An Efficient Large Resource-User Scale SCMA Codebook Design Method

Abstract

Sparse code multiple access (SCMA) is an attractive non-orthogonal multiple access (NOMA) technology. An effective SCMA codebook design metrics is maximizing minimum Euclidean distance between superimposed codewords. However, the number of superimposed codewords will significantly expand with the growing number of resources and users, making the max–min problem impractical. To this end, an efficient suboptimal SCMA codebook design method is proposed for large source-user scale. We first project multi-dimensional superimposed codewords to resources to reduce the number of superimposed codewords in the constellation. Then we formulate the max–min optimization as a nonconvex quadratically constrained quadratic programming (QCQP) problem and solve it by semidefinite relaxation (SDR). Simulation result shows that the method, which can dramatically reduce computational complexity, is able to solve large-scale codebook design problem while incurring little loss in performance compared with the existing codebook design method.