and is illustrated by generalized practical examples.