Hardware-Efficient Time-Domain-Enhanced Channel Estimation for 5G New Radio RedCap

Enhanced battery lifetime, cost and chip-size reduction drive 5G New Radio (NR) to relax requirements down to single-antenna reception making channel estimation pivotal for a Reduced Capability (RedCap) device. Thereby, potential channel-estimation algorithms have to be evaluated in terms of performance as well as in terms of their silicon-area footprint. In this paper, the suitability of two channel-estimation algorithms is analyzed: A standard Minimum Mean Square Error (MMSE) receiver with Moving Average(MA) filtering as well as a less common Time-Domain-Enhanced (TDE) MMSE approach, which especially includes an unallocated spectrum extrapolation as well as a threshold-based significant tap selection. The presented system-level performance results demonstratethat the TDE algorithm outperforms the standard one by 0.9 dB under strong multipath conditions. In addition, it does not require any additional channel knowledge making it very robust in the field. On the contrary, the standard MMSE algorithm is highly sensitive to the window size that depends on channel knowledge. Accurate silicon-area estimations based on an optimized architecture indicate an only $9 \%$ overhead over a MMSE realization which can be considered neglectable relative to the chip size of a completeRedCap modem system-on-a-chip.