Yu Ted Su

Model-based channel estimation for OFDM signals in Rayleigh fading

This paper proposes a robust pilot-assisted channel estimation method for orthogonal frequency division multiplexing (OFDM) signals in Rayleigh fading. Our estimation method is based on nonlinear regression channel models. Unlike the linear minimum mean-squared error (LMMSE) channel estimate, the method proposed does not have to know or estimate channel statistics like the channel correlation matrix and the average signal-to-noise ratio (SNR) per bit. Numerical results indicate that the performance of the proposed channel estimator is very close to the theoretical bit error propagation lower bound that is obtained by a receiver with perfect channel response information.

Performance analysis of equalized OFDM systems in Rayleigh fading

Channel estimation is usually needed to compensate for the amplitude and phase distortions associated with a received orthogonal frequency-division multiplexing (OFDM) waveform. This paper presents a systematic approach for analyzing the bit-error probability (BEP) of equalized OFDM signals in Rayleigh fading. Closed-form expressions for BEP performance of various signal constellations [phase-shift keying (PSK), differential phase-shift keying (DPSK), quaternary phase-shift keying (QPSK)] are provided for receivers that use a linear pilot-assisted channel estimate. We also derive the optimal linear channel estimates that yield the minimum BEP and show that some previous known results are special cases of our general formulae. The results obtained here can be applied to evaluate the performance of equalized single-carrier narrowband systems as well.

Blind and semiblind detections of OFDM signals in fading channels

This paper considers the problem of blind joint channel estimation and data detection for orthogonal frequency-division multiplexing (OFDM) systems in a fading environment. Employing a regression model for a time-varying channel, we convert the problem into one that finds the data sequence x whose associated least-squares (LS) channel estimate z(x) is closest to the space of some regression curves (surfaces). We apply the branch-and-bound principle to solve the nonlinear integer programming problem associated with finding the curve that fits a subchannel in the LS sense. A recursive formula for fast metric update is obtained by exploiting the intrinsic characteristic of our objective function. The impacts of reordering the data sequence and selective detection are addressed. By employing a preferred order along with a selective detection method, we greatly reduce the detector complexity while giving up little performance loss. Both the complete and the reduced-complexity algorithms can be used for blind and semiblind detections of OFDM signals in a subchannel-by-subchannel manner. To further reduce the complexity and exploit the frequency-domain channel correlation, we suggest a two-stage approach that detects a few selected positions in some subchannels first, and then, treating the detected symbols as pilots, determines the remaining symbols within a properly chosen time-frequency block by a two-dimensional model-based pilot-assisted algorithm. The proposed methods do not require the information of the channel statistics like signal-to-noise ratio or channel correlation function. Performance of differential modulations like differential quaternary phase-shift keying and STAR 16-ary quadrature amplitude modulation are provided. Both blind and semiblind schemes yield satisfactory performance.

MIMO channel estimation in correlated fading environments

This paper presents two analytic correlated multiple-input multiple-output (MIMO) block fading channel models and their time-variant extensions that encompass the popular Kronecker model and the more general Weichselberger model as special cases. Both static and time-variant models offer compact representations of spatial- and/or time-correlated channels. When the transmit antenna array is such that the associated MIMO channel has a small angle spread (AS), which occurs quite often in a cellular downlink, our models admit reduced-rank channel representations. They also provide compact channel state information (CSI) descriptions which are needed in feedback systems and in many post channel estimation applications. The latter has the important implication of reduced feedback channel bandwidth requirement and lower post-processing complexity. Based on one of the proposed channel models we present novel iterative algorithms for estimating static and time-variant MIMO channels. The proposed models make it natural to decompose each iteration of our algorithms into two successive stages that are responsible for estimating the correlation coefficients and the signal direction, respectively. Using popular industry-approved standard channel models, we verify through simulations that our algorithms yield good MSE performance which, in many practical cases, is better than that achievable by a conventional least-square estimator. The mean-squared error (MSE) performance of our estimators are analyzed and the resulting predictions are consistent with those estimated by simulations.

Multiuser detection for MFSK frequency-hopped multiple access system

The capacity of a multilevel FSK (MFSK), frequency-hopped multiple access (FHMA) system is dictated by two major design concerns: the hopping pattern and the receiver structure. This paper studies the impact of both factors. We present a maximum likelihood (ML) diversity combiner for asynchronous FHMA systems in Rayleigh fading and compare systems using random hopping patterns and those that use a set of optimal hopping patterns. On the other hand, a multiuser detector for an FHMA/MFSK system usually consists of two stages. The first stage makes an initial decision while the second stage tries to resolve the ambiguity and reduce the cochannel interference (CCI) left by the first stage detector. After analyze the performance of the first stage detector that uses appropriate diversity combining, we then examine the effectiveness of a second stage detector which is a modification of an earlier structure developed by Timor (1980). This receiver is of modest complexity and is capable of removing most of the CCI. Moreover, we also find that it is insensitive to power variations of the system users.

New Signal Designs for Enhanced Spatial Modulation

In this paper, we present three new signal designs for enhanced spatial modulation (ESM), which was recently introduced by the present authors. The basic idea of ESM is to convey information bits not only by the index(es) of the active transmit antenna(s) as in conventional SM, but also by the types of the signal constellations used. The original ESM schemes were designed with reference to single-stream SM and involved one or more secondary modulations in addition to the primary modulation. Compared with single-stream SM, they provided either higher throughput or improved signal-to-noise ratio (SNR). In this paper, we focus on multi-stream SM (MSM) and present three new ESM designs leading to increasing SNR gains when they are operated at the same spectral efficiency. The secondary signal constellations used in the first two designs are derived through a single geometric interpolation step in the signal constellation plane, while the third design also makes use of additional constellations derived through a second interpolation step. The new ESM signal designs are described for MIMO systems with four transmit antennas out of which two are active, but we also briefly present extensions to higher numbers of antennas. Theoretical analysis and simulation results indicate that the proposed designs provide a significant SNR gain over MSM.

Enhanced spatial modulation with multiple constellations

Spatial modulation (SM) has recently drawn a great deal of attention, particularly due to the low complexity that it promises for both the transmitter and the receiver sides. However, this technique has a significant spectral efficiency loss with respect to spatial multiplexing (SMX) with the same number of transmit (Tx) antennas, and when the modulation order is increased to achieve the same spectral efficiency, it loses in terms of the bit error rate (BER). In this paper, a new type of SM (referred to as Enhanced SM) is proposed which increases the number of bits transmitted per channel use compared to conventional SM. Note that conventional MIσMO techniques including SMX and SM employ a fixed signal constellation. In our proposed technique, some information bits select not only the index(es) of the active antenna(s), but also the constellations to be transmitted from each of them. Both the closed-form analysis and the numerical results demonstrate that the proposed technique achieves better performance than conventional SM and that in most cases it also outperforms SMX.

MIMO channel estimation in spatially correlated environments

We present in this paper a novel model-based scheme for estimating MIMO block fading channels. The proposed method can be applied to estimate the conventional independent and identically distributed (i.i.d.) complex Gaussian as well as the more realistic spatial-correlated channels. Excellent mean squared identification error (MSIE) performance can be attained with proper designed system parameters. We also derive an adaptive least squares (RLS) version that can be used for estimating a static channel or tracking time-varying channels. Furthermore, when the angle spread (AS) at the transmit site is small and thus the correlations amongst the transmit antennas are high, we show that it is possible to extract the information about the transmitters angle of departure (AOD) for potential postprocessing applications.

Modelling and Estimation of Correlated MIMO-OFDM Fading Channels

This paper presents a general reduced-rank channel model and a corresponding low-complexity estimation scheme for wideband spatial-correlated multiple-input multiple-output (MIMO) systems. We focus on orthogonal frequency division multiplex (OFDM) based systems. The proposed reduced-rank model is useful for many post-channel-estimation applications such as channel state information (CSI) feedback, precoder design and user/channel selection. Our work is an extension of an earlier investigation on narrowband MIMO channels. Like the narrowband case, the proposed wide channel estimator also offer the advantage of rendering both channel coefficients and mean angle of departure (AoD) simultaneously. BY exploiting the time, frequency, and spatial correlations of the channel and with continuous-type pilot symbols, we found that, even with as high as a compression ratio of 1\%, our channel estimator is capable of maintaining an acceptable mean squared errors (MSE) in highly correlated environments. Both mathematical analysis and computer simulation, based on some industry-approved standard channel models, indicate that our algorithm outperform the conventional least-square estimator within most range of interest.

The effects of Rician fading and multiple CCI on differentially detected GMSK signals

The paper presents the bit error rate (BER) analysis of a GMSK system employing various forms of differential detection. The desired signal is assumed to experience Rician fading while multiple co-channel interference (CCI) suffer from Rayleigh fading. The effects of fading rate (fast or slow fading) and time and frequency dispersions are also examined.