Nonlinear readout signal processing using the DataWave video signal processor

Implementation of digital signal processing operations as part of the data acquisition process in high-energy physics detectors requires very high-speed devices. Signal processors are not typically used in these applications due to inadequate processing performance. New components developed for high-definition television (HDTV) are considered to be applicable for these purposes. By using programmable signal processors a superior reconfigurability can be obtained, as opposed to VLSI-based custom implementations. Where the algorithms are fixed when-the design has been completed. This paper presents one such processor, the DataWave video signal processor, and describes how some nonlinear algorithms for readout signal processing can be implemented with it. The operations selected include an order statistic filter structure and an FIR-order statistic hybrid (FIR-OS) filter with trainable coefficients and rank operator. Comments on the suitability of the DataWave processor for different kinds of algorithms are also given.<
>