A Seysen's algorithm–based incremental lattice reduction. Issue 7 (22nd March 2019)

• Record Type: Journal Article
• Authors: He, Qinwei; Hu, Yulin; Schmeink, Anke
• Is Part Of: Transactions on emerging telecommunications technologies. Volume 30:Issue 7(2019)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-03-22
• Subjects: Telecommunication -- Periodicals; 384.05
• DOI: 10.1002/ett.3596
• Languages: English
• ISSNs: 2161-5748

Abstract

Abstract: Lattice reduction (LR)–aided detections have attracted great attention in the symbol detection of multiple‐input–multiple‐output communication, especially for systems with larger scale antennas. Recently, the Lenstra‐Lenstra‐Lovász (LLL)–based incremental LR (ILR) algorithms have been proposed to employ an early termination to jointly conduct the LR and retrieve the symbol based on a partial successive interference cancellation (SIC) detection, which induces a significant performance improvement. In this paper, we propose to apply the Seysen's algorithm (SA) to the ILR. After providing the feasibility analysis, two novel SA involved algorithms are introduced and studied. In particular, a new sorted QR decomposition‐based SA algorithm is proposed and tested as it plays a key role in the SIC detection. Subsequently, a novel SA‐based ILR scheme is further developed and investigated. By simulations, we show that the proposed approach outperforms conventional LLL‐based ILR in various aspects, while only a marginal increment in the complexity, restricted to medium signal‐to‐noise‐ratio regions, is observed. Abstract : In this paper, two novel algorithms, i.e., sorted QR decomposition (SQR) based Seysen's algorithm (SA) and SA based incremental lattice reduction (ILR) algorithm are introduced and evaluated. By simulations, we first validate the effectiveness of the SQR‐SA algorithm as it plays the key role in the SA‐ILR scheme. Afterwards, we demonstrate that the proposed SA‐ILR outperforms conventional LLL‐ILR in various aspects, while only a marginal increment in the complexity, restricted to medium SNR regions, is observed.