This study presents a method of generating long-period pseudo-random sequence based on a plaintext and negative selection algorithm for real applications. It is a new method that extracts the generated factor from the plaintext. Moreover, the generated factor as a group is expanded by the way of the m-sequence and generates pseudo-random. The autocorrelation and statistical properties of the method are better. If it used for data encrypted, it has a good anti-attack performance. In addition, the generate factor can also be used for immune identification in receiver end, to verify data integrity. So there will be a good prospect.