Efficient algorithms for throughput maximization on the MIMO interference channels

• Main Author: Anand, Kushal
• Other Authors: Erry Gunawan
• Format: Theses and Dissertations
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Wireless communication systems
• Online Access: http://hdl.handle.net/10356/51185
• Institution: Nanyang Technological University

This research focuses on the algorithm designs for throughput enhancement on the multiple-input-multiple-output interference channels (MIMO-IC) under the cases of both the perfect and imperfect channel state information at the transmitters (CSIT). The overall network sum rate is the measure of performance. We study various standard beamforming design algorithms in the literature for throughput enhancement on the distributed MIMO-IC and identify their strengths and weaknesses under different network scenarios. It is observed that most of these algorithms perform well in the symmetric networks where the desired and interfering signal strengths at the receivers are of equal strengths. However, these algorithms show significant degradation in the network sum rate in the asymmetric and the overloaded networks (when the number of users or data streams in the network exceeds a certain limit). Note that in a distributed network, few beamformer exchanges between the transmitters and the receivers should be considered for the beamforming design in order to reduce the system overhead but the existing algorithms require lot of iterative beamformer exchanges to converge to their locally optimal beamformer solution. We propose two beamforming methods which show higher throughput compared to the existing methods in the asymmetric and overloaded networks. Moreover, the proposed methods show faster convergence rate and thus, attain a higher sum rate with less overhead compared to other methods when we consider less number of iterative beamformer exchanges. We address the practical scenario of throughput enhancement on the MIMO-IC under the case of imperfect or quantized CSIT. In practice, each user has limited feedback budget for quantizing its channels. In this work, we derive the dynamic bit allocation policy for the MIMO-IC under limited feedback resources constraint at the receivers. The idea is to minimize the system rate loss metric by optimally allocating the feedback bits to the desired and the interfering links. The bit allocation policy depends on the channel strengths, SNR of the users and the network parameters such as the number of users and antennas in the network. We get some useful insights based on our analysis of the bit allocation policy. We find that under very limited feedback budget, most of the bits should be allocated to the desired links. However, as the feedback budget begins to increase, some bits should be allocated to the interfering links and at high feedback budget, most of the bits are allocated to the interfering links. Furthermore, for a fixed feedback budget, more bits are allocated to the interfering links at higher SNR. Based on the bit allocation policy, improved beamforming strategies are proposed both for the multiple-input-single-output interference channels (MISO-IC) and the MIMO-IC networks under the case of limited feedback resources at the users. In the proposed improved beamformer design, the transmitter exploits the channel quantization error statistics as decided by the bit allocation policy at the receivers. Thus, the improved beamformer design shows higher throughput compared to any existing schemes by making optimal use of the limited feedback resources at the users.