Math @ Duke

Publications [#236043] of Robert Calderbank
Papers Published
 Calderbank, AR; Fishburn, PC; Rabinovich, A, Covering properties of convolutional codes and associated lattices,
IEEE Transactions on Information Theory, vol. 41 no. 3
(1995),
pp. 732746 [doi]
(last updated on 2018/10/23)
Abstract: This study proposes several Markov methods for the analysis of the expected and worst case performance of sequencebased methods of quantization. The quantization algorithm is assumed as a dynamic programming where the current step is dependent on metric functions. The main objective is to obtain a concise representation of these metric functions including the possible trajectories of the dynamic programming algorithm. To demonstrate this, the quantization of equiprobable binary data using a convolutional code is considered. In addition, these methods are also applicable to the quantization of arbitrary symmetric probability distributions using convolutional codes. For certain convolutional codes a formula that depends only on the distribution of differences for a single pair of path metrics is derived for expected use.


dept@math.duke.edu
ph: 919.660.2800
fax: 919.660.2821
 
Mathematics Department
Duke University, Box 90320
Durham, NC 277080320

