Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Computer Engineering and Sciences

First Advisor

Josko Zec

Second Advisor

Veton Z. Kepuska

Third Advisor

Ivica Kostanic

Fourth Advisor

Ersoy Subasi


This dissertation has proposed several algorithms to optimize the allocation of pilots to the users’ equipment (UEs) to mitigate the effect of the pilot contamination problem in the massive MIMO systems. Pilot contamination reduces the performance of massive MIMO systems due to the reduction in the quality of the estimated channel between a UE and the serving base station (BS). The limitation of the number of samples in a coherence block limits the number of unique mutually orthogonal pilots, and hence, reusing the set of pilots across the cells causes inter-cell interference during pilot transmission, which is called pilot contamination. For this purpose, algorithms that aim to maximize the minimum asymptotic signal to noise ratio (SIR) by optimizing the pilot allocation have been proposed. The goal is to enhance the SIRs for the UEs that suffer the most from the pilot contamination problem. On the other hand, pilots’ allocation is a combinatorial problem, and generally nondeterministic polynomial-time hard (NP-hard). Hence, finding the optimal solution is obtained by the exhaustive search, which is not feasible when the number of UEs and cells are large. Thus, we have regenerated the state-of-the-art scheme to evaluate the proposed algorithms when the optimal solution could not be found due to computational complexity. While all the proposed algorithms have performed better than the related work, two types of the proposed algorithms stand out by providing a near-optimal solution with polynomial complexity in the scenarios when the exhaustive search was applicable, and much outperformed the state-of-art algorithm especially in the simulations of the large systems. Also, we have noticed from the simulations of large systems, that applying the proposed pilot allocation schemes to the system that uses the zero-forcing combining scheme, where a BS is deployed with a very large number of antennas, has greatly raised the spectral efficiency of the UEs that have low spectral efficiency using conventional and state-of-art pilot allocation schemes.


Copyright held by author