The application-specific multiprocessor system-on-chip (MPSoC) architecture is becoming an attractive solution to deal with increasingly complex embedded applications, which require both high performance and flexible programmability. As an effective method for MPSoC development, we present a gradual refinement flow starting from a high-level Simulink model to a synthesizable and executable hardware and software specification. The proposed methodology consists of five different abstract levels: Simulink combined algorithm and architecture model (CAAM), virtual architecture (VA), transactional accurate architecture (TA), virtual prototype (VP) and field-programmable gate array (FPGA) emulation. Experimental results of Motion-JPEG and H.264 show that the proposed gradual refinement flow can generate various MPSoC architectures from an original Simulink model, allowing processor, communication and tasks design space exploration.

We analyze the drawbacks of generally distributed time transition stochastic Petri nets (GDTT_SPN) in evaluating the performance of parallel systems, and propose a more general model, stochastic individual predicate/transition nets (SIPTN). SIPTN has higher modeling power and could provide more realistic models compared to GDTT_SPN, because in SIPTN the sojourn time distribution is determined not only by the transition, but also by the individuals. It is further proved that GDTT_SPN is a subset of SIPTN. As SIPTN introduces folding techniques from predicate/transition nets, SIPTN models have simpler and more intuitive graphic notations and accordingly higher usability, and thus are suitable for constructing simulation models for parallel systems.

We propose a new design scheme for a Booth encoder based on clocked transmission gate adiabatic logic (CTGAL). In the new design the structural complexity of the Booth encoder is reduced while the speed of the multiplier is improved. The adiabatic two’s complement multiplier-accumulator (MAC) is furthermore a design based on the CTGAL. The computer simulation results indicate that the designed circuit has the correct logic function and remarkably less energy consumption compared to that of the MAC based on complementary metal oxide semiconductor (CMOS) logic.

We present a new sense amplifier circuit for EEPROM memory. The topology of the sense amplifier uses a voltage sensing method, having low cost and low power consumption as well as high reliability. The sense amplifier was implemented in an EEPROM realized with an SMIC 0.35-μm 2P3M CMOS embedded EEPROM process. Under the condition that the power supply is 3.3 V, simulation results showed that the charge time is 35 ns in the proposed sense amplifier, and that the maximum average current consumption during the read period is 40 μA. The novel topology allows the circuit to function with power supplies as low as 1.4 V. The sense amplifier has been implemented in 2-kb EEPROM memory for RFID tag IC applications, and has a silicon area of only 240 μm².

Neurons with complex-valued weights have stronger capability because of their multi-valued threshold logic. Neurons with such features may be suitable for solution of different kinds of problems including associative memory, image recognition and digital logical mapping. In this paper, robustness or tolerance is introduced and newly defined for this kind of neuron according to both their mathematical model and the perceptron neuron’s definition of robustness. Also, the most robust design for basic digital logics of multiple variables is proposed based on these robust neurons. Our proof procedure shows that, in robust design each weight only takes the value of i or −i, while the value of threshold is with respect to the number of variables. The results demonstrate the validity and simplicity of using robust neurons for realizing arbitrary digital logical functions.

The performance of complementary feeders, running in parallel, can be significantly improved by installing static transfer switches (STSs) at critical locations. We develop the STS control logic, which transfers the critical load from the preferred feeder to the alternate feeder when a voltage sag or a fault occurs on the preferred feeder. A forced commutation technique is proposed and implemented to turn off the preferred feeders’ thyristor, thus avoiding cross current to flow and minimizing the transfer time. Simulation results show that the forced commutation technique is more effective as compared to the recently proposed time delay technique for STS operation. Two different feeders, namely New Exchange, the preferred feeder, and Sector I-10/2, the alternate feeder of Islamabad Electric Supply COmpany (IESCO), Pakistan, have been selected for case studies. The software PSCAD/EMTDC professional package has been used for simulation.

We propose a position sensorless control scheme for a four-switch, three-phase brushless DC motor drive, based on the zero crossing point detection of phase back-EMF voltages using newly defined error functions (EFs). The commutation instants are 30° after detected zero crossing points of the EFs. Developed EFs have greater magnitude rather than phase or line voltages so that the sensorless control can work at a lower speed range. Moreover, EFs have smooth transitions around zero voltage level that reduces the commutation errors. EFs are derived from the filtered terminal voltages v_ao and v_bo of two low-pass filters, which are used to eliminate high frequency noises for calculation of the average terminal voltages. The feasibility of the proposed sensorless control is demonstrated by simulation and experimental results.

We present a derived grid-based model for the simulation of pedestrian flow. Interactions among pedestrians are considered as the result of forces within a certain neighbourhood. Unlike the social force model, the forces here, as in Newtonian physics, are proportional to the inverse of the square of the distance. Despite the notion of neighbourhood and the underlying grid, this model differs from the existing cellular automaton (CA) models in that the pedestrians are treated as individuals. Bresenham’s algorithm for line rastering is applied in the step calculation.

In the daily life, people often repeat regular routes in certain periods. In this paper, a mining system is developed to find the continuous route patterns of personal past trips. In order to count the diversity of personal moving status, the mining system employs the adaptive GPS data recording and five data filters to guarantee the clean trips data. The mining system uses a client/server architecture to protect personal privacy and to reduce the computational load. The server conducts the main mining procedure but with insufficient information to recover real personal routes. In order to improve the scalability of sequential pattern mining, a novel pattern mining algorithm, continuous route pattern mining (CRPM), is proposed. This algorithm can tolerate the different disturbances in real routes and extract the frequent patterns. Experimental results based on nine persons’ trips show that CRPM can extract more than two times longer route patterns than the traditional route pattern mining algorithms.

Images with human faces comprise an essential part in the imaging realm. Occlusion or damage in facial portions will bring a remarkable discomfort and information loss. We propose an algorithm that can repair occluded or damaged facial images automatically, named ‘facial image inpainting’. Inpainting is a set of image processing methods to recover missing image portions. We extend the image inpainting methods by introducing facial domain knowledge. With the support of a face database, our approach propagates structural information, i.e., feature points and edge maps, from similar faces to the missing facial regions. Using the inferred structural information as guidance, an exemplar-based image inpainting algorithm is employed to copy patches in the same face from the source portion to the missing portion. This newly proposed concept of facial image inpainting outperforms the traditional inpainting methods by propagating the facial shapes from a face database, and avoids the problem of variations in imaging conditions from different images by inferring colors and textures from the same face image. Our system produces seamless faces that are hardly seen drawbacks.

Segmentation of the bladder in computerized tomography (CT) images is an important step in radiation therapy planning of prostate cancer. We present a new segmentation scheme to automatically delineate the bladder contour in CT images with three major steps. First, we use the mean shift algorithm to obtain a clustered image containing the rough contour of the bladder, which is then extracted in the second step by applying a region-growing algorithm with the initial seed point selected from a line-by-line scanning process. The third step is to refine the bladder contour more accurately using the rolling-ball algorithm. These steps are then extended to segment the bladder volume in a slice-by-slice manner. The obtained results were compared to manual segmentation by radiation oncologists. The average values of sensitivity, specificity, positive predictive value, negative predictive value, and Hausdorff distance are 86.5%, 96.3%, 90.5%, 96.5%, and 2.8 pixels, respectively. The results show that the bladder can be accurately segmented.

We propose a novel texture clustering method. A classical type of (approximate) shift invariant discrete wavelet transform (DWT), dual tree DWT, is used to decompose texture images. Multiple signatures are generated from the obtained high-frequency bands. A locality preserving approach is applied subsequently to project data from high-dimensional space to low-dimensional space. Shift invariant DWT can represent image texture information efficiently in combination with a histogram signature, and the local geometrical structure of the dataset is preserved well during clustering. Experimental results show that the proposed method remarkably outperforms traditional ones.

Currently there are two approaches for a multi-class support vector classifier (SVC). One is to construct and combine several binary classifiers while the other is to directly consider all classes of data in one optimization formulation. For a K-class problem (K>2), the first approach has to construct at least K classifiers, and the second approach has to solve a much larger optimization problem proportional to K by the algorithms developed so far. In this paper, following the second approach, we present a novel multi-class large margin classifier (MLMC). This new machine can solve K-class problems in one optimization formulation without increasing the size of the quadratic programming (QP) problem proportional to K. This property allows us to construct just one classifier with as few variables in the QP problem as possible to classify multi-class data, and we can gain the advantage of speed from it especially when K is large. Our experiments indicate that MLMC almost works as well as (sometimes better than) many other multi-class SVCs for some benchmark data classification problems, and obtains a reasonable performance in face recognition application on the AR face database.

To extract the maximum power from a photovoltaic (PV) energy system, the real-time maximum power point (MPP) of the PV array must be tracked closely. The non-linear and time-variant characteristics of the PV array and the non-linear and non-minimum phase characteristics of a boost converter make it difficult to track the MPP for traditional control strategies. We propose a fuzzy neural network controller (FNNC), which combines the reasoning capability of fuzzy logical systems and the learning capability of neural networks, to track the MPP. With a derived learning algorithm, the parameters of the FNNC are updated adaptively. A gradient estimator based on a radial basis function neural network is developed to provide the reference information to the FNNC. Simulation results show that the proposed control algorithm provides much better tracking performance compared with the fuzzy logic control algorithm.

We propose a medium access control (MAC) protocol for uplink transmissions in wireless local area networks (WLANs), where both stations and access points (APs) are equipped with multiple antennas. The protocol solves some common problems in utilizing multiple input multiple output (MIMO) under the 802.11 protocol, e.g., how to deploy preamble (training sequence) used for channel estimation and how to enable simultaneous data transmissions, and facilitates two simultaneous uplink data transmissions via a cross-layer approach. Furthermore, we develop a 3D discrete-time Markov model to analyze the performance of the proposed WLAN scheme. The analytical results are verified by simulation, and numerical results show that the system throughput can be significantly improved by our proposed scheme as compared with conventional schemes.

As the applications of wireless sensor networks (WSNs) diversify, providing secure communication is emerging as a critical requirement. In this paper, we investigate the detection of wormhole attack, a serious security issue for WSNs. Wormhole attack is difficult to detect and prevent, as it can work without compromising sensor nodes or breaching the encryption key. We present a wormhole attack detection approach based on the probability distribution of the neighboring-node-number, WAPN, which helps the sensor nodes to judge distributively whether a wormhole attack is taking place and whether they are in the influencing area of the attack. WAPN can be easily implemented in resource-constrained WSNs without any additional requirements, such as node localization, tight synchronization, or directional antennas. WAPN uses the neighboring-node-number as the judging criterion, since a wormhole usually results in a significant increase of the neighboring-node-number due to the extra attacking link. Firstly, we model the distribution of the neighboring-node-number in the form of a Bernoulli distribution. Then the model is simplified to meet the sensor nodes’ constraints in computing and memory capacity. Finally, we propose a simple method to obtain the threshold number, which is used to detect the existence of a wormhole. Simulation results show that WAPN is effective under the conditions of different network topologies and wormhole parameters.

Multi-proxy signature schemes allow the original signer to delegate his/her signing power to n proxy signers such that all proxy signers must corporately generate a valid proxy signature on behalf of the original signer. We first propose a multi-proxy signature scheme based on discrete logarithms and then adapt it to the elliptic curve cryptosystem. With the integration of self-certified public-key systems and the message recovery signature schemes, our proposed schemes have the following advantages: (1) They do not require the signing message to be transmitted, since the verifier can recover it from the signature; (2) The authentication of the public keys, verification of the signature, and recovery of the message can be simultaneously carried out in a single logical step; (3) No certificate is needed for validating the public keys. Further, the elliptic curve variant with short key lengths especially suits the cryptographic applications with limited computing power and storage space, e.g., smart cards. As compared with the previous work that was implemented with the certificate-based public-key systems, the proposed schemes give better performance in terms of communication bandwidth and computation efforts.

We propose a novel high-performance hardware architecture of processor for elliptic curve scalar multiplication based on the Lopez-Dahab algorithm over GF(2¹⁶³) in polynomial basis representation. The processor can do all the operations using an efficient modular arithmetic logic unit, which includes an addition unit, a square and a carefully designed multiplication unit. In the proposed architecture, multiplication, addition, and square can be performed in parallel by the decomposition of computation. The point addition and point doubling iteration operations can be performed in six multiplications by optimization and solution of data dependency. The implementation results based on Xilinx VirtexII XC2V6000 FPGA show that the proposed design can do random elliptic curve scalar multiplication GF(2¹⁶³) in 34.11 μs, occupying 2821 registers and 13 376 LUTs.

We construct the membership functions of the fuzzy objective values of a controllable queueing model, in which cost elements, arrival rate and service rate are all fuzzy numbers. Based on Zadeh’s extension principle, a set of parametric nonlinear programs is developed to find the upper and lower bounds of the minimal average total cost per unit time at the possibility level. The membership functions of the minimal average total cost are further constructed using different values of the possibility level. A numerical example is solved successfully to illustrate the validity of the proposed approach. Because the object value is expressed and governed by the membership functions, the optimization problem in a fuzzy environment for the controllable queueing models is represented more accurately and analytical results are more useful for system designers and practitioners.