On the Capacity of Special Classes of Gaussian Relay Networks with Orthogonal Components and Noncausal State Information at Source

TL;DR

This paper analyzes the capacity of specific Gaussian relay networks with orthogonal components and noncausal channel state information at the source, deriving bounds and exact capacities for special cases.

Contribution

It introduces an upper bound on the capacity of relay networks with noncausal CSI and derives exact capacities for two special Gaussian classes.

Findings

Upper bound on the capacity for the discrete memoryless model.

Capacity characterization for a class with no interference at relays and destination.

Exact capacity results for specific Gaussian relay network classes.

Abstract

In this paper, we study relay networks with orthogonal components in presence of noncausal channel state information (CSI) available at the source. We propose an upper bound on the capacity of the discrete memoryless model (DM)for the case in which just the source component intended for the destination is encoded against the CSI known non-causally at the source. Also, we derive capacity for two special classes of the Gaussian structure of the model. The first class is the one for which we have obtained the upper bound and the second class is the one in which all of the source components intended for the relays and destination are encoded against the noncausal CSI, however, no interference at the relays and destination exists in this case.