Kalman Filter Based Channel Tracking for Precoded GFDM Systems

Generalized Frequency Division Multiplexing (GFDM) offers the simultaneous benefits of flexibility, low latency and excellent spectral properties over conventional OFDM for modern wireless systems, making it a promising candidate for 5G and other wireless technologies. However, several implementation related issues remain, including estimation and equalization. We consider the problem of efficient estimation and channel tracking in GFDM systems. In particular, we discuss a scattered (and tessellated) pilot based channel estimation approach, and derive a Kalman filter based channel tracker to handle time varying channels. We extend this work to the GFDM-MIMO case. Simulations reveal that the proposed tracker minimizes estimation errors which significantly improves the BER performance of the system.