The stability theory of stream ciphers suggests that good key stream sequences must not only have a large linear complexity, but also change of a few terms must not cause a significant drop of the Linear complexity. This unfavorable property leads to the concept of k-error linear complexity. In this project we propose to extend this to the case of multisequences.
