TY - JOUR
T1 - Precoder Design and Statistical Power Allocation for MIMO-NOMA via User-Assisted Simultaneous Diagonalization
AU - Krishnamoorthy, Aravindh
AU - Ding, Zhiguo
AU - Schober, Robert
N1 - Funding Information:
Manuscript received May 6, 2020; revised September 7, 2020; accepted October 26, 2020. Date of publication November 6, 2020; date of current version February 17, 2021. The work of Zhiguo Ding was supported by the UK EPSRC under Grant EP/P009719/2. This article was presented in part at the 2019 IEEE International Conference on Communication (ICC) and the 24th International ITG Workshop on Smart Antennas. The associate editor coordinating the review of this article and approving it for publication was F. Zhou. (Corresponding author: Aravindh Krishnamoorthy.) Aravindh Krishnamoorthy is with the Institute for Digital Communications, Friedrich-Alexander-Universität Erlangen-Nürnberg, 91058 Erlangen, Germany, and also with the Fraunhofer Institute for Integrated Circuits (IIS), 91058 Erlangen, Germany (e-mail: [email protected]).
Publisher Copyright:
© 1972-2012 IEEE.
PY - 2021/2
Y1 - 2021/2
N2 - In this paper, we investigate the downlink precoder design for two-user power-domain multiple-input multiple-output (MIMO) non-orthogonal multiple access (NOMA). We propose a novel user-assisted (UA) simultaneous diagonalization (SD) based MIMO-NOMA scheme that achieves SD of the MIMO channels of both users through a combination of precoder design and low-complexity self-interference cancellation at the users, thereby considerably lowering the overall decoding complexity compared to joint decoding. The achievable ergodic user rates of the proposed scheme are analyzed for Rayleigh fading channels based on a finite-size random matrix theory framework, which is further exploited to develop a statistical power allocation algorithm. Simulation and numerical results show that the proposed UA-SD MIMO-NOMA scheme significantly outperforms orthogonal multiple access and a benchmark precoder design performing SD via generalized singular value decomposition in terms of the achievable ergodic rate region for most user rates. The ergodic rate region is further enhanced by a hybrid scheme which performs time sharing between the proposed UA-SD MIMO-NOMA scheme and single-user MIMO.
AB - In this paper, we investigate the downlink precoder design for two-user power-domain multiple-input multiple-output (MIMO) non-orthogonal multiple access (NOMA). We propose a novel user-assisted (UA) simultaneous diagonalization (SD) based MIMO-NOMA scheme that achieves SD of the MIMO channels of both users through a combination of precoder design and low-complexity self-interference cancellation at the users, thereby considerably lowering the overall decoding complexity compared to joint decoding. The achievable ergodic user rates of the proposed scheme are analyzed for Rayleigh fading channels based on a finite-size random matrix theory framework, which is further exploited to develop a statistical power allocation algorithm. Simulation and numerical results show that the proposed UA-SD MIMO-NOMA scheme significantly outperforms orthogonal multiple access and a benchmark precoder design performing SD via generalized singular value decomposition in terms of the achievable ergodic rate region for most user rates. The ergodic rate region is further enhanced by a hybrid scheme which performs time sharing between the proposed UA-SD MIMO-NOMA scheme and single-user MIMO.
KW - linear algebra
KW - multiple access
KW - Multiple-input multiple-output (MIMO)
KW - non-orthogonal multiple access (NOMA)
UR - https://www.scopus.com/pages/publications/85096833340
U2 - 10.1109/TCOMM.2020.3036453
DO - 10.1109/TCOMM.2020.3036453
M3 - Article
AN - SCOPUS:85096833340
SN - 0090-6778
VL - 69
SP - 929
EP - 945
JO - IEEE Transactions on Communications
JF - IEEE Transactions on Communications
IS - 2
M1 - 9250605
ER -