Optimal designs of collaborative relay-assisted multiuser beamforming for cellular systems

Da Wang, Chen Chen, Lin Bai, Jianhua He, Ye Jin, Ying Li

Research output: Contribution to journalArticlepeer-review


With careful calculation of signal forwarding weights, relay nodes can be used to work collaboratively to enhance downlink transmission performance by forming a virtual multiple-input multiple-output beamforming system. Although collaborative relay beamforming schemes for single user have been widely investigated for cellular systems in previous literatures, there are few studies on the relay beamforming for multiusers. In this paper, we study the collaborative downlink signal transmission with multiple amplify-and-forward relay nodes for multiusers in cellular systems. We propose two new algorithms to determine the beamforming weights with the same objective of minimizing power consumption of the relay nodes. In the first algorithm, we aim to guarantee the received signal-to-noise ratio at multiusers for the relay beamforming with orthogonal channels. We prove that the solution obtained by a semidefinite relaxation technology is optimal. In the second algorithm, we propose an iterative algorithm that jointly selects the base station antennas and optimizes the relay beamforming weights to reach the target signal-to-interference-and-noise ratio at multiusers with nonorthogonal channels. Numerical results validate our theoretical analysis and demonstrate that the proposed optimal schemes can effectively reduce the relay power consumption compared with several other beamforming approaches.
Original languageEnglish
Pages (from-to)1484–1499
Number of pages15
JournalWireless Communications and Mobile Computing
Issue number15
Early online date25 Sept 2012
Publication statusPublished - 25 Oct 2014


  • collaborative relay beamforming
  • multiuser system
  • semidefinite relaxation technology


Dive into the research topics of 'Optimal designs of collaborative relay-assisted multiuser beamforming for cellular systems'. Together they form a unique fingerprint.

Cite this