Michiharu MAEDA Hiromi MIYAJIMA
This paper presents novel algorithms for image restoration by state sharing methods with the stochastic model. For inferring the original image, in the first approach, a degraded image with gray scale transforms into binary images. Each binary image is independently inferred according to the statistical fluctuation of stochastic model. The inferred images are returned to a gray-scale image. Furthermore the restored image is constructed from the average of the plural inferred images. In the second approach, the binary state is extended to a multi-state, that is, the degraded image with Q state is transformed into n images with τ state and image restoration is performed. The restoration procedure is described as follows. The degraded image with Q state is prepared and is transformed into n images with τ state. The n images with τ state are independently inferred by the stochastic model and are returned to one image. Moreover the restored image is constructed from the average of the plural inferred images. Finally, the properties of the present approaches are described and the validity of them is confirmed through numerical experiments.
Digital watermarks on pictures must have the ability to survive various image processing operations while not causing degradation of picture quality. Random geometric distortion is one of the most difficult kinds of image processing for a watermark to survive, and this problem has become a central issue in watermarking research. Previous methods for dealing with random geometric distortion have been based on searches, special watermark patterns, learning, or additional data such as original pictures. Their use, however, is accompanied by large computational overhead or by operational inconvenience. This paper therefore proposes a method based on embedding watermark patterns in two of the three color planes constituting a color picture so that these two planes have a specific covariance. The detection of the embedded information is based on the covariance between these two planes. Random geometric distortion distorts all the constituent color planes of a picture in the same way and thus does not affect the covariance between any two. The covariance-based detection is therefore immune to the distortion. The paper clarifies that detection error would occur whenever the inherent covariance (the covariance in the original picture) overrides the covariance made by watermarking. The two constituent planes having the minimum inherent covariance are therefore selected and their inherent covariance is reduced by shifting one of them and using a noise-reduction preprocess. Experimental evaluations using StirMark confirmed that 64 bits embedded in 256256-pixel pictures can be correctly detected without using searches, special patterns, learning, or additional data.
In this study, a CG animation tool was designed that allows interpolation and extrapolation of two types of repeated motions including finger actions, for quantitative analyses of the relationship between features of human motions and subjective impressions. Three-dimensional human motions are measured with a magnetic motion capture and a pair of data gloves, and then relatively accurate time-series joint data are generated utilizing statistical characteristics. Based on the data thus obtained, time-series angular data of each joint for two dancing motions is transformed into frequency domain by Fourier transform, and spectral shape of each dancing action is interpolated. The interpolation and extrapolation of two motions can be synthesized with simple manner by changing an weight parameter while keeping good harmony of actions. Using this CG animation tool as a motion synthesizer, repeated human motions such as a dancing action that gives particular impressions on the observers can be quantitatively measured and analyzed by the synthesis of actions.
Atsushi HONDA Kei SAKAGUCHI Jun-ichi TAKADA Kiyomichi ARAKI
An RF front-end using a six-port circuit is a promising technology for realization of a compact software defined radio (SDR) receiver. Such a receiver, called a six-port direct conversion receiver (DCR), consists of analog circuit and digital signal processing components. The six-port DCR itself outputs four different linear combinations of received and local signals. The output powers are measured at each port, and the received signal is recovered by solving a set of linear equations. This receiver can easily cover a wide frequency band unlike the conventional DCR since it does not require the precise orthogonality that the conventional one does. In this paper, we propose a novel calibration method for a six-port system that includes nonlinear circuits such as diode detectors. We demonstrated the demodulation performance of a six-port DCR by computer simulation and experiments at 1.9, 2.45, and 5.85 GHz.
We propose a method to determine accurate motion fields and illumination conditions such as non-uniform or non-stationary illuminations. The method extends a stabilization method using reliability indices of optical flow to combine with a gradient-based approach that determines a motion field and illumination conditions simultaneously. We applied the proposed method to two synthetic image sequences and a standard image sequence. The method is effective for image sequences including poorly textured areas, edges of brightness variation, and almost dark objects.
In this paper, we study a work-conserving multicast scheduling with fanout splitting in a switch, which routes incoming packets asynchronously without fragmentation into cells. A new switch architecture is proposed, which distributes the input links to P variable length packet switching fabrics (VPS) with every G input links sharing GR inlets of the VPS. The system performance is analyzed by queueing analysis to express the maximum throughput and packet delay in terms of the system parameters and traffic characteristics. A practical switch design is also proposed to realize almost the same scheduling as the work-conserving one. We have surveyed how the fanout distribution affects the performance of the switch through Fanout Function, which is defined and studied to help the design of a multicast switch. We show how Fanout Function determines the maximum throughput and packet delay. Various fanout distributions are compared. The mixed fanout distribution exhibits better performance while the deterministic fanout can be used as a bound in the design of a multicast switch. We optimize R and P to attain 100% maximum throughput under limited switch complexity. When the mean fanout size is large, we can use less hardware to achieve the optimal performance by using our architecture. The proposed realization of this switch can be implemented easily due to its modular design. It is scalable because distributed output contention resolution and routing are used instead of a central arbitrator. Its performance is verified by simulation. The result matches the theoretical work-conserving scheduling very well.
Scattering of the two dimensional electromagnetic waves is studied by the infinite sequences of zeros arising on the complex plane, which just correspond to the null points of the far field pattern given as a function of the azimuthal angle θ. The convergent sequences of zeros around the point of infinity are evaluated when the scattering objects are assumed to be N-polygonal cylinders. Every edge condition can be satisfied if the locations of zeros are determined appropriately. The parameters, which allow us to calculate the exact positions of zeros, are given by the asymptotic analysis. It is also shown that there are N-directions of convergence, which tend to infinity. An illustrative example is presented.
Masaru KAMADA Mitsuhiro MATSUO
A digital/analog hybrid system is presented which implements the cardinal polynomial spline interpolation of arbitrary degree. Based on the fact that the (m-1)st derivative of a spline of degree m-1 is a staircase function, this system generates a cardinal spline of degree m-1 by m-1 cascaded integrators with a staircase function input. A given sequence of sampled values are transformed by a digital filter into coefficients for the B-spline representation of the spline interpolating the sampled values. The values of its (m-1)st derivative with respect to time are computed by the recurrence formula interpreting differentiation of the spline as difference of the coefficients. Then a digital-to-analog converter generates a staircase function representing the (m-1)st derivative, which is integrated by a cascade of m-1 analog integrators to make the expected spline. In order to cope with the offset errors involved in the integrators, a dynamical sampled-data control is attached. An analog-to-digital converter is employed to sample the output of the cascaded integrators. Target state of the cascaded integrators at each sampling instance is computed from the coefficients for the B-spline representation. The state error between the target and the estimated is compensated by feeding back a weighted sum of the state error to the staircase input.
In this letter, we show the effects of the chip waveform selection on the detection performance of the energy detector in DS/SS communications. Three chip waveforms such as rectangular, half-sine and raised-cosine are examined as the DS/SS chip waveform. It is demonstrated that the partial-band detection can enhance the detection performance of the energy detector approximately 50-70% compared with the full-band detection. When the chip rate is identical, the raised-cosine waveform shows lower detection probability due to its wider spreading bandwidth. However, when the spreading bandwidth is identical, the rectangular waveform shows lower detection probability due to its lower partial-band energy factor.
The AODV (Ad Hoc On-Demand Distance Vector) protocol is one of the typical reactive routing protocols in Ad Hoc networks, in that mobile nodes initiate routing activities only in the presence of data packets in need of a route. In this paper, we focus upon the local repair mechanism of AODV. When a link is broken, the upstream node of the broken link repairs the route to the destination by initiating a local route discovery process. The process involves the flooding of AODV control messages in every node within a radius of the length from the initiating node to the destination. In this paper, we propose an efficient local repair scheme for AODV, called AELR (AODV-based Efficient Local Repair). AELR utilizes the existing routing information in the downstream intermediate nodes which have been on the active route to the destination before a link break occurs. AELR can reduce the flooding range of AODV control messages and the route recovery time for route recovery because it can repair a route through the nearby downstream intermediate nodes. The performance results show that AELR can achieve faster route recovery than the local repair mechanism of AODV.
This work explores generative models of handwritten digit images using natural elastic nets. The analysis aims to extract global features as well as distributed local features of handwritten digits. These features are expected to form a basis that is significant for discriminant analysis of handwritten digits and related analysis of character images or natural images.
This paper presents a framework for the analysis of multi-antenna communication systems with mutually-coupled elements. The approach uses a network model that includes the coupled antennas, the propagation channel, the receiver matching network, a realistic noise model for the receive amplifiers, and externally generated interference. The general scheme is applied to diversity receivers, multiple-input multiple-output, and adaptive array architectures. Application of the formulation to coupled dipole antennas illustrates the impact of both mutual coupling and receiver matching on the performance of several representative multi-antenna applications.
Johan BAUWELINCK Dieter VERHULST Peter OSSIEUR Xing-Zhi QIU Jan VANDEWEGE Benoit DE VOS
This paper presents a new approach based on current mode circuits for fast and accurate optical level monitoring with wide dynamic range of a gigabit burst-mode laser driver chip. Our proposed solution overcomes the drawbacks that voltage mode implementations show at higher bit rates or in other technologies. The main speed-limiting factor of the level monitoring circuitry is the parasitic capacitance of the back facet monitor photodiode. We propose the use of an active-input current mirror to reduce the impact of this parasitic capacitance. The mirror produces two copies of the photo current, one to be used for the "0" level measurement and another for the "1" level measurement. The mirrored currents are compared to two reference currents by two current comparators. Every reference current needs only one calibration at room temperature. A pattern detection block scans the incoming data for patterns of sufficiently long consecutive 0's or 1's. At the end of such a pattern a valid measurement is present at the output of one of the current comparators. Based on these measurements the digital Automatic Power Control (APC) will adjust the bias (IBIAS) and modulation current (IMOD) setting of the laser driver. Tests show that the chip can stabilize and track the launched optical power with a tolerance of less than 1 dB. In these tests the pattern detection was programmed to sample the current comparators after 5 bytes (32 ns at 1.25 Gbps) of consecutive 1's and 0's. Automatic power control on such short strings of data has not been demonstrated before. Although this laser transmitter was developed for FSAN GPON applications at a speed of 1.25 Gbps upstream, the design concept is generic and can be applied for developing a wide range of burst mode laser transmitters. This chip was developed in a 0.35 µm SiGe BiCMOS process.
Yong-Jae KWAK So-Young PARK Joon-Ho LIM Hae-Chang RIM
In this paper, we propose a naïve probabilistic shift-reduce parsing model which can use contextual information more flexibly than the previous probabilistic GLR parsing models, and utilize the characteristics of agglutinative language in which the functional words are highly developed. Experimental results on Korean have shown that our model using the proposed contextual information improves the parsing accuracy more effectively than the previous models. Moreover, it is compact in model size, and is robust with a small training set.
Kenichi MASE Satoshi NARITA Sota YOSHIDA
We propose an efficient IP address assignment protocol in mobile ad hoc networks, which use a proactive routing protocol. In this method, which is termed the Bisected-Range based Assignment (BRA), a node repeatedly broadcasts an Agent Request to ask for address assignment. If there are one or more neighbor MANET nodes, one of them becomes an agent to select and assign an IP address to the requesting node. We use address location in the IP address space so that each agent maintains its own exclusive address range to be used for address selection, resulting to decrease the possibility of address conflict. If the requesting node cannot discover any neighbor MANET node over pre-determined random agent-search time, it selects by itself an IP address at random from the given address block. We evaluate performance of the basic and enhanced BRAs by computer simulation. It is shown that the basic and enhanced BRAs can reduce address conflict compared with random assignment. It is also shown that the enhanced BRA is superior in terms of control traffic overhead as well as address assignment delay over the random assignment with the strong Duplicate Address Detection.
Wenbin LUO Gregory L. HEILEMAN
The chaotic property of a new open addressing hash function, called exponential hashing, is presented. Our analysis indicates the connection between ergodic theory and hashing. Based on that, concepts from ergodic theory are applied to predict the performance of exponential hashing. Experimental results are presented to verify our theoretic analysis and the prediction.
Taichi YUKI Takayuki YAMAMOTO Masashi SUGANO Masayuki MURATA Hideo MIYAHARA Takaaki HATAUCHI
Since a radio channel is shared among terminals in an ad hoc network, packet collisions are frequent. In case of transmitting packets especially using TCP, data and ACK packets are transmitted in opposite directions on the same radio channel. Therefore, frequent collisions are unavoidable, and this seriously degrades TCP throughput. It is possible to transmit to two or more nodes which adjoin from a certain node simultaneously on the radio channel. To reduce the likelihood of packet collisions when an intermediate node transmits both data and ACK packets, these two types of packet can be combined and transmitted at the same time to increase the efficiency of radio channel utilization. In this paper, we propose a new technique to improve TCP performance by combining data and ACK packets. Our proposed technique is applicable to generic ad hoc networks easily. By means of a simulation using networks with various topologies, we have found that throughput can be improved by up to 60% by applying our proposed technique.
Whereas the autocorrelation is frequently used for pitch estimation, the resultant estimates usually suffer from inaccuracy. Instead of upsampling, we can improve the accuracy of the estimates by applying polynomial interpolation to the autocorrelation directly. For that purpose, four kernels, which are interpolating quadratic, quadratic-B spline, cubic-B spline, and cubic convolution kernels respectively, have been compared. Experiments show that the cubic B spline kernel shows the best performance, a little inferior to the computationally intensive upsampling procedure. The quadratic B spline kernel shows also reasonable performance with the merit of the further reduced computational complexities compared with the cubic B spline kernel.
The single path routing protocol, known as the Ad Hoc On-demand Distance Vector, has been widely studied for use in mobile ad hoc networks. AODV requires a new route discovery whenever a path breaks. Such frequent route discoveries cause a delay due to route discovery latency. To avoid such inefficiency, a multipath routing protocol has been proposed that attempts to find link-disjoint paths in a route discovery. However, when there are two or more common intermediate nodes on the path, the protocol can not find a pair of link-disjoint paths even if the paths actually exist. To reduce this route discovery latency, it is necessary to increase the opportunities for finding a pair of link-disjoint paths. In this paper, we focus on AODV and propose an AODV-based new multipath routing protocol for mobile ad hoc networks. The proposed routing protocol uses a new method to find a pair of link-disjoint paths by selecting a route having a small number of common intermediate nodes on its path. Using simulation models, we evaluate the proposed routing protocol and compare it with AODV and the existing multipath routing protocol. Results show that the proposed routing protocol achieves better performance in terms of delay than other protocols because it increases the number of cases where a pair of link-disjoint paths can be established.
Jaikwan JOO Sunghak LEE Hyunsoo YOON
Since IEEE 802.11 has many problems such as hidden node, exposed node problem, larger sensing function and BEB (Binary Exponential Back-off), it is not suitable for use in multi-hop wireless ad hoc network. When an on-demand routing protocol is used with 802.11 DCF (Distributed Coordination Function), the route to transmit the packet will be formed by nodes which have less competition (fewer neighbors) than others for the medium access. This effect will make that the routing path will be longer and decrease network performance. Therefore, we propose a new MAC (Medium Access Control) protocol that makes a shorter routing path, enabling better performance in multi-hop wireless ad hoc networks. The protocol modifying IEEE 802.11 DCF gives priority to the node with more neighbors and with less transmission. Through simulations, we have demonstrated that the proposed algorithm improves performance in terms of transmission rate, transmission delay and total consumption energy.