Blind Source Separation for OFDM with Gradient Algorithms

Signal recovery mechanism for Orthogonal Frequency Division Multiplexing (OFDM) with the assistance of principles of Blind Source Separation (BSS) and Gradient Algorithm (GA) is presented. Elimination or reduction of undesirable influences encountered with in the wireless interface is targeted by using a set of filter coefficients. Four scalar energy functions are used in deriving the filter coefficients. Energy criterion functions to be optimized and the performance is justified. Time correlation properties of the channel are used as advantages in introducing the energy functions. Performance of the algorithms is evaluated when they are used for a synchronous downlink OFDM model. Simulations are carried out under slow fading channel conditions with a receiver containing Equal Gain Combining (EGC) and BSS algorithms. It could be noted that, better performance for this predominant multicarrier air interface communication technique can be achieved with this combined scheme. It is important grasp the fact that, these schemes can be promoted as low complexity simple matrices based processing mechanisms.