We derive a new outer bound on the capacity region of broadcast traffic in multiple input broadcast packet erasure channels with feedback, and extend this outer bound to packet erasure relay networks with feedback. We show the tightness of the outer bound for various classes of networks. An important engineering implication of this work is that for network coding schemes for parallel broadcast channels, the “xor” packets should be sent over correlated broadcast subchannels.

This paper studies and derives new results for the interference channel with a relay (ICR). Three inner bounds for the discrete memoryless ICR are proposed, based on three coding strategies that employ layered code at the relay. The first scheme is inspired by layered noisy network coding, proposed by Lim et al. for the two-way relay channel, the second and the third schemes rely on simpler encoding and decoding processes, dubbed layered quantize-forward. Performance of the proposed schemes is investigated for two classes of channels with Gaussian noise: the interference channel with in-band relay reception/out-of-band relay transmission and the interference with in-band relay reception/in-band relay transmission. For the former class of channels, it is shown that the first proposed scheme achieves the same inner bound as the generalized hash-forward scheme with incremental binning. In addition, the inner bound is within 0.5 bit of the capacity region under certain conditions on the channel parameters. For the latter class of channels, new upper bounds on sum-rate are established by extending known upper bounds for symmetric channels. The first inner bound is shown to be within 0.5 bit of the capacity region if the relay's power exceeds a certain threshold, which depends on channel parameters. Numerical examples show that the proposed schemes can achieve significantly higher sum-rates when compared with other compress-forward schemes. Analysis also reveals a tradeoff between achievable rates, coding delay, and complexity of the proposed schemes. Results in this paper provide a better understanding of coding for the ICR, in particular, they show that layered coding is a beneficial element in multiuser networks with relays.

This paper studies a new model for cooperative communication, the interfering relay channels. We show that the hash-forward scheme introduced by Kim for the primitive relay channel is capacity achieving for a class of semideterministic interfering relay channels. The obtained capacity result generalizes and unifies earlier capacity results for a class of primitive relay channels and a class of deterministic interference channels.

This paper studies a model for communications in wireless networks supported by designated cooperation links. In particular, a 2-user Gaussian one-sided interference channel with two rate-limited and orthogonal communication links between the receivers is considered. A communication protocol for the channel is proposed, which combines rate-splitting and superposition encoding techniques with the conventional decode-forward and compress-forward strategies. It is shown that a careful design of codebooks and coding scheme, which is obtained from intuition based on superposition coding, can greatly reduce the complexity of the strategy. Analytical and numerical results show that the proposed scheme, although not universally optimal, can achieve the capacity region or sum capacity exactly or asymptotically in certain scenarios. Various limits of sum capacity gain due to cooperation are also discussed.

This paper proposes two new coding schemes for the discrete memoryless two-way relay channel. The main target is to show the benefits of compress-forward without Wyner-Ziv binning and of layered relaying in networks wherein a relay is to help multiple destinations, that may have unequal channel quality and/or have access to different side information. Numerical results for a Gaussian channel show that the new coding schemes outperform variants of compress-forward relaying and offer a good trade-off between achievable rates and complexity and decoding delay. The idea can also be applied to other relay networks.

We present a new achievable rate region for the discrete memoryless Z channel (DM-ZC) using Marton coding with rate splitting. The region is shown to include previously known achievable rate regions. Secondly we study a class of degraded Z channels, the bijective degraded Z channel (BDZC). An outer bound for the BDZC is proved, which is shown to meet the inner bound for the deterministic settings. For the Gaussian Z channel with weak crossover link, we show that if Gaussian inputs are optimal then a coding scheme based on Marton coding without rate splitting achieves to within half a bit per real dimension from the boundary of the capacity region.

We derive achievable rate regions for a 2-user Gaussian Z-interference channel with conferencing decoders. We identify different cases where the rate-limitedness of the conference link from the interference-free receiver to the interfered receiver affects the conferencing strategy as well as the achievable rate region. Furthermore, an outer bound to the capacity region based on cut-set and genie-aided bounds is presented.

In this work we provide an achievable rate region for the cellular downlink with three users where two users want to communicate with each other. Due to the side information from the prior uplink, gains are achievable by combining bidirectional and classical broadcast channel coding strategies. A coding theorem for the bidirectional broadcast channel with random state non-causally known at the encoder is generalized to continuous alphabets and applied to Gaussian channels with an average power constraint. The single-user capacities are achievable if one decoder additionally knows the channel state. Accordingly, we see that a more comprehensive view on the information flow in a multi-user network can lead to a larger achievable rate region.

This paper considers a discrete memoryless four-node network where two nodes want to send three independent messages to the other two nodes. The two receiving nodes are allowed to cooperate by means of a unidirectional noiseless link with finite capacity. A coding scheme is proposed which combines rate splitting, block Markov multi-level superposition coding with binning and joint decoding. The general achievable rates are then specialized to degraded channel and Gaussian channel, where it is shown that the sum capacity for the Gaussian channel is achieved under certain conditions. Results in this paper recover and unify previously known results for the discrete memoryless Z channel without cooperation, and results for the Gaussian Z-interference channel with a digital relay link.

A Gaussian multiple-input multiple-output broadcast channel (MIMO GBC) is considered. Throughout the paper it is assumed that 1) input signals are Gaussian and 2) perfect channel state information is available at the transmitter and at the receivers. By considering each data stream as a single user, the uplink-downlink signal-to-interference-plus-noise (SINR) duality is generalized to the MIMO case with general cross-talk matrix. The duality is subsequently applied to finding the solution for the SINR-balancing problem. The result serves as a tool for characterizing achievable rate regions of different coding strategies. Next, we investigate a superposition coding scheme proposed by Cover-van der Meulen-Hajek and Pursley (nicknamed CMHP), where there is a common message to both users. We consider a MIMO broadcast channel with two users, each user has two antennas and the transmitter has four antennas. Assuming one common stream is sent by CMHP coding and successive decoding, a lower bound to the CMHP rate region is found. Behaviors of the CMHP rate region and sumrate are analyzed. We find the sumrate gaps between DPC, CMHP, and MMSE at high SNR for general 2-user multiple-input singleoutput (MISO) Gaussian broadcast channel. The result suggests when CMHP is beneficial for sumrate.