Immune Evolutionary Computation And Its Application

In recent years, more and more scientists and experts in many research fields are paying attention to the secret of life phenomena and biologic intelligent behaviors. Especially the research on biologic information system, has been gradually regarded as an important area in artificial intelligent research field. The above research, with characters of multilevels in development of modem sciences, as well as interinfiltration and interacceleration among multisubjects and multifields. takes an important role in the development of information science and computer science, and at the same time proposes many techniques and methods for engineers in many fields. Based on this consideration, this paper first gives a deep analysis on characteristics of existing evolutionary algorithms, then uses the concept of immunity in biology for reference, and finally proposes a series of novel algorithms, namely evolutionary algorithms based on immune mechanism. There are three kinds of algorithms, i.e.. immune algorithm, immune programming and immune strategy, whose core concentrates in constructing an immune operator, and this operator is composed of vaccination and immune selection. The goal of introducing the concept of immunity into the canonical evolutionary algorithms is to utilize some characteristics and knowledge in the pending problems for restraining the degenerative phenomena during evolution, so as to improve the algorithmic efficiency. The theoretical analysis and many simulations show that immune evolutionary algorithms are not on;\ feasible but also effective and are conducive to alleviating the degeneration phenomenon in the original algorithms, thus greatly increasing the converging speed. It can be proved in theory that the above three algorithm are all convergent.Immune evolutionary algorithms are optimal algorithms in essence, therefore, they can be used in some fields, such as cybernation, pattern recognition, optimal design, meshing learning, network security, etc. There are also some examples of these attempts described in this paper.First, in the aspect of designing an artificial neural network (ANN), the existing neural network models are all based on understanding of natural neural system, and established by highly simplifying and abstracting this system, which is propitious to its development and application in engineering practice, but losses some original functions of the natural systemat the same time. With the development and wide spread of ANN's applying, there continually appear some problems, such as the system is prone to plunging into locally extreme state when the learning algorithm is not selected suitably, there exists a conflict between the network complexity and its generalization and so on. From the deep analysis of the existing network models and algorithms, we can learn about that their methods of setting parameters lack the capability of meeting an actual situation, so that some torpidity appears when solving problems, which is conducive to the universality of the structure or algorithm but neglects the assistant function of the characteristics or knowledge. Based on this consideration, this paper aims at introducing the concept of immunity into some existing artificial neural networks, so as to design a novel network model which can use the characteristic knowledge for solving problem. This model is presently called immune neural network (INN) and it is used for improving the capability of dealing with some difficult problems.Secondly, in the research of communication, the direct-sequence code division multiple access (DS-CDMA) is an important component of numerous recent communications systems, proposed and implemented. However, due to nonorthogonality of practical spreading sequences, the conventional correlator suffers from the near-far problem. This implies that the cross correlation between the spreading sequence of the user of interest and the signal from a strong interfere can be larger than the correlation with the signal from the desired user. Detection is re...