Orthogonal precoding constitutes a powerful technique to reduce spectrum sidelobes of multicarrier signals. This reduction is bought at the cost of introducing precoder redundancy, which results in some throughput loss and additional precoding/decoding complexity. When the goal is to meet some spectral emission mask constraints, it is desirable to avoid unnecessary sidelobe suppression in order to keep precoder redundancy at a minimum. In this context, we introduce a general framework under which we develop a novel Lagrange multiplier-based mask-compliant orthogonal precoder design targeting minimal redundancy. We also adapt to this framework two previously proposed designs based on spectral notches and minimum out-of-band emission, respectively, to explicitly incorporate mask constraints. Simulation results are provided to show the effectiveness of the proposed designs under different practical masks for multicarrier wireless systems.

Abstract One of the major drawback of an orthogonal frequency division multiplexing based cognitive radio is high out‐of‐band leakage that can severely affect primary users operating in adjacent bands. The paper aims to design a new spectral precoding scheme that minimizes Adjacent Channel Leakage Ratio (ACLR) while retaining the orthogonality among data symbols. The problem is formulated as trace ratio minimization subjected to unitary constraint that can be effectively solved using iterative methods. The proposed precoder's error performance is similar to uncoded orthogonal frequency division multiplexing system due to orthogonal multiplexing. Also, it is practically realizable due to lower complexity. The scheme also extends for distinct ACLR requirements in upper and lower adjacent bands. A relative important factor is introduced so as to flexibly control the interference leakage in upper and lower adjacent channels. An algorithm is suggested for choosing suitable values of this factor. Simulation results display the superiority of proposed scheme over existing techniques in terms of ACLR and throughput. Copyright © 2016 John Wiley & Sons, Ltd.