Energy Efficient Sparse Precoding Design for Satellite Communication System

Through precoding, the spectral efficiency of the
system can be improved; thus, more users can benefit from
5G and beyond broadband services. However, complete precoding (using all precoding coefficients) may not be possible in
practice due to the high signal processing complexity involved
in calculating a large number of precoding coefficients and
combining them with symbols for transmission. In this paper,
we propose an energy-efficient sparse precoding design, where
only a few precoding coefficients are used with lower power
consumption depending on the demand. In this context, we
formulate an optimization problem that minimizes the number of
in-use precoding coefficients and the system power consumption
while matching the per beam demand. This problem is nonconvex. Hence, we apply Lagrangian relaxation and successive convex approximation to convexify it. The proposed solution
outperforms the benchmark scheme in power consumption and demand satisfaction with the additional advantage of sparse precoding design.