Heon Huh

A Unified Approach to Optimum Frame Synchronization

In this paper, we present a unified approach for optimum frame synchronization where the observed symbols are modeled as the output of a Markov chain corrupted by additive white Gaussian noise (AWGN). This model encompasses many different frame synchronization situations e.g., convolutionally coded transmissions and nonlinear modulations with memory, such as continuous-phase modulation. The proposed frame synchronizer is implemented with trellis path search used for Markov chain decoding and provides better performance than conventional synchronizers because it is able to exploit coding gain. The proposed synchronizer includes many previously proposed synchronizers either as special cases or as suboptimal approximations. The numerical results show the performance gain of the proposed frame synchronizer for convolutional codes with linear modulation, transmission over known intersymbol-interference channels, and nonlinear continuous-phase modulation signals. The proposed frame synchronizer is derived for the case of continuous transmission but may also be applied to certain packet transmission scenarios

Comparison of error probability for OFDM and SC-FDE

The purpose of this paper is to quantify the effects of frequency and timing offsets in orthogonal frequency division multiplexing (OFDM) and single carrier systems and to compare their relative sensitivities. This is done in two ways. First, by simulation and second, by a numerical computation using a method first introduced by Beaulieu. We show that the standard approach, which is based on approximating the ICI (or ISI) term as a Gaussian random variable, is inadequate to characterize performance degradation. Our results show that the Gaussian approximation can be overly optimistic or overly pessimistic, depending upon the case. In contrast the numerical method is shown to agree closely with simulation in all cases. While the numerical method used here has been previously employed for calculation with OFDM, it has not been used in the comparison of single carrier and OFDM.

Non-coherent demodulation for orthogonal space-time coded CPM

In this paper non-coherent receivers for orthogonal space-time coded continuous phase modulated signals transmitted over additive white Gaussian noise and quasi-static fading channels are presented. The receivers are found to perform less than 1-dB away from the corresponding coherent detectors.

Sensing and Signal Processing for a Distributed Pavement Monitoring System

Estimation of road profile parameters and detection of road pavement failures using in-vehicle sensors such as vertical accelerometers or position sensors is of immense value in a futuristic distributed highway monitoring system. In this paper, a novel road elevation profile model, well suited for analysis using statistical signal processing tools, is developed to characterize road features. We then apply maximum likelihood (ML) and minimum mean square error (MMSE) estimation techniques to estimate model parameters using in-vehicle vertical acceleration and displacement measurements. In addition, we show that the estimates of the road model parameters can be used to detect sudden disturbances in the road or pavement failures

Non-Coherent Receiver Performance for Orthogonal Space-Time Coded CPM in Fading Channels

In mobile communication systems where bandwidth and power efficiency are at a premium, continuous phase modulation (CPM) is a viable modulation technique. In this paper, non-coherent receivers for orthogonal space-time coded continuous phase modulated signals transmitted over two antennas are derived using a maximum likelihood (ML) approach. The performance of these receivers is numerically characterized in Rayleigh fading channels

Frame Synchronization of Coded Modulations in Time-Varying Channels via Per-Survivor Processing

In this letter, an optimum frame synchronizer is proposed for coded modulations in channels with uncertainties. Coded modulations include various frame synchronization scenarios, e.g., convolutionally coded transmissions and nonlinear modulations with memory. Frame synchronization is proposed as a maximum a posteriori probability estimation implemented using trellis path search for Markov chain decoding. In addition, time-varying uncertainties such as frequency offset and phase noise are jointly estimated via per-survivor processing as frame synchronization proceeds. The proposed frame synchronizer exploits the coding gain of coded modulations to achieve better performance than conventional frame synchronizers. We show that the resulting frame synchronizer consists of a correlation term and two data correction terms. Numerical results show that the proposed frame synchronizer is robust to uncertainties at the receiver and it exhibits improved performance.

Suboptimal symbol-by-symbol demodulation of continuous phase modulated signals using the Laurent decomposition

Symbol-by-symbol detection is often used in systems with interleaved coded modulation in order to provide the needed soft-decision metrics for the decoding of the outer code. The paper develops a suboptimal symbol-by-symbol detection method for the demodulation of continuous phase modulated (CPM) signals. The proposed algorithm offers a reduced-complexity symbol-by-symbol detection in which the numbers of matched filters and trellis states are appreciably reduced. The complexity reduction is achieved by approximating the CPM signal using the Laurent representation. A simulation study shows that the performance degradation relative to the optimal full complexity algorithm is small for a variety of CPM signals.

Efficient Timing and Frequency Offset Estimation in OFDM Systems

Orthogonal frequency division multiplexing (OFDM) is a bandwidth efficient modulation scheme used widely for high speed data communication. Fast and ro- bust synchronization algorithms are necessary for OFDM systems operating in burst mode like WLANs. In this paper, we propose an efficient timing and carrier frequency synchronization method for OFDM. The proposed scheme needs only one OFDM symbol for both symbol/frame timing and frequency offset estimation, while previous synchronization algorithms need more than two OFDM symbols as a training sequence. Numerical analysis shows that the proposed frequency estimator nearly achieves the Cramer-Rao lower bound for the variance of the frequency offset estimate, despite the reduction in the training sequence length.

On training-symbol design and efficient synchronization in OFDM systems

We show the equivalence between the conventional frame synchronization in single-carrier systems and integer part estimation of frequency offset in OFDM systems and propose an efficient synchronization scheme. The proposed scheme achieves both OFDM symbol/frame timing and frequency offset estimation with only one well-designed OFDM training symbol, while previous synchronization algorithms need two OFDM training symbols at least. Numerical analysis shows that the proposed frequency estimator nearly achieves the Cramer-Rao lower bound for the variance of the frequency offset estimate, despite the reduction in the training sequence length.

Decoder-assisted frame synchronization in the presence of phase/frequency noise

The optimum decoder-assisted frame synchronization is proposed for convolutionally coded modulation under the perfect channel state information. The proposed synchronization is extended to channels with phase instabilities by incorporating per-survivor processing. The phase noise is jointly estimated by per-survivor processing as frame synchronization proceeds. The proposed frame synchronizer is robust to phase noise at the receiver and shows performance improvement due to coding gain.