Bangwon Seo

Relay Precoding for Non-Regenerative MIMO Relay Systems with Partial CSI Feedback

We consider a relay precoding problem in a non-regenerative multiple-input multiple-output (MIMO) relay system, where the relay has the perfect channel state information of the source-relay link and only the covariance information of the relay-destination link. The average capacity seen by the relay is maximized under a relay transmit power constraint. The relay precoding problem is transformed to its equivalent power allocation problem at the relay. The optimal power allocation solution is then derived by formulating the optimization problem via difference of convex (DC) functions programming. To reduce the computational complexity, we also propose a suboptimal scheme by modifying the naive Jensen method. Furthermore, we propose the maximum eigenmode relaying (MER) scheme, which allocates the relay transmit power only to the maximum eigenmode. A necessary and sufficient condition for the MER to be asymptotically optimal is derived in the high signal-to-noise ratio (SNR) regime of the source-relay link. The performance of the proposed schemes is investigated through extensive numerical simulations.

Fast Convergent LMS Adaptive Receiver for MC-CDMA Systems with Space-Time Block Coding

For multicarrier code-division multiple access (MC-CDMA) systems employing Alamoutis simple space-time block coding (STBC), a special relationship between two minimum mean-square error (MMSE) filter weight vectors is derived in which one filter is for detecting odd-indexed symbols and the other for even-indexed symbols. Using this relationship, a modified MSE cost function is presented and an improved least mean-square (LMS) adaptive receiver is proposed. Since the number of filter taps independently updated in the proposed adaptive receiver is only half that of the conventional LMS adaptive receiver, the proposed adaptive receiver has a faster convergence rate than the conventional LMS adaptive receiver while keeping the same steady-state MSE performance. This is verified through the simulation.

SINR Lower Bound Based Multiuser Detector for Uplink MC-CDMA Systems with Residual Frequency Offset

For uplink multi-carrier code-division multiple access (MC-CDMA) systems, we propose a multiuser detector that is robust to a small residual frequency offset existing after frequency offset estimation and compensation. In this paper, when the residual frequency offset is normalized by subcarrier spacing, it is called a normalized residual frequency offset (NRFO). In the proposed scheme, we first derive a lower bound of the signal-to-interference plus noise ratio (SINR) of the desired user when the NRFO of the desired user is bounded by a small value and the value is known to the receiver. We then design a detection filter to maximize the SINR lower bound. Simulation results show that the proposed scheme has better SINR and bit error rate (BER) performances in a high signal-to-noise ratio (SNR) region than a conventional minimum mean square error (MMSE) receiver that ignores the NRFO.

Improved Precoding-Assisted Blind Channel Estimation Scheme for OFDM Systems

In this paper, we proposes an improved preceding- assisted blind channel estimation scheme for single-input single- output (SISO) orthogonal frequency-division multiplexing (OFDM) systems. In the proposed scheme, a modified autocorrelation matrix is obtained by dividing the autocorrelation matrix of the received signal vector by the precoding matrix element-wise. The channel vector is estimated as a principal eigenvector of the modified autocorrelation matrix, and it is simply obtained by the computationally simple power method without the need of the complicated eigenvalue decomposition. Simulation results show that the normalized mean square error of the proposed scheme is much lower than that of the conventional preceding-assisted blind channel estimation scheme.

Joint design of precoder and receiver in cognitive radio networks using an MSE criterion

We consider a single-user multiple-input multiple output (MIMO) transmission from secondary base station (SBS) in cognitive radio network where there is one primary user to which spectrum is licensed. It is assumed that coexistence of secondary network is allowed if interference power from the secondary network measured at primary user (PU) can be kept below a given threshold. In this paper, joint design problem of precoder and receiver is treated using mean-squared error (MSE) criterion between transmitted signal and its estimate at the receiver. Due to constraint on interference power measured at the PU, it is not possible to find an optimal solution in closed form. Therefore, a suboptimal solution with low complexity is proposed in closed form. Simulation results show that the MSE and bit error rate (BER) performances of the proposed scheme are 1.5-4.0dB better than those of a conventional capacity-maximization scheme when the received signal-to-noise ratio (SNR) is high, and they converge to their lower bounds, i.e., those of an MSE-based scheme without considering the PU interference constraint, as the received SNR increases.

Fast communication: Precoder design in cognitive radio networks with channel covariance information

In cognitive radio communications, the coexistence of a secondary network is allowed as long as the interference power from the secondary network measured at the primary user (PU) is kept below a given threshold. In this paper, we consider a communication scenario for cognitive radio networks where the channel state information (CSI) between the secondary base station (SBS) and the PU is perfectly known at the SBS transmitter whereas only the covariance information is known for the channel of the SBS-secondary user (SU) link at the SBS transmitter. A closed-form precoder design method is proposed for the purpose of maximizing the upper bound of the expected achievable rate of the SU. Even though the proposed scheme uses only partial CSI of the SBS-SU link, the simulation results show that the expected achievable rate of the proposed scheme approaches to that of a conventional scheme with full CSI as the channel correlation increases.

Non-redundant Precoding Based Blind Channel Estimation Scheme for OFDM Systems

For orthogonal frequency-division multiplexing (OFDM) systems, we propose a blind channel estimation scheme based on non-redundant precoding. In the proposed scheme, a modified covariance matrix is first obtained by dividing the covariance matrix of the received signal vector by the precoding matrix element-by-element. Then, the channel vector is estimated as an eigenvector corresponding to the largest eigenvalue of the modified covariance matrix. The eigenvector can be obtained by power method with low computational complexity instead of the complicated eigenvalue decomposition. We analytically derive a mean square error (MSE) of the proposed channel estimation scheme and show that the analysis result coincides well with the simulation result. Also, simulation results show that the proposed scheme has better MSE and bit error rate (BER) performance than conventional channel estimation schemes.

Improved Blind Channel Estimation for OFDM Systems

In this paper, we propose an improved blind channel estimation scheme for single-input single-output (SISO) orthogonal frequency-division multiplexing (OFDM) systems. In the proposed scheme, a modified cross-correlation matrix is obtained by dividing, element-wise, the cross-correlation matrix by the preceding matrix and then the channel vector is estimated by applying the principal eigenvector method to the modified cross-correlation matrix. The principal eigenvector is obtained by the power method with low complexity instead of the complicated eigenvalue decomposition. Simulation results show that the channel estimation and bit error rate (BER) performance of the proposed scheme is superior to that of the conventional preceding-assisted blind channel estimation scheme.

Simple Blind Channel Estimation Scheme for Downlink MC-CDMA Systems

In multicarrier code-division multiple access (MC-CDMA) systems, conventional blind channel estimation schemes require the inverse matrix calculation or eigenvalue decomposition of the received signal covariance matrix. Therefore, computational complexity of the conventional schemes is too high and they cannot be employed in downlink systems. In this paper, we propose a simple blind channel estimation scheme with very low computational complexity. Simulation results show that the proposed scheme has better channel estimation and bit error rate (BER) performance than the conventional schemes.

Blind Channel Estimation for Improper OFDM Systems with Linear Precoding

In this paper, we propose a non-redundant preceding-assisted blind channel estimation scheme for the improper orthogonal frequency-division multiplexing (OFDM) systems. In the proposed scheme, a modified cross-correlation matrix of the augmented received signal is obtained by dividing the cross- correlation matrix of the even-indexed and odd-indexed OFDM received signal by the precoding matrix element-wise. The channel coefficient vector is estimated as a principal eigenvector of the modified cross-correlation matrix, and it is obtained by the computationally simple power method without the need of the complicate eigenvalue decomposition. Simulation results show that the channel estimation and bit error rate (BER) performance of the proposed scheme is superior to that of the conventional preceding-assisted blind channel estimation scheme.