Cell search for a millimeter wave cellular system with 1-bit quantization at the receiver

For 5G it will be important to leverage the avail- able millimeter wave spectrum. Communication at high carrier frequencies requires antenna arrays at both the base and mobile station. A 1-bit analog to digital converter can effectively reduce the complexity and power consumption of the analog frontend. This is especially interesting in the context of large antenna arrays with a sizable signal bandwidth. In this paper we investigate the cell search capabilities of a mobile station with an antenna array and an analog frontend with 1-bit quantization behind each antenna. We derive the Likelihood Ratio Test (LRT) for this problem with full channel knowledge. We also show the generalized Likelihood Ratio Test (GLRT) for a case without channel knowledge, and compare the performance to practical algorithm based on correlation. We show that the practical correlation achieves a similar performance compared to the theoretical optimal GLRT. Compared to a system with high resolution analog to digital conversion our results show the theoretical derived SNR gap of 2 dB.