This paper presents a fuzzy control-based intelligent medium access controller (FiMAC), which optimizes random access control between heterogeneous traffic aiming at more efficient voice/data integrated services in dynamic reservation TDMA-based broadband radio access networks. In order to achieve the design objective, viz. a differentiated quality-of-service (QoS) guarantee for individual service plus maximal system resource utilization, the FiMAC intelligently and independently controls the random access parameters such as the lengths of random access regions dedicated to respective service traffic and the corresponding permission probabilities, frame-by-frame basis. In addition, we have adopted a mobile-assisted random access mechanism where the voice terminal readjusts a global permission probability from the FiMAC, to handle the 'fair access' issue resulting from distributed queueing problems inherent in the access network. Our extensive simulation results indicate that the FiMAC is well coordinated with a mobile-assisted mechanism such that significant improvements are achieved in terms of voice capacity, delay, and fairness over most of the existing MAC schemes for the integrated services.

The adaptive phase tracking scheme for orthogonal frequency division multiplexing (OFDM) signals can provide superior PER performance in channels with varying phase noise power. It is an effective technique for achieving high-rate and high quality wireless transmission. This paper proposes a new simple adaptive phase tracking scheme for OFDM signals in order to realize high-rate wireless local area networks (LANs). The proposed scheme measures the integrated phase rotation in order to appropriately set the properties of the FIR filter in the phase tracking circuits. This scheme uses the fact that the integrated phase rotation is correlated to the phase noise power. Assuming an RMS delay spread of 100 ns, computer simulations show that the proposed scheme offers superior required Eb/N0 performance (with regard to the phase noise power) compared to the conventional fixed-tap scheme, where the phase noise to signal power ratios are below -18 dB. It also offers excellent PER performance at the packet length of 1000 bytes unlike the conventional schemes, which suffer degraded PER performance.

OFDM transmission performance in mobile communications suffers severe degradation caused by multipath delay difference greater than the Guard Interval (GI). This is because the excess delay results in considerable Inter-Symbol Interference (ISI) between temporally adjacent symbols and Inter-Carrier Interference (ICI) among subcarriers in the same symbol. This paper proposes a robust OFDM receiver for the scattered pilot OFDM signal that can effectively suppress both ISI and ICI by using two types of equalization and a smoothed FFT-window. In order to verify the performance of the proposed receiver, computer simulations are conducted in accordance with the scattered pilot OFDM signal format of the Digital Terrestrial Television Broadcasting (DTTB). The simulation results demonstrate that the proposed receiver shows much better performance than the conventional receiver in multipath fading environments with the delay difference greater than GI duration.

We have designed a 44 Banyan switch using SFQ logic circuits. The switch is composed of three parts; one is an input buffer, the second is a contention solver which checks packet contention in a distribution network, and the third is a packet distribution network which distributes contention-free packets to their destination address. The packet distribution network is composed of Batcher-Banyan switch with the input buffer. The contention solver decides to send a data packet to the distribution network, using only internal routing tags which are added to packets in the switch. As the circuit is composed of two parts, the contention solver and the packet distribution network, the transfer rate is raised because it doesn't need to wait any more while a data packet passes through the distribution network. Simulation results using JSIM show that the switch circuit can operate at a clock frequency of 40 GHz.

We investigated single flux quantum sinc filters with multistage decimation structure in order to realize high-speed sinc filter operation. Second- and third-order (k=2, 3) sinc filters with a decimation factor N=2 were designed and confirmed their proper operations. These sinc filters with N=2 are utilized as elementary circuit blocks of our multistage decimation sinc filters with N=2M, where M indicates the number of the stage of the decimation. As an example of the multistage decimation filter, we designed a k=2, N=4 sinc filter which was formed from a two-stage decimation structure using k=2, N=2 sinc filters, and confirmed its proper operation. The k=2, N=4 sinc filter consisted of 1372 Josephson junctions with the power consumption of 191 µW.

A new distributed medium access control (MAC) protocol--Soft Reservation Multiple Access with Priority Assignment (SRMA/PA) protocol--is introduced for supporting the integrated services of real-time and non-real-time applications in mobile ad-hoc networks. The SRMA/PA protocol allows the distributed nodes to contend for and reserve time slots with RTS/CTS-like "collision-avoidance" handshake and "soft reservation" mechanism augmented with distributed and dynamic access priority control. The SRMA/PA protocol realizes distributed scheduling for guaranteeing QoS requirements of integrated services and maximizes statistical multiplexing gain. We have demonstrated by simulation studies that the multiplexing gain can be improved significantly without unduly compromising on system complexity. Moreover, we have shown that the proposed back-off mechanism designed for delay-constrained services is useful for further improving utilization of the channel.

To improve the CT image degraded by radiographic noise (such as quantum mottle), we propose a method based on the wavelet transform modulus sum (WTMS). The noise and regular parts of a signal can be observed by tracing the evolution of its WTMS across scales. Our results show that most of the quantum mottle in the projections of Shepp-Logan phantom has been removed by the proposed method with the supposed cranium well preserved. The denoised CT images show good signal to noise ratio in the region of interest. We also have investigated the relation between the number of X-ray photons and the quality of images reconstructed from denoised projections. From experimental results, this method shows the possibility to reduce a patient's dose about 1/10 with the same visual quality.

A wireless communications system with a transmission speed of 18 Mbit/s is presented using the 2.4 GHz ISM band. This system employs the Carrier Frequency Offset-Spread Spectrum (CFO-SS) scheme and the Dual-Polarization Staggered Transmission (DPST) scheme. The 18 Mbit/s CFO-SS system (named CFO-SS18) was developed and its performance evaluated in fields. In this paper, the detailed operating principle of CFO-SS and DPST schemes, together with the specifications and structures of CFO-SS18, are presented. Results of indoor and field tests obtained by using CFO-SS18 are also presented.

A new approach to build up a real-time multiprocessing system that is configuration flexible for evaluating space-time (ST) equalizers is described. The core of the system consists of fully programmable devices such as digital signal processors (DSPs), field-programmable gate arrays (FPGAs), and reduced instruction set computers (RISCs) with a real-time operating system (RTOS). The RTOS facilitates flexibility in the multi-processor configuration for the system conforming with ST processing algorithms. Timing jitter synchronization caused by use of the RTOS-embedded system is shown, and an adjustable frame format for a transmission system is described as a measure to avoid the jitter problem. Bit error rate (BER) performances measured in uncorrelated frequency-selective fading channels show that an ST equalizer provides a significantly lower BER than an array processor does.

In an orthogonal frequency division multiplexing (OFDM) system, the bit error performance is degraded in the presence of multiple propagation paths whose excess delays are longer than the Guard Interval (GI), because the orthogonality between subcarriers cannot be maintained. In this paper, we propose a new OFDM demodulation method with a variable-length effective symbol and a multi-stage inter-carrier interference (ICI) canceller, in order to improve the bit error performance in the presence of multipaths whose excess delays are longer than the GI. The influence of the inter-symbol interference (ISI) is eliminated by the variable-length effective symbol, and then the ICI component is reduced by the multi-stage ICI canceller. The principle of the proposed method is explained, and the performance of the proposed method is then evaluated by computer simulation. The results show that the proposed method improves the system availability under more various multipath fading environments without changing the system parameters.

The sequence-pair was proposed as a representation method of block placement to determine the densest possible placement of rectangular modules in VLSI layout design. A method of achieving bottom left corner packing in O(n2) time based on a given sequence-pair of n rectangles was proposed using horizontal/vertical constraint graphs. Also, a method of determining packing from a sequence-pair in O(n log n) time was proposed. Another method of obtaining packing in O(n log log n) time was recently proposed, but further improvement was still required. In this paper, we propose a method of obtaining packing via the Q-sequence (representation of rectangular dissection) in O(n+k) time from a given sequence-pair of n rectangles with k subsequences called adjacent crosses, given the position of adjacent crosses and the insertion order of dummy modules into adjacent crosses. The position of adjacent crosses and insertion order of dummy modules can be obtained from a sequence-pair in O(n+k) time using the conventional method. Here, we prove that arbitrary packing can be represented by a sequence-pair, keeping the value of k not more than n-3. Therefore, we can determine packing from a sequence-pair with k of O(n) in linear time using the proposed method and the conventional method.

This paper presents an adaptive burst-mode M-QAM modem architecture suitable for variable rate broadband wireless packet data networks. The core signal processing functions for the modem are common to all constellations resulting in an efficient hardware architecture for field programmable gate array (FPGA) implementation.

A blind equalizer which uses the differential constant modulus algorithm (DCMA) is introduced. An anchored FIR equalizer applied to a first-order autoregressive channel and updated according to the DCMA is shown to converge to the inverse of that channel regardless of the initial tap-weights and the gain along the direct path.

The jaw movements can be measured by estimating the position and orientation of two small permanent magnets attached on the upper and lower jaws. It is a difficult problem to estimate the positions and orientations of the magnets from magnetic field because it is a typical inverse problem. The back propagation neural networks (BPNN) are applicable to solve this problem in short processing time. But its precision is not enough to apply to practical measurement. In the other hand, precise estimation is possible by using the nonlinear least-square (NLS) method. However, it takes long processing time for iterative calculation, and the solutions may be trapped in the local minima. In this paper, we propose a precise and fast measurement system which makes use of the estimation algorithm combining BPNN with NLS method. In this method, the BPNN performs an approximate estimation of magnet parameters in short processing time, and its result is used as the initial value of iterative calculation of NLS method. The cost function is solved by Gauss-Newton iteration algorithm. Precision, processing time and noise immunity were examined by computer simulations. These results shows the proposed system has satisfactory ability to be applied to practical measurement.

In this paper, a new approach is proposed to improve the channel estimation accuracy with channel tracking capability for adaptive multicarrier equalization systems under time-variant multipath fading channel. The improvement is carried out based on the assumption that the channel is static over a transmitted block period, and slowly linearly changing over several block periods. By applying IFFT to the concatenated channel transfer function derived from different blocks, the noise-averaging improvement is achieved, and a better estimation of the channel coefficients with some delay can be obtained. A multi-step channel predictor and a smoothing filter is utilized to compensate for the delay and make the system more robust in terms of channel tracking performance. Adaptive time domain equalization is jointly performed with this approach to avoid the channel invertibility problem found in the frequency domain approach. A short period of training sequences is utilized resulting in more efficient use of available communication capacity. The effectiveness of the proposed approach is evaluated through simulation for multicarrier systems in time-variant multipath fading channels. Results show improvement over previous channel estimation schemes.

When we use an adaptive array antenna (AAA) with the minimum mean square error (MMSE) criterion under the multipath environment, where the receiving signal level varies, it is difficult for the AAA to converge because of the distortion of the desired wave. Then, we need the equalization both in space and time domains. A tapped-delay-line adaptive array antenna (TDL-AAA) and the AAA with linear equalizer (AAA-LE) have been proposed as simple space-temporal equalization. The AAA-LE has not utilized the recursive least square (RLS) algorithm. In this paper, we propose a space-temporal simultaneous processing equalizer (ST-SPE) that is an AAA-LE with the RLS algorithm. We proposed that the first tap weight of the LE should be fixed and the necessity of that is derived from a normal equation in the MMSE criterion. We achieved the space-temporal simultaneous equalization with the RLS algorithm by this configuration. The ST-SPE can reduce the computational complexity of the space-temporal joint equalization in comparison to the TDL-AAA, when the ST-SPE has almost the same performance as the TDL-AAA in multipath environment with minimum phase condition such as appeared at line-of-sight (LOS).

In this paper, we analyze the convergence and steady-state behavior of the least mean-square (LMS) adaptive filtering algorithm for a finite-length phase-splitting hybrid-type decision feedback equalizer (H-DFE). With some approximations, we derive an iterative expression for the excess mean-square error (MSE) of the H-DFE, which is composed of three statistically dependent excess MSEs; that is, the excess MSEs of the feedforward filter (FFF), intersymbol interference predictive feedback filter (ISI-FBF), and noise predictive feedback filter (NP-FBF) taps. Computer simulation and analytical results show that the average eigenvalue of the input signal for the NP-FBF taps of the H-DFE is time-varying, whereas those for the FFF and ISI-FBF taps are fixed. Nevertheless, the H-DFE can be implemented with fixed step sizes that ensure the convergence of the LMS algorithm without performance degradation from the standpoint of convergence speed, as well as steady-state performance for digital subscriber line (xDSL) applications.

The modern network service of finding the optimal path subject to multiple constraints on performance metrics such as delay, jitter, loss probability, etc. gives rise to the multi-constrained optimal-path (MCOP) QoS routing problem, which is NP-complete. In this paper, this problem is solved through both exact and heuristic algorithms. We propose an exact algorithm E_MCOP, which first constructs an aggregate weight and then uses a K-shortest-path algorithm to find the optimal solution. By means of E_MCOP, the performance of the heuristic algorithm H_MCOP proposed by Korkmaz et al. in a recent work is evaluated. H_MCOP only runs Dijkstra's algorithm (with slight modifications) twice, but it can find feasible paths with a success ratio very close to that of the exact algorithm. However, we notice that in certain cases its feasible solution has an unsatisfactorily high average cost deviation from the corresponding optimal solution. For this reason, we propose some modified algorithms based on H_MCOP that can significantly improve the performance by running Dijkstra's algorithm a few more times. The performance of the exact algorithm and heuristics is investigated through computer simulations on networks of various sizes.

In this paper, we propose a packet-scheduling algorithm, called the Class-level Service Lagging (CSL) algorithm, that guarantees multiple delay bounds for multi-class traffic in packet networks. We derive the associated schedulability test conditions, which are used to determine call admission. We first introduce a novel implementation of priority control, which has a conventional and simple form. We show how the efforts to confirm the logical validity of that implementation are managed to reach the definition of the CSL algorithm. The priority control is realized by imposing class-level unfairness in service provisioning, while the underlying service mechanism is carried out using the notion of fair queueing. The adoption of fair queueing allows the capability to maintain the service quality of the well-behaving traffic even in the presence of misbehaving traffic. We call this the firewall property. Simulation results demonstrate the superiority of the CSL algorithm in both priority control and firewall functionality. We also describe how the CSL algorithm is implementable with a computational complexity of O(1). Those features as well as the enhanced scalability, which results from the class-level approach, confirm the adequacy of the CSL algorithm for the fast packet networks.

In this paper, we propose an efficient quality-providing scheme to satisfy delay bound and loss ratio requirements for real-time video applications. To utilize network resources more efficiently while meeting service requirements, the network resources are dynamically allocated to each video connection based on the predicted traffic and currently provided quality of service degree. With the proposed bandwidth allocation method, a fair quality of service support in terms of packet loss ratio and maximum packet transfer delay to each video source can be achieved. To avoid possible quality violation by incoming new video connections, we present a connection admission control based on the provided QoS for existing connections and the measured traffic statistics. Simulation results show that our proposed dynamic method is able to provide accurate quality control.