In this paper, an intercarrier interference (ICI) suppression method for frequency domain differential orthogonal frequency division multiplexing (OFDM) systems with multiple symbol differential detection (MSDD) in time-varying frequency-selective fading channels is proposed. The decision metric is derived by considering the statistical effects of ICI from all the subcarriers as well as considering the fading autocorrelations. Knowledges of transmitted signals from both inner and outer subcarriers of an MSDD observation interval can be used to estimate and mitigate the effects of ICI. The effectiveness of the proposed method is demonstrated by computer simulations.

In this paper, the diversity reception of differential space-time block coding (STBC) system using multiple-symbol differential detection (MSDD) method with multiple receive antennas is considered. Postdetection maximum ratio combining (MRC), postdetection equal-gain combining (EGC) and predetection selective combining (SC) are used as diversity combining techniques. Their performance is analyzed by computer simulations

In OFDM systems with differential phase shift keying (DPSK), it is possible to apply differential modulation in either the time or frequency domain, depending on fading channel conditions, such as Doppler frequency shift and delay spread. Meanwhile, postdetection diversity for DPSK is attractive because the cophasing function necessary in predetection combiners is not required. The paper proposes a simple calculation method, that is an approximate closed-form equation of the bit error rate (BER) in DPSK/OFDM systems with postdetection diversity over both time and frequency selective Rayleigh fading channels. The validity of the proposed method is demonstrated by the fact that the BER performances given by the derived equation coincide with those by Monte Carlo simulation.

We had previously proposed a fuzzy logic-based buffer management scheme (BMS) called fuzzy early detection (FED), designed to improve transmission control protocol (TCP) performance over the unspecified bit rate service of asynchronous transfer mode networks. Since a weakness in FED was discovered later, we present a refined version of it named FED+ here. Maintaining the design architecture and algorithm of FED, FED+ further adopts per virtual connection accounting to achieve its design goals. The effects of TCP implementation, TCP maximum segment size, switch buffer size and network propagation delay on FED+ performance are studied through simulation. Its performance is then compared with those of pure early packet discard, P-random early detection and FED. Our evaluations show that FED+ is superior to the others if the issues of efficiency, fairness, robustness, buffer requirement and the ease of tuning control parameters of a BMS are considered collectively.

We have proposed a fuzzy logic-based buffer management scheme (BMS) called fuzzy early detection (FED) to improve transmission control protocol (TCP) performance over the unspecified bit rate service of asynchronous transfer mode networks. FED uses early congestion detection and selective packet discard strategies to achieve its design goals, and adopts early packet discard (EPD) to prevent frequent buffer overflows. Its performance is compared with those of pure EPD and P-random early detection (P-RED) through various simulations. The effects of TCP implementation, TCP maximum segment size, switch buffer size and network propagation delay on their performances are studied. If the issues of efficiency, fairness, robustness, ease of tuning parameters, and buffer requirement of a BMS are considered collectively, we conclude that FED is better than the others.

Orthogonal frequency division multiplexing (OFDM) systems avoid the effect of multipath fading by absorbing delayed waves into guard interval. But as the effect of delayed waves over guard interval becomes stronger, it causes the bit error rate (BER) performance in OFDM systems to be more degraded. This paper proposes the approximate derivation method of the BER in DQPSK/OFDM systems over frequency selective fading channels. In this study, we will derive the approximate equation of the BER with differential coding/differential detection in the time domain. We confirm that the BER derived from the proposed approximate equation coincides with that by computer simulation even when such parameters as the Doppler frequency shift, the delay spread, the Rician factor and so on are varied.

The transmission quality in mobile wireless communications is affected not only by additive white Gaussian noise, but also by multipath fading, which drastically changes the amplitudes and phases of wireless signals. It is one of the key themes in direct-sequence code-division multiple-access systems to evaluate average bit error rate (BER) performances as parameters of the Doppler frequency, the delay profile, the number of simultaneous access users, and so on. This paper proposes the approximate equation for easily calculating the average BER over fading channels (multiray fading models). The performance evaluations are carried out in the two-ray model, the IMT2000 model, and the COST207 model. It is confirmed from the coincidence of approximate results with computer simulation results that the proposed approach is applicable to a variety of parameters.

In previous work we have presented a model of genetic algorithm (GA) that applies varying mutations parallel to standard crossover & mutation putting them in a cooperative-competitive standing with each other (Aguirre et al., 1999). An improved GA based on this model (GA-SRM) using an adaptive mechanism for parallel mutation significantly improves the performance of GAs (Aguirre et al., 2001). Now, we introduce a self-adaptive mechanism within the parallel mutation operator of GA-SRM and show that the model is an appropriate framework to effectively use and develop further self-adaptation within GAs.

We have presented an empirical model of genetic algorithms (GA) that puts parallel genetic operators in a cooperative-competitive stand with each other. An improved GA (GA-SRM) based on this model remarkably improves the search performance of a single population GA. We extend GA-SRM to distributed GAs in order to improve the performance of multiple population GAs. Simulation results verify that the parallel genetic operators in GA-SRM, CM and SRM, can successfully contribute to improve the search performance of distributed GAs.

A fully automatic method of detecting new lung nodules based on the subtraction method between two serial mass chest radiographs is proposed. Nodule detection is carried out following the three steps: pre-processing, subtraction and post-processing. In the pre-processing step, the outline of the bilateral lungs is extracted and modified in order to perform good subtraction between the two images to be compared. In the post-processing step, limitation and localization of candidate nodules of lung cancer are carried out. In these two steps, an evaluator function called "simpleness" of a single connection region is applied to identify outline of the lung and to select candidates of cancer(s).

An optimum spectrum is presented for a multilevel QAM system perturbed by timing error, phase error, and additive noise. We propose a double-jump spectrum (DJS) whose segment height is generally different from that of Franks' spectrum. The performance is analyzed with Gauss quadrature rules and a computer simulation. The results are compared to Franks' spectrum and the conventional cosine rolloff spectrum.

The general probability of error expressions of coherentM-ary PSK (M = 2, 4, 8, 16) systems in the presence of both impulsive and Gaussian noise are derived. This work is an extension of our paper related to BPSK and DPSK systems in a complex interference environment [1]. The presented impulsive and Gaussian channel disturbance is a first-order approximation to operational multichannel satellite and terrestrial microwave systems in which, in addition to front-end Gaussian noise, out-of-band intermodulation noise may be present. Our analysis contains an extension of the method developed by Bello and Esposito [2], [3], to include the Gaussian as well as the impulsive noise environment. The numerical results show that a well defined threshold region exists, above which the effect of the impulsive noise on the system performance is predominant.

In this paper we investigate the combined effect of sampling offset and phase error on the probability of error Peperformance of quadrature partial response (QPR) and offset QPR systems. The Gram-Charlier series expansion method is used to get accurate probability of error expressions. As an example, class 1 and class 4 partial response signaling schemes are examined.

General probability of error expressions of coherent PSK and differentially coherent DPSK systems in the presence of both impulsive and Gaussian noise are derived. The presented impulsive and Gaussian channel disturbance is a first-order approximation to operational multichannel satellite and terrestrial microwave systems in which in addition to front-end Gaussian noise, out-of-band intermodulation noise is also present. Our analysis contains an extension of the method developed by Bello and Esposito [1], [2] to include the Gaussian as well as the impulsive noise environment. Computed results are presented for a one-pole bandpass filter and a lognormal impulse amplitude distribution. The numerical results show that a welldefined threshold region exists, above which the effect of the impulsive noise on the system performance is predominant. A comparative study of system performance having integrate-sample-and-dump receivers with sample-only receivers is presented.

Signal-to-noise ratios (SNR's) at the output of the squarelaw detector are determined as the functions of modulation degree and bandwidth ratio of the modulating signal and the noise carrier, when the input is an amplitude-modulated noise carrier.