Cooperative MIMO relay beamforming in overloaded DS/CDMA multicast networks

• Published in: IET communications 2017-02, Vol.11 (3), p.344-354
• Main Authors: Mehrizi, Sajad, Mahboobi, Behrad
• Format: Article
• Language: English
• Subjects: antenna number; array signal processing; Beamforming; code division multiple access; computational complexity; concave programming; cooperative communication; cooperative MIMO relay beamforming; DC programming method; difference‐of‐two convex functions; direct sequence code division multiple access network; indefinite quadratic programming; Interference; MIMO communication; MIMO‐CDMA scheme; Multicast; multiinput multioutput nonregenerative relays; multiple‐access interference; Networks; nonconvex programming; optimisation problem; overloaded synchronous DS‐CDMA multicast networks; Programming; radiofrequency interference; Receivers; Relay; relay networks (telecommunication); SDP; SINR constraint; SOCP; spread spectrum communication; total relay transmission power minimization; Transmitters
• ISSN: 1751-8628; 1751-8636; 1751-8636
• DOI: 10.1049/iet-com.2016.0610

A new beamforming scheme is proposed for multi-input multi-output (MIMO) non-regenerative relays in an overloaded synchronous direct sequence code division multiple access (DS-CDMA) network in which the number of users is larger than the processing gain. In this scheme, several transmitters multicast their data to their groups of receivers through MIMO relays and DS-CDMA technique is utilised to decrease the level of multiple access interference (MAI) at the relays and the receivers. At the relays, the data of transmitters are despread, beamformed and re-spread. The aim of the beamformer is to minimise the total relays transmission power subject to the signal to interference and noise ratio (SINR) constraints at the receivers. This optimisation problem is an indefinite quadratic (non-convex) programming. Semi-definite programming (SDP) and second-order cone programming (SOCP) are two existing methods to solve the problem. However, they, respectively, suffer from high computational complexity and performance degradation when the number of antennas or users increase. To effectively mitigate these deficiencies, we propose a novel DC (difference of two convex functions) programming method. Our simulation results demonstrate the computational advantage of the proposed DC programming over SDP, while a better performance over SOCP is achieved. Moreover, the superiority of the proposed MIMO-CDMA scheme over other multiplexing schemes is shown by several scenarios.