Kok Ann Donny Teo

Pilot tone design for peak-to-average power ratio reduction in OFDM

In orthogonal frequency division multiplexing (OFDM), the composite time signal exhibits a high peak-to-average power ratio (PAPR). Due to non-linearities of the transmit power amplifiers, this high PAPR generates spectral spreading, in-band distortion and out of band noise (OBN) which degrades the bit-error rate (BER). In this paper, we propose a way to combat this problem without sacrificing channel estimation and frequency-offset tracking accuracy, by designing a sub-optimal configuration of the pilot tones. The effectiveness of the proposed method to reduce PAPR and improve BER is validated by computer simulations based on the conditions in the IEEE 802.11a standard.

Subcarrier Allocation for multi-user OFDM system

Orthogonal frequency division multiple access (OFDMA) allows multiple users to make use of the same bandwidth as the single-user OFDM for data transmission and is a promising candidate to be used for future cellular systems. A key issue at hand is the rate-adaptive resource allocation problem. In this paper, we propose two subcarrier allocation schemes with low complexities based on the magnitude of the channel frequency responses. The proposed methods ensure a fair resource allocation in terms of the number of subcarriers with affordable bit-rates. Monte-Carlo simulations show that our proposed methods achieve a higher overall bit-rate for all users compared to previously derived subcarrier allocation schemes

Performance of single-carrier block transmissions over multipath fading channels with MMSE equalization

We study uncoded bit-error-rate (BER) performances of single-carrier block transmissions, zero-padded (ZP), and cyclic-prefixed (CP) transmission, when linear equalizers are applied and the BERs are averaged over one block. We show analytically that the BER of ZP transmission with linear equalization degrades as the bandwidth efficiency increases, i.e., there is a tradeoff between BER and bandwidth efficiency in ZP transmission. It is also proven that when minimum mean-squared-error (MMSE) equalization is adopted, ZP transmission outperforms CP transmission and uncoded orthogonal frequency-division multiplexing (OFDM) on the average over random channels. However, the difference between the ZP and the CP transmission becomes smaller as the block size gets larger, since the average BER performance of the ZP transmission degrades, while the average BER performance of CP transmission improves, as a function of the block size. Numerical examples are provided to validate our theoretical findings and to compare the block transmission systems

Kalman channel estimation based on oversampled polynomial model for OFDM over doubly-selective channels

To take intercarrier interference (ICI) attributed to time variations of the channel into consideration, we propose to parameterize the doubly-selective channel by the oversampled polynomial model (OPM). The OPM outperforms the over-sampled basis expansion model (OBEM). On top of that, we also propose to incorporate the effective ICI suppressing, reduced-complexity Viterbi-type maximum likelihood (RML) equalizer into Kalman filter for recursive channel tracking and equalization. An enhancement in the channel tracking ability and a significant improvement in BER performance are validated by numerical simulations.

Adaptive Subcarrier Allocation for Multi-User OFDM System*This paper was presented in part at IEEE Asia-Pacific Conference on Comunications (IEEE APCC) in Perth, October 2005.

Orthogonal Frequency Division Multiple Access (OFDMA) allows multiple users to make use of the same bandwidth as the single-user OFDM for data transmission and is a promising candidate to be used for future cellular systems. A key issue at hand is the rate-adaptive resource allocation problem. In this paper, we propose two basic subcarrier allocation schemes with low complexities based on the magnitude of the channel frequency responses. The proposed algorithms ensure a fair resource allocation in terms of the number of subcarriers with affordable bit-rates. Through extensive discussions and Monte Carlo simulations, a comprehensive comparison with previously derived subcarrier allocation schemes is performed which depicts the pros and cons of our proposed algorithms.

Approximate ber expression of ML equalizer for OFDM over doubly selective channels

Maximum Likelihood (ML) equalization for Orthogonal Frequency Division Multiplexing (OFDM) over time- and frequency- selective channels is analyzed in this paper. An approximate expression for bit error rate (BER) performance of the ML equalizer with a limited number of taps is developed, which subsumes the matched filter bound (MFB) equalization performance as a special case, and which saves on intensive time-consuming empirical simulations. Numerical simulations validate our approximate expression.

Universal BER Performance Ordering of MIMO Systems over Flat Channels

We present universal BER performance ordering for different antenna sizes in Multiple-Input Multiple-Output (MIMO) wireless systems with zero-forcing (ZF) equalization, which hold for all SNR. We first show that when the number of transmit antennas is fixed, BER of each symbol degrades with a decrease in the number of receive antennas even if the received SNR is kept constant. Then, we prove that when the number of receive antennas is fixed, the average BER improves with a decrease in the number of transmit antennas, which shows the tradeoff between BER and bandwidth efficiency in MIMO with ZF equalization. Furthermore, discussions on the relation of BER with the same change in the number of receive and transmit antennas show that there is no universal order in BER for any particular channel. These highlight the advantage and the limit of MIMO with ZF equalization.

SPC02-5: Pilot-Aided Channel Estimation and Viterbi Equalization for OFDM over Doubly-Selective Channel

In Orthogonal Frequency Division Multiplexing (OFDM) transmissions over doubly-selective channels, pilot-aided channel estimation is a practical approach to obtain channel estimates. In this paper, we apply Bessel model for channel modeling based on which pilot-aided channel estimation is performed. The utilization of Bessel model enables the estimation of intercarrier interference (ICI) effect of the neighboring subcarriers. A low complexity Viterbi-type Maximum Likelihood (ML) equalizer then uses the Bessel model-based channel estimates for equalization, displaying good bit-error rate (BER) performance. Monte-Carlo simulations show that the combination of Viterbi-type ML equalization and Bessel model-based estimates yields improved BER performance over other combinations of linear equalization and finite parameter model-based estimates.

Simple but Efficient Antenna Selection for MISO-OFDM Systems

Simple but efficient antenna selection schemes are proposed for the downlink of Orthogonal Frequency Division Multiplexing (OFDM) transmission with multiple transmit antennas over frequency selective fading channels, where transmit antennas are selected at the mobile terminal and the base station is informed of the selected antennas through feedback channel. To obtain the optimal antenna selection, channel frequency responses are required and performances have to be evaluated at all the subcarriers. To reduce the computational complexity at mobile terminal, time-domain channels are utilized for antenna selection in place of channel frequency responses. Our scheme does not guarantee the optimal antenna selection but is shown by numerical simulations to yield reasonable selections. Moreover, by using a specially designed pilot OFDM preamble, an antenna selection without channel estimation is developed. Efficiencies of our suboptimal antenna selections with less computational complexities are verified by numerical simulations.

Pilot Tone Design with Low Peak-to-Average Power Ratio in OFDM

In Orthogonal Frequency Division Multiplexing (OFDM), the composite time signal exhibits a high peak-to-average power ratio (PAPR). Due to non-linearities of the transmit power amplifiers, this high PAPR generates in-band distortion, out of band noise (OBN) or spectral spreading, which degrades the bit-error rate (BER) performance. In this paper, we propose a simple way to combat this problem without sacrificing channel estimation and frequency-offset tracking accuracy, by designing a sub-optimal configuration of the pilot tones. The effectiveness of the newly designed pilot tones in reducing PAPR is validated by Monte-Carlo simulations. The corresponding improvement in BER is also verified by simulations under IEEE 802.11a standard settings, by using the channel with perfect CSI and the designed pilot-aided estimated channel for coherent detection.