Optimized Blind Algorithms for Widely Linear Beamforming

The use of array beamforming in modern wireless communication systems has been increasingly investigated, due to their potential in rejecting interference and improving the system capacity. The so-called blind algorithms have been considered in such application in order to avoid training procedures and improve the overall transmission throughput. Most works in the field consider circular signals and classical linear signal processing framework. However, the alternative scenario of Widely Linear Processing was shown to be more suitable under other circularity conditions, particularly in rectilinear modulations schemes. This paper deals with blind algorithms for array beamforming in a Widely Linear Processing framework. We consider the well-known CMA and NCMA algorithms and derive corresponding Widely Linear formulations with optimized performance. In addition, we show that their implementation requires less computational cost than the original CMA and NCMA. Simulation results confirm that both proposed algorithms perform better than their strictly linear counterparts.