Relays in Interference Limited Cellular Systems

A relay-aided cellular system
A relay-aided cellular system. In addition to the desired signal from the home base station, each relay or user receives interference from other-cell base stations or relays.

Relays in Interference Limited Cellular Systems

Cellular systems are still not enjoying the high capacity and coverage gains that were heralded by advances in communication theory including multiple antennas, better modulation, and near-Shannon error control codes. The reason is that many of these advances largely neglect conditions found at the cell edge: high path loss, shadowing, and significant interference. Current cellular networks do little to mitigate or exploit the interference at the edge of cells other than spatial reuse and soft handoff. Future networks are being designed to exploit base station coordination or network MIMO, on the uplink and downlink, at the expense of a potentially excessive amount of overhead. While a promising technology, coordinating the base stations (see our work on multicell MIMO communication and multicell MIMO limited feedback) requires architectural changes to the network including higher bandwidth backhaul connections with lower latencies. A near-term solution is to employ relays, infrastructure equipment that assists in conveying signals between the base station and the subscriber station, but which is not connected via wire to the network backbone. Standard relay approaches, though, neglect interference and thus do not perform as well in cellular environments as predicted by the theory derived under interference free assumptions.

In our recent work we have been revisiting relay performance in interference limited cellular systems. We have studied how several common relay architectures degrade in the presence of out-of-cell interference including one-way and two-way relays. To solve this problem we are investigating a variety of different approaches to reducing the impact of interference.

One approach is to mitigate interference at the relay, using spatial interference cancellation, combined with transmit beamforming to reduce the out-of-cell interference caused by relays. Another approach is to consider shared relays, which decode signals from multiple base stations, and thus are designed to exploit not combat interference. Yet another alternative is to consider hybrids of different approaches, for example a three-phase two-way shared relay that combines the benefits of several promising architectures.


Several of our recent results and publications are summarized below.

R. W. Heath, Jr.A. Y. Panah, K. T. Truong,and  S. W. Peters, “Where are the Relay Gains in Cellular Systems?” invited presentation given at the IEEE Communication Theory Workshop, May 2010.

This presentation summarizes several of our recent results on improving relays in the presence of out-of-cell interference. First it reviews potential capacity gains from relaying and illustrates how these gains behave in the presence of out-of-cell interference. It then proposes several different strategies, using multiple antennas, to improve the downlink in relay-based cellular systems. Multiple antenna relays, operating in interference cancelation mode, cancel strong interferers and improve cell edge performance. Shared relays, where relays support transmissions from multiple base stations, are shown to improve cell edge throughput. Two-way extensions of the shared relay are suggested to reduce the half-duplex loss.

A. Y. Panah, K. T. Truong, S. W. Peters, and  R. W. Heath, Jr., “Interference Management Schemes for the Shared Relay Concept,” EURASIP Journal on Advances in Signal Processing, special issue on Cooperative MIMO Multicell Networks, vol. 2010, Article ID 269817, 14 pages, doi:10.1155/2011/269817, 2010.

This paper proposes several interference management strategies for the shared relays ranging. It considers different relay operation including amplify-and-forward, decode-and-forward, and two-way extensions. This paper provides several comparisons between strategies and provides throughput as a function of distance for the uplink and downlink with different strategies. The simulation framework is the same as in our other EURASIP relay paper.

S. W. Peters, A. Y. Panah, K. T. Truong, and  R. W. Heath, Jr., “Relaying Architectures for 3GPP LTE-Advanced,” EURASIP Journal on Advances in Signal Processing, special issue on 3GPP LTE and LTE Advanced, vol. 2009, Article ID 618787, 14 pages, doi:10.1155/2009/618787, 2009.

This paper compares the performance of one-way, two-way, and shared relays in a 3GPP LTE-inspired deployment. A major assumption in the paper is single stream transmission; sophisticated MIMO techniques are not used for transmission. The shared relay is found to have performance that is some cases is not too far from complete base station coordination. The half-duplex one-way relay provides better performance than the direct path in only a few cases. The shared-relay is found to be too sensitive to out-of-cell interference to provide much gain. An additional contribution of the paper is a simulation framework that we are using in subsequent papers.

S. W. Peters and  R. W. Heath, Jr., “The Future of WiMAX: Multihop Relaying with IEEE 802.16j,” IEEE Communications Magazine, vol. 47, no. 1, pp. 104-111, Jan. 2009.

The IEEE 802.16j standard seems to be the first commercial cellular standard to fully embrace relay technology. While IEEE 802.16j has not seen widespread commercial success, it is a good starting point for studying how relays might be included in a cellular communication protocol. This paper reviews key features of the standard and highlights interesting design features.


Current and past sponsors of this work include Huawei Technologies.

Comments are closed.