Blind Pilot Decontamination

Abstract

A subspace projection to improve channel estimation in massive multi-antenna\nsystems is proposed and analyzed. Together with power-controlled hand-off, it\ncan mitigate the pilot contamination problem without the need for coordination\namong cells. The proposed method is blind in the sense that it does not require\npilot data to find the appropriate subspace. It is based on the theory of large\nrandom matrices that predicts that the eigenvalue spectra of large sample\ncovariance matrices can asymptotically decompose into disjoint bulks as the\nmatrix size grows large. Random matrix and free probability theory are utilized\nto predict under which system parameters such a bulk decomposition takes place.\nSimulation results are provided to confirm that the proposed method outperforms\nconventional linear channel estimation if bulk separation occurs.\n

Keywords

Affiliated Institutions

• Aalto University FI
• Norwegian University of Science and Technology
• EURECOM FR
• Friedrich-Alexander-Universität Erlangen-Nürnberg DE
• KTH Royal Institute of Technology SE

Related Publications