A genetic algorithm to design error correcting codes

During the digital information transmitting through a channel, practically inevitable errors due to the presence of noise and to other factors are produced, such as interferences, echoes, etc. Then, it is necessary to establish ways, if not to avoid the errors, at least for be able to recognize its presence and, if it is possible, to correct them. Correcting of some manner the message in the receiver side cheapens the price of a transmission process, therefore otherwise there would be that to repeat the message when is corrupted along the way. In these cases the use of error correcting codes is suitable. In the actual communication systems the control of error is performed by means of the application of special codes that add redundancy. This aggregate redundancy permits to detect and/or to correct the errors occurred during the data transmission. In this work we center us in the binary linear block codes. The problem to find an error correcting code that corrects a maximum number of errors given is a NP-complete problem. By this reason, we have undertaken this problem by means of a genetic algorithm. In this article we expose the implementation and results of the genetic algorithm, having obtained some good values compared with other works.