Can Reconstruction Fir Filter Based On Genetic Algorithm Research And Design,

Reconfigurable computing system developed on the basis of all-purpose computer and special-purpose Application Specific Integrated Circuit (ASIC). It has combined the advantages of flexibility of the all-purpose computer and the efficiency of special-purpose ASIC. There is important application in a great deal of fields such as deciphering, signal processing and pattern-recognition. Algorithm-level function reconfigurable circuit is a new reconfigurable digital system. Under the stimulus by external condition and the parameter, this kind of circuit can metamorphose and evolve automatically, organically, to form intellectual hardware circuit with a certain meaning.Presently, in the research on algorithm-level function reconfigurable circuit, evolvable circuit's design is the most typical hardware design method and be widely studied. It has combined with the advantages of evolutionary algorithm and programmable logic device, but evolvable hardware is mostly realized by full hardware form. The full hardware implemental method has not only increased difficulty of realizing the evolutionary algorithm but also greatly reduced the application flexibility of the algorithm-level function reconfigurable circuit.Overcoming the deficiencies of evolutionary hardware described above, this paper proposed an improved method by deeply researching the typical reconfigurable system configuration. The idea that genetic algorithm can effectively solve the system optimization problem has been applied to programmable system-on-chip. The shortcomings of full hardware implemental method had been overcomed by using embedded processors as a reconfigurable control unit, adopting corresponding evolutionary models and chromosome encoding mode according to different applications, and then the reconfigurable Finite Impulse Response(FIR) digital filter system has been designed based on the genetic algorithm.In this paper, the current research situation, the meaning of reconfigurable computing technology and the existing problem has been summaried at first. Then we study the genetic algorithm and its improved methods, analyse the characteristic of the existing typical algorithm-level reconfigurable circuit structure, and put forward the self-reconfigurable system-on-chip based on genetic algorithm. By embedded processor in reconfigurable logic device, it can achieve evolutionary computation based on the genetic algorithm, implement auto-configuration process with the IP core which hold on reconfigurable unit, and then an algorithm-level functional reconfigurable system-on-chip is constructed.Through the design and implementation of the FIR self-reconfigurable system, the validity of the design method mentioned in this paper is verified. A new high-speed reconfigurable FIR processor structure with coarse grains has been proposed after analyzing the FIR algorithm and its implementary hardware structure. The experimental hardware platform which implements the reconfigurable FIR with algorithm-level reconfiguration has been constructed with proper devices. The reconfigurable optimal FIR had been implemented in the platform using genetic algorithm and the test performance indicate that the alrithm and design method are reasonable and successful.