Myeong Jin Kim

Transceiver design based on interference alignment for MIMO interfering broadcast channels

In this paper, we propose a non-iterative transceiver design based on interference alignment (IA) for MIMO interfering broadcast channels with arbitrary number of cells and users per cell. We consider the case where a base station has more antennas than a user, which is a typical scenario of downlink transmission in the cellular network. By aligning the interference from other cells efficiently, we show that the proposed transceiver design can achieve the degrees of freedom (DOF) of d per user, i.e., total DOF of Σi=1C Kid, where C is the number of cells, Ki is the number of users for the i-th cell, and d is the number of streams per user. Accordingly, we investigate conditions for the antenna configuration to attain the total DOF of Σi=1CKid. We also analyze the computational complexity and achievable DOF of the proposed transceiver design. Finally, simulation results are provided to assess the sum-rate performance and achievable DOF of the proposed transceiver design, which validates the advantage of our design over existing schemes.

Limited Feedback Design for Interference Alignment on Two-Cell Interfering MIMO-MAC

This paper considers limited feedback for a two-cell interfering multiple-input-multiple-output multiple-access channel (MIMO-MAC). We propose a QR decomposition-based interference alignment (QRD-IA) scheme for efficient limited feedback and investigate the conditions for the antenna configuration. We characterize the interference leakage caused by quantization error as a function of the number of feedback bits. To minimize the interference leakage, we further propose a quantized-precoder-based IA (QP-IA) scheme by modifying the proposed QRD-IA scheme. By characterizing the interference leakage of the QP-IA scheme, we investigate the number of feedback bits to maintain the degrees of freedom (DOFs) in the two-cell interfering MIMO-MAC and a feedback bit allocation algorithm to improve the sum rate. From simulation results, we demonstrate that our proposed schemes can provide better sum-rate performance than the existing scheme using the same number of feedback bits, and our feedback bit allocation algorithm provides a higher sum rate compared with equal feedback bit allocation.

Sequential transceiver design based on block diagonalization and interference alignment for full duplex multiuser MIMO system

In this paper, we propose a sequential transceiver design based on block diagonalization (BD) and interference alignment (IA), which provides a superior performance of the full duplex multiuser multiple-input multiple-output (FD MU-MIMO) system over the half duplex (HD) system The proposed sequential transceiver design first obtains the precoder for the downlink (DL) channel by adopting the BD scheme. Next, by using IA, we suppress the self interference that causes the performance degradation to the uplink (UL) channel. From simulation results, we demonstrate that our proposed sequential transceiver design offers better sum rate performance than the conventional HD system.

Robust Transceiver Design Based on Joint Signal and Interference Alignment for MIMO Interference Channels With Imperfect Channel Knowledge

In this letter, we propose a new transceiver design based on joint signal and interference alignment, which provides the robustness against imperfect channel knowledge in multiple-input multiple-output (MIMO) interference channels. Under the assumption of norm-bounded channel uncertainties, the proposed robust transceiver design minimizes the interference signals that leak outside the interference subspace and makes the desired signal fall into the orthogonal complement of the interference subspace. From simulation results, we demonstrate that our proposed robust transceiver design offers not only better sum rate performance but also higher robustness against imperfect channel knowledge than the conventional scheme.

Novel node pair selection algorithms for sum-rate enhancement in MIMO interference channel

In multiple-input multiple-output(MIMO) interference channel, multiple transmitters and receivers communicate simultaneously as a pair. We establish the metric for node pair selection algorithm to maximize the sum rate in MIMO interference channel. In this paper, by adjusting node pairs properly, we show that the sum rate can be enhanced. We propose two different node pair selection algorithms to improve the sum rate of MIMO interference channel. Firstly, we propose an optimum node pair selection algorithm by formulating a node pair selection metric to select the best node pairs among all possible node pairs. Then we propose a sub-optimum node pair selection algorithm to lower computational complexity. We confirm from some numerical simulations that the proposed algorithms offer significant sum rate enhancement compared to the case when node pair selection is not considered.

Resource allocation for spatial-frequency domain based interference alignment scheme in MIMO-OFDM systems

Interference alignment (IA) needs multiple dimensions, which can be achieved by extending time, frequency or spatial domain. However, using time or frequency domain for alignment requires an exponential number of resources as the degrees of freedom increases. In addition, it is impractical to employ many number of antennas in the spatial-domain-based interference alignment system. In this paper, we consider MIMO-OFDM systems over K-user interference channels and present resource allocation criteria based on the IA scheme utilizing spatial and frequency domain simultaneously. Then, we propose a low complexity resource allocation scheme appropriate to both spatial and frequency domain IA. We show that a single additional subcarrier in the proposed resource allocation scheme is enough to improve the sum-rate performance close to the optimal performance.

Filter design of interference alignment scheme with limited feedback for two-cell interfering MIMO-MAC

In this paper, considering a two-cell interfering MIMO multiple access channel (MIMO-MAC), we propose an interference alignment scheme with an efficient limited feedback and investigate the conditions for the antenna configuration. In addition, we characterize the interference leakage power caused by quantization error as a function of feedback bits. From simulation results, we demonstrate that the proposed scheme can provide the better sum rate performance than the existing scheme under the same number of feedback bits.

Node pair selection schemes in MIMO interference channel using interference alignment

In this paper, we propose two different node pair selection schemes to enhance the system performances such as the sum rate and the outage probability in multiple-input multiple-output (MIMO) interference channel where multiple transmitters and multiple receivers communicate simultaneously as a pair. Firstly, we propose the optimum node pair selection scheme in terms of the outage probability. Then the max-min node pair selection scheme is proposed to improve both the sum rate and the outage probability simultaneously. From some numerical results, we confirm that the proposed schemes provide significant performance enhancement compared to the system when node pairs are fixed.

IA-based hybrid beamforming design in MIMO interference channel

In this paper, we develop a hybrid beamforming design over multiple input multiple output (MIMO) interference channels. Since the performance of MIMO interference channels depends on the desired channel and the leakage of interference signals, the proposed hybrid beamforming design first selects radio frequency (RF) precoders/combiners and optimizes the corresponding baseband precoders/combiners to maximize the rate over the desired channel by adopting the concept of orthogonal matching pursuit (OMP). Then, by using interference alignment (IA), we modify the baseband precoders/combiners to suppress the interference. From simulation results, we demonstrate that the proposed hybrid beamforming design offers the better sum rate than the baseline scheme in signal to noise ratio (SNR) region of interest where the interference is the dominant factor to the performance while preserving the degrees of freedom (DOF).

Interference Alignment Scheme Based on Limited Feedback for Two-Cell Interfering MIMO-MAC

In this paper, we propose an interference alignment (IA) scheme based on limited feedback for the two-cell interfering multiple input multiple output multiple access channel (MIMOMAC), which obtains the precoding vectors in the codebook directly so that the inter-cell interference (ICI) channels are most aligned. With the quantized precoding vectors, the receive beamforming filters are designed to cancel the impact of interuser interference (IUI) and reduce the residual ICI. In addition, we characterize the complexity of our proposed IA scheme. From simulation results, we demonstrate that the proposed scheme can provide the better sum rate performance under the same system configurations than the existing schemes which mitigate the interference leakage by aligning the ICI channels under limited feedback.