Anytime algorithms in embedded signal processing systems1

In this paper a family of so-called anytime signal processing algorithms is introduced to improve the overall performance of larger scale embedded digital signal processing (DSP) systems. In such systems there are cases where due to abrupt changes within the environment and/or the processing system temporal shortage of computational power and/or loss of some data may occnr. It is an obvious requirement that even in such situations the actual processing should be continued to insure appropriate performance. This means that signal processing of somewhat simpler complexity should provide outputs of acceptable quality to continue the operation of the complete embedded system. The accuracy of the processing will be temporarily lower but possibly still enough to produce data for qualitative evaluations and supporting decisions. Consequently anytime algorithms should provide short response time and be very flexible with respect to the available input information and computational power. The paper presents such algorithms based on the re-configurable version of the recursive signal transformer structure described in [1].