Blind 2-channel identification based on discrete orthogonal learning rule
Huaqiu Deng, Li Liu, Wenhua Weng, Jiping Yu, Ying-Lin Yu · 2002
A discrete neural network is proposed to solve a blind channel identification problem under wireless communication. Under a discrete orthogonal learning rule, the synapses vector of the discrete neural network globally converges to a null space of an input data matrix for almost all initial conditions. Simulation results show that the discrete neural network functions very well.