1st International Conference on 5G for Ubiquitous Connectivity

Research Article

Performance Analysis of Block and Comb Type Channel Estimation for Massive MIMO Systems

Download661 downloads
  • @INPROCEEDINGS{10.4108/icst.5gu.2014.258076,
        author={Gabor Fodor and Piegiuseppe Marco and Miklos Telek},
        title={Performance Analysis of Block and Comb Type Channel Estimation for Massive MIMO Systems},
        proceedings={1st International Conference on 5G for Ubiquitous Connectivity},
        publisher={IEEE},
        proceedings_a={5GU},
        year={2014},
        month={12},
        keywords={channel estimation mimo least square minimum mean square error},
        doi={10.4108/icst.5gu.2014.258076}
    }
    
  • Gabor Fodor
    Piegiuseppe Marco
    Miklos Telek
    Year: 2014
    Performance Analysis of Block and Comb Type Channel Estimation for Massive MIMO Systems
    5GU
    IEEE
    DOI: 10.4108/icst.5gu.2014.258076
Gabor Fodor,*, Piegiuseppe Marco1, Miklos Telek2
  • 1: Royal Institute of Technology and Ericsson Research
  • 2: Budapest University of Technology and Economics
*Contact email: gabor.fodor@ericsson.com

Abstract

For pilot sequence based multiple input multiple output (MIMO) channel estimation, the arrangements of pilot symbols, such as the block or comb type arrangement,is known to play an important role. In this paper we compare the performance of block and comb pilot symbol patterns in terms of uplink mean square error (MSE) and spectral efficiency when the receiver at the base station employs least square (LS) or minimum mean square error (MMSE) channel estimation and MMSE equalizer for uplink data reception. For this system, we derive a closed form solution for the MSE and spectral efficiency that allows us to obtain exact results for an arbitrary number of antennas. Our key observation is that the comb pilot arrangement allows for unequal pilot-data power allocation in the frequency domain, which leads to a significant spectral efficiency increase. This spectral efficiency increase is particularly important with LS estimation and as the number of base station antennas grows large. It also gives noticeable gains with MMSE estimation. Our main conclusion is that with a large number of antennas, unequal power allocation facilitated by comb arrangement can give large gains over alternative pilot arrangements.