Non-Orthogonal Multiple Access with Improper Gaussian Signaling

    Research output: Contribution to journalArticle

    Published

    View graph of relations

    Improper Gaussian signaling (IGS) helps to improve the throughput of a wireless communication network by taking advantage of the additional degrees of freedom in signal processing at the transmitter. This paper exploits IGS in a general multiuser multi-cell network, which is subject to both intra-cell and inter-cell interference. With IGS under orthogonal multiple access (OMA) or non-orthogonal multiple access (NOMA), designs of transmit beamforming to maximize the users’ minimum throughput subject to transmit power constraints are addressed. Such designs are mathematically formulated as nonconvex optimization problems of structured matrix variables, which cannot be solved by popular techniques such as weighted minimum mean square error or convex relaxation. By exploiting the lowest computational complexity of2×2 linear matrix inequalities, lower concave approximations are developed for throughput functions, which are the main ingredients for devising efficient algorithms for finding solution of these difficult optimization problems. Numerical results obtained under practical scenarios reveal that there is an almost two-fold gain in the throughput by employing IGS instead of the conventional proper Gaussian signaling under both OMA and NOMA; and NOMA-IGS offers better throughput compared to that achieved by OMA-IGS.

    Documents

    DOI

    Original languageEnglish
    Number of pages12
    Pages (from-to)496-507
    JournalIEEE Journal of Selected Topics in Signal Processing
    Journal publication date22 May 2019
    Issue number3
    Volume13
    Early online date27 Feb 2019
    DOIs
    Publication statusPublished - 22 May 2019

    ID: 170076027