Title:  Efficient LLL-based Lattice Reduction Algorithms for MIMO Detection

Committee: 

Dr. Ma, Advisor  

Dr. Chang, Chair

Dr. Baxley

Abstract: The objective of the proposed research is to design efficient Lenstra-Lenstra-Lovasz (LLL) based Lattice Reduction (LR) algorithms for multiple-input multiple-output (MIMO) detection in wireless communications. The original LLL algorithm is challenging to directly implement in practice due to its nondeterministic structure and slow convergence. To solve these problems, we first investigate the relationship between the LLL algorithms and the error performance of LR-aided MIMO detectors, as well as the inherent characteristics of the LLL algorithm. By exploiting these investigations, we then develop enhanced fixed-complexity LLL (fcLLL) and greedy LLL algorithms, which not only converge faster but also exhibit much lower complexity than the existing fcLLL and greedy LLL algorithms while the error performance of the MIMO system is maintained.