SPARSE AND LOW-RANK OPTIMIZATION FOR PLIABLE INDEX CODING

In this paper, we investigate the pliable index coding problem, where clients are interested in receiving any messages (instead of specific messages) that they do not have. The motivating applications including caching networks, recommendation systems and distributed computing systems, where the clients are happy to receive any messages not available in them. However, the pliable index coding problem turns out to be computationally intractable, for which we propose a novel sparse and low-rank optimization framework to assist efficient algorithms design in real field, thereby minimizing the number of channel uses for message delivery. To address the nonconvex challenges in this framework, we further propose the alternating projection algorithm to solve the sparse and low-rank optimization problem with local convergence guarantees. Simulation results demonstrate that the number of channel uses can be significantly reduced for message delivery via the sparse and low-rank optimization.