- author = "Ken C. Sharman and Anna I. {Esparcia Alcazar} and Yun Li",
- title = "Evolving Signal Processing Algorithms by Genetic Programming",
- booktitle = "First International Conference on Genetic Algorithms in Engineering Systems: Innovations and Applications, GALESIA",
- year = "1995",
- editor = "A. M. S. Zalzala",
- volume = "414",
- pages = "473--480",
- address = "Sheffield, UK",
- publisher_address = "London, UK",
- month = "12-14 " # sep,
- publisher = "IEE",
- keywords = "genetic algorithms, genetic programming, simulated annealing, adaptive signal processing, neural networks, memory",
- ISBN = "0-85296-650-4",
-
URL = "
http://www.iti.upv.es/~anna/papers/galesi95.ps",
- abstract = "We introduce a novel Genetic Programming (GP) technique to evolve both the structure and parameters of adaptive Digital Signal Processing (DSP) algorithms. This is accomplished by defining a set of node functions and terminals to implement the basic operations commonly used in a large class of DSP algorithms. In addition, we show how Simulated Annealing may be employed to assist the GP in optimising the numerical parameters of expression trees. The concepts are illustrated by using GP to evolve high performance algorithms for detecting binary data sequences at the output of a noisy, non-linear communications channel.",
-
notes = "12--14 September 1995, Halifax Hall, University of
Sheffield, UK see also
http://www.iee.org.uk/LSboard/Conf/program/galprog.htm
Use simulated annealing to assist GP optimise numerical componets of expression trees.
DSP uses push and pop (any point in stack stkN) in function set
Previous outputs (Yn) used as inputs - {"}time recursion{"}. Single sample delay node Z. {"}infinitley many different tree structures for a particular system function{"}
{"}Time varying fitness function{"} Sigmoidal transfer function with beta evolable.
Pop 250, tournament 10, max gene 100, mutation, 2 ADFs, Evolved filter has zero bit error rate!
{"}Perliminary results...GP significantly out performed existing systems.{"}
GP computationaly expensive",
