Low complexity Sub-optimal GFDM receiver for remote areas

Abstract

Generalized Frequency Division Multiplexing (GFDM) is a waveform, initially proposed for 5G, that can play an important role in some applications. It offers low out-of-band emission, good spectral efficiency and has the flexibility to cover the 5G requirements in remote area scenarios. The GFDM receiver implementation complexity poses restrictions to the use of this powerful waveform in constrained resources devices, such as the ones used in Internet of Things (IoT). This thesis proposes, a sub-optimal receiver chain using frequency-domain equalization, offering a complexity comparable to the Zero-Forcing (ZF) receiver while achieving performance close to the Minimum Mean Square Error (MMSE) receiver. The study also evaluates the proposed equalizer performance under frequency selective fading channel for the SingleInput/Single-Output (SISO) and Multiple-Input/Multiple-Output (MIMO) cases. Simulation results show that with a small increment in complexity, in comparison with ZF, a performance gain in Symbol Error Rate (SER) can be obtained. An implementation of the proposed chain using the Software Defined Radio (SDR) approach, suggests that the proposed scheme is feasible for practical real-time deployments.