Novel VLSI implementation of Peano-Hilbert curve address generator

This paper presents a fast algorithm for generating Hilbert address for hardware implementation with low storage requirement. This work avoids the use of recursive functions as compared with Quinqueton's work, and eliminates complicated bit manipulations as proposed by Butz, and does not use any look-up-tables as implemented by Kamata. Each address can be obtained in one clock cycle by one-to-one mapping using a simple incremental counter and cascading of multiplexers. The merit of our method is that it achieves very high speed when computing the Hilbert address which requires little memory storage.