A reduced-complexity algorithm and its systolic architecture are presented for computation of the discrete cosine transform. The proposed scheme not only leads to a fully-pipelined regular and modular hardware, but also offers significantly higher throughput, lower latency and lower area-time complexity over the existing structures. The proposed design is devoid of complicated input/output mapping and complex control structure. Moreover, it does not have any restriction on the transform-length and it is easily scalable for higher transform-length as well.
Bitte loggen Sie sich ein, um Zugang zu diesem Inhalt zu erhalten