Spatial modulation and recurrent neural network based equalizer for MIMO communication systems

Spatial multiplexing is a prominent feature of MIMO communication system to achieve higher spectral efficiency. Spatial modulation is an indexing-based approach to achieve effective spatial multiplexing and reduce the effect of inter channel interference (ICI). The objective of the research in this paper lies in developing an efficient equalizer by using recurrent neural network (RNN) architecture for both outdoor and indoor transmission considering the practical fading channel statistics. The equalizer performance is tested using standard channel models like Rayleigh, Rician and IEEE 802.11 for single user MIMO. Simultaneous use of spatial PSK modulation and adaptive equalization has the capability to mitigate the channel impairments and enhance the receiver performance.