Noncoherent BCJR algorithm, in which multiple symbol windowing is used to recover deterioration caused by channel phase noise, is proposed for single input and single output (SISO) systems in AWGN channels. In this article, we first apply noncoherent BCJR algorithm to multiple input and multiple output (MIMO) systems and generalize the algorithm from scalar to matrix based. In addition, we advance its competency from AWGN channels to Rayleigh fading channels and formulate corresponding branch metric. Based on our theoretical derivation, we propose noncoherent and non-CSI iterative decoding for serial concatenated single differential space-time block coding (SDST) and develop two types of decoding techniques using rectangular windowing (RW) and exponential decaying windowing (EDW), respectively. The efficiency of our proposed schemes is validated and compared via simulations.

Weak atmospheric turbulence causes pulse broadening of an optical signal. In this article, we drive the pulse broadening as a function of turbulence strength by temporal moments directly. We derive pulse broadening dependency on turbulence strength and propagation distance when an optical signal propagates through atmosphere. This pulse broadening induces intersymbol interference (ISI) between adjacent pulses. Based on this derived results, we develop some observation patterns using a practical example.

We propose a noncoherent decoder-assisted (NDA) frame synchronizer, a two-stage, synchronization scheme based on the list-synchronization principle [P. Robertson, 1992], [M. Howlader and B. Woerner, 2001], for serially concatenated coded differential modulation (SCCD) system [K. R. Narayanan and G. L. Stuber, 1999] in packet transmission. For the first stage, we derive a suitable noncoherent maximum likelihood (SNML) rule, more effective for the noncoherent decoder's operating region of smaller channel phase offsets, to provide a reliable list of potential frame starting positions. For the second stage, we propose a hybrid sync method by exploiting the combined sign difference ratio (SDR) and sync word log-likelihood ratio (SW LLR) information [M. Howlader and B. Woerner, 2001] to obtain more reliable final decisions and reduce computational complexity. Performance and complexity of the proposed scheme are evaluated under constant and time-varying phase offsets in the AWGN channel via Monte Carlo simulation.

This paper proposes a noncoherent decoder-assisted (NDA) frame synchronization methodology for OFDM packets using the list-synchronization technique [MMK Howlander et al., 2001] for serially concatenated coded differential modulation (SCCD) systems [KR Narayanan et al, 1999], where a conventional MMSE rule based on the supplementary guard interval [M Speth et al., 1998] synchronizer in the time domain generates a list of most probable packet starting positions, and the following synchronizer makes a refined decision. The second synchronizer is decoder-assisted, accepting soft-decision from the decoder in the frequency domain based on the combination of choosing and discarding technique. Preliminary simulation results of the proposed scheme are provided to demonstrate the methodology

In this paper, we propose a detection technique for a serially concatenated coded MSK (SC-MSK) modulated signal, based on the noncoherent BCJR (N-BCJR) algorithm, where an exponentially decaying observation window (EDW) in the N-BCJR algorithm is used to get better use of the channel phase memory and reduce the state as well as metric computational complexity. We also derive the decoding metrics of N-BCJR for the Rayleigh fading channel without channel state information (CSI) and incorporate the EDW in the metrics. The BER performance of the proposed noncoherent SC-MSK, using EDW and a rectangular window (RW), are compared via Monte Carlo simulation in both AWGN and slow, flat Rayleigh fading channels.

We propose a noncoherent, decoder-assisted (NDA) frame synchronizer, a two-stage synchronization scheme based on the list-synchronization principle, for the serially concatenated coded differential modulation (SCCD) system in the packet transmission. For the first stage, we derive a suitable noncoherent, maximum likelihood (SNML) rule, more effective for the noncoherent decoder's operating region of smaller channel phase changes, to provide a reliable list of potential frame starting positions. For the second stage, we propose a hybrid sync method by exploiting the combined sign difference ratio (SDR) and sync word Log-likelihood ratio (SW LLR) information to obtain more reliable final decisions and reduce computational complexity. Performance and complexity of the proposed scheme are evaluated under constant and time-varying phase changes in the AWGN channel via Monte Carlo simulation.

In this article, we propose iterative noncoherent (without knowledge of fading phase) and non-CSI (without knowledge of fading phase and amplitude) decoding for space-time block coding (STBC) by employing MSK modulation instead of commonly used PSK. Beyond a simple replacement, MSK modulated STBC (MSK-STBC) reduces the smallest resolvable time interval from STBC code duration, necessary for differential space-time coding (DST), to MSK symbol duration, and thus presents consistent and enhanced system performance in fast fading channels. The efficiency of our schemes is validated and illustrated via theoretical analysis and numerical simulations.

In this paper, a generic differential detection technique for DS-CDMA systems. Generalized fractional symbol differential detection (GFSD), is proposed and studied in Rayleigh fading channels. The symbol-level differential detection (SLD) and chip-level differential detection (CLD) are also discussed as special cases of GFSD. The decision variable of GFSD is derived as a function of processing gain and fractional factor. The average bit error rate (BER) performance is investigated in multipath time-varying channels via simulations with different parameters, such as signal to noise ratio (SNR), fading rate and number of paths. The results show that for any specific channel situation, there is an optimal fractional factor to obtain the best system performance.

In this article, we propose a new hybrid interference cancellation (HIC) receiver for multirate DS-CDMA system based on multiple processing gain (MPG). This improved HIC receiver divides the active users with different data rates into a number of groups, where each group consists of users with different data rates. First, the groupwise parallel interference cancellation (GW-PIC) is performed among the groups, and is followed by the successive interference cancellation (SIC) within each group. We derive the bit error probability (P/sub b/) of this HIC scheme. The P/sub b/, of the HIC receiver in AWGN channel is better than that of the adaptive multistage PIC or conventional SIC receiver.

Space-time trellis codes (STTC) have been proposed as the constituent codes of turbo codes (D Cui et al., 1999) (W Firmanto et al., 2001). In this article, we propose a similar scheme, the iteratively decoded space-time trellis codes (ISTTC), and derive the performance upper bound on BER. When the STTC are combined with space-time block codes (STTC-STBC), it provides an improved BER over the system with STTC only. Here, we also propose the iteratively decoded STTC-STBC (ISTTC-STBC) to improve both the performance and the rate. It can compensate the reduced transmission rate caused from the iteratively decoded STTC to combine the STBC having more transmitters than the constituent STTC. This paper presents the details of computing the extrinsic information shared between two constituent decoders, design schemes of ISTTC with or without STBC, and their performance results shown via simulation.

Differential space-time block coding provides an efficient decoding technique under unknown fading channels. However, the difficulties regarding blind decoding schemes for space-time trellis coding (STTC) remain unsolved. In this article, we propose to apply noncoherent BCJR algorithm to serial concatenated STTC systems, where an average or integration operation is employed to recover the deterioration caused by channel phase and amplitude. Two types of decoding algorithms are developed, referred to as noncoherent (without knowledge of channel phase) and nonCSI (without knowledge of channel state information (CSI)). The efficiency of our proposed schemes is validated and compared via simulations.

In this article, we propose a serial concatenated single differential space-time (SDST) coded OFDM systems, in which blind SDST coding is serially concatenated with convolutional coding. The OFDM is incorporated for maintaining satisfactory performance in both time and frequency selective fading channels at low SNR. The encoding and iterative decoding processes are thoroughly studied and the improved performance is demonstrated via simulations.

Adaptive orthogonal frequency division multiplexing (OFDM) has been shown to have significant improvement over basic OFDM in terms of spectral efficiency and bit error rate performance. However, the most important condition to achieve these benefits is the availability of accurate channel state information (CSI) at the transmitter. Imperfect CSI arises from noise at the receiver and due to a delay in providing the CSI from the receiver to the transmitter for the next data transmission. In this paper, we have studied the impact of imperfect CSI on the spectral efficiency of adaptive OFDM systems. We have proposed a robust loading algorithm to improve the performance of such systems. This algorithm applies the bit under loading phenomenon and makes efficient use of M-ary Phase Shift Keying (MPSK) modulation. Plots of spectral efficiency via simulations are provided.

Some previous research has demonstrated the feasibility of decoder-assisted frame synchronization and channel estimation as separate processes for coded systems. Here a joint decoder-assisted process of channel estimation and frame synchronization using the same sync word (SW) is proposed. In brief, the method entails using a SW embedded in the mid-packet for both channel estimation and frame synchronization. The method is based on a list-synchronization scheme, where a correlation-based frame synchronizer passes a list of possible delays to the turbo decoder. The turbo decoder makes the final decision about frame synchronization. The turbo decoder is draw upon pilot symbol assisted modulation (PSAM) techniques for channel estimation using the coded SW of frame synchronization. The performance of channel estimation technique using the SW is presented via simulation.

Higher transmission rate and reliable communication over fading channels are major research pursuits of current wireless communications. The space-time block codes (STBC) have been introduced to improve those research goals by simple processing algorithms at the transmitter and the receiver ends. STBC designs with the help of an orthogonal design method are described. The STBC designed in a quasi-orthogonal method is focused on keeping the higher transmission rate at the cost of the diversity gain. On the other hand, STBC designed in complex orthogonal methods emphasize the high diversity gain rather than the rate. Design schemes of turbo concatenation to STBC using various methods are introduced and detailed performance analyses are provided.

Channel estimation has been considered difficult for a time-varying Rayleigh fading channel. The statistics of fading channels are so variant that it is difficult to get an optimal channel model. Overall system complexity reduces without the channel estimation. Concatenating turbo codes to the transmit diversity achieves the coding gain from the concatenated turbo codes and the diversity gain from the transmit diversity. For normalized Doppler frequency, f/sub d/T/sub s/ up to 10/sup -4/ BER degrades by 3 dB for the transmit diversity without channel estimation over the transmit diversity with channel estimation. However, when the transmit diversity with no channel estimation is concatenated with turbo codes, the optimal performance can be obtained even above f/sub d/T/sub s/, 10/sup -4/.

Recently space-time coding (STC) has received increasing attention because of its ability to combat fading effects. Among existing schemes, double differential space-time coding is of special interest because its decoding process does not require channel state information. Furthermore, it is applicable to continuous fast time varying channels. Our article focuses on the improvement of double differential space-time coding by considering multiple symbol decoding. We develop a trellis-based decoding algorithm for multiple symbol double differential space-time coding, derive the branch metric, and validate the improvement via simulations. However, the performance of this scheme degrades considerably in frequency selective channels. Another focus of this article is to construct a suitable model in which OFDM can be concatenated with multiple symbol double differential space-time coding.

This paper proposes a novel frame synchronization methodology for packets with convolutionally encoded data. Rather than by placing sync bits in a separate header, the sync bits are placed in a midamble and encoded as part of the data sequence, using the error correction encoder to resolve time ambiguities. The scheme is based on the principle that an error in trellis termination may result in decoding a wrong information bit sequence, while the starting portion of the sequence can be decoded as error free. This frame synchronization technique is extended to the synchronization of turbo-coded systems using a list-based approach, where the first synchronizer provides a list of potential packet starting positions, and the following synchronizer makes the final decision. The performance improvement over conventional synchronization techniques is quantified via simulation.

This paper proposes a novel frame synchronization scheme for convolutionally encoded data packets. Rather than placing sync bits in a separate header, the sync bits are placed in a mid-amble and encoded as part of the data sequence, using the error correction encoder to resolve time ambiguities. This technique requires fewer bits for frame synchronisation. The performance improvement over conventional synchronization techniques is explored via simulation. The applicability of the scheme for synchronization of turbo codes is also discussed.