This paper proposes a distributed observer on a sensor network, where communication on the network is randomly performed. This work is a natural extension of Kalman consensus filter approach to the cases involving random communication. In both bidirectional and unidirectional communication cases, gain conditions that guarantee improvement of estimation error convergence compared to the case with no communication are obtained. The obtained conditions are more practical than those of previous studies and give appropriate cooperative gains for a given communication probability. The effectiveness of the proposed method is confirmed by computer simulations.

This paper presents a 120-GHz-band split ring resonator (SRR) bandstop filter whose insertion loss can be controlled by coupling another lattice pattern substrate. The SRR bandstop filter and lattice pattern substrate is composed of 200-µm-thick quartz substrate and 5-µm-thick gold patterns. S21 of the SRR bandstop filter is -37.8 dB, and its -10-dB bandwidth is 115-130 GHz. S21 of the SRR bandstop filter changes to -4.1 dB at 125 GHz by arranging the lattice pattern substrate in close proximity to the SRR stopband filter, because coupling between the SRR and the lattice pattern occurs when the SRR and lattice pattern are opposed in close proximity. It was found that 10 Gbit/s data transmission can be achieved by setting the lattice pattern substrate just above the SRR bandstop filter with a spacer thickness of 50 µm, even though data transmission is impossible when only the SRR bandstop filter is inserted between the transmitter and the receiver.

Universal filtered multicarrier (UFMC) systems offer a flexibility of filtering sub-bands with arbitrary bandwidth to suppress out-of-band (OoB) emission, while keeping the orthogonality between subcarriers in one sub-band. Oscillator discrepancies between the transmitter and receiver induce carrier frequency offset (CFO) in practical systems. In this paper, we propose a novel CFO estimation method for UFMC systems that has very low computational complexity and can then be used in practical systems. In order to fully exploit the coherence bandwidth of the channel, the training symbols are designed to have several identical segments in the frequency domain. As a result, the integral part of CFO can be estimated by simply determining the correlation between received signal and the training symbol. Simulation results show that the proposed method can achieve almost the same performance as an existing method and even a better performance in channels that have small decay parameter values. The proposed method can also be used in other multicarrier systems, such as orthogonal frequency division multiplexing (OFDM).

Recently, visual trackers based on the framework of kernelized correlation filter (KCF) achieve the robustness and accuracy results. These trackers need to learn information on the object from each frame, thus the state change of the object affects the tracking performances. In order to deal with the state change, we propose a novel KCF tracker using the filter response map, namely a confidence map, and adaptive model. This method firstly takes a skipped scale pool method which utilizes variable window size at every two frames. Secondly, the location of the object is estimated using the combination of the filter response and the similarity of the luminance histogram at multiple points in the confidence map. Moreover, we use the re-detection of the multiple peaks of the confidence map to prevent the target drift and reduce the influence of illumination. Thirdly, the learning rate to obtain the model of the object is adjusted, using the filter response and the similarity of the luminance histogram, considering the state of the object. Experimentally, the proposed tracker (CFCA) achieves outstanding performance for the challenging benchmark sequence (OTB2013 and OTB2015).

In the realization of convolutional neural networks (CNNs) in resource-constrained embedded hardware, the memory footprint of weights is one of the primary problems. Pruning techniques are often used to reduce the number of weights. However, the distribution of nonzero weights is highly skewed, which makes it more difficult to utilize the underlying parallelism. To address this problem, we present SENTEI*, filter-wise pruning with distillation, to realize hardware-aware network architecture with comparable accuracy. The filter-wise pruning eliminates weights such that each filter has the same number of nonzero weights, and retraining with distillation retains the accuracy. Further, we develop a zero-weight skipping inter-layer pipelined accelerator on an FPGA. The equalization enables inter-filter parallelism, where a processing block for a layer executes filters concurrently with straightforward architecture. Our evaluation of semantic-segmentation tasks indicates that the resulting mIoU only decreased by 0.4 points. Additionally, the speedup and power efficiency of our FPGA implementation were 33.2× and 87.9× higher than those of the mobile GPU. Therefore, our technique realizes hardware-aware network with comparable accuracy.

In this paper, we propose a method which enables us to control the variance of the coefficients of the LMS-type adaptive filters. In the method, each coefficient of the adaptive filter is modeled as an random variable with a Gaussian distribution, and its value is estimated as the mean value of the distribution. Besides, at each time, we check if the updated value exists within the predefined range of distribution. The update of a coefficient will be canceled when its updated value exceeds the range. We propose an implementation method which has similar formula as the Gaussian mixture model (GMM) widely used in signal processing and machine learning. The effectiveness of the proposed method is evaluated by the computer simulations.

Nowadays recommender systems (RS) keep drawing attention from academia, and collaborative filtering (CF) is the most successful technique for building RS. To overcome the inherent limitation, which is referred to as data sparsity in CF, various solutions are proposed to incorporate additional social information into recommendation processes, such as trust networks. However, existing methods suffer from multi-source data integration (i.e., fusion of social information and ratings), which is the basis for similarity calculation of user preferences. To this end, we propose a social collaborative filtering method based on novel trust metrics. Firstly, we use Graph Convolutional Networks (GCNs) to learn the associations between social information and user ratings while considering the underlying social network structures. Secondly, we measure the direct-trust values between neighbors by representing multi-source data as user ratings on popular items, and then calculate the indirect-trust values based on trust propagations. Thirdly, we employ all trust values to create a social regularization in user-item rating matrix factorization in order to avoid overfittings. The experiments on real datasets show that our approach outperforms the other state-of-the-art methods on usage of multi-source data to alleviate data sparsity.

The flooding DDoS attack is a serious problem these days. In order to detect the flooding DDoS attack, the survival approaches and the mitigation approaches have been investigated. Since the survival approach occurs the burden on the victims, the mitigation approach is mainly studied. As for the mitigation approaches, to detect the flooding DDoS attack, the conventional schemes using the bloom filter, machine learning, and pattern analyzation have been investigated. However, those schemes are not effective to ensure the high accuracy (ACC), the high true positive rate (TPR), and the low false positive rate (FPR). In addition, the data size and calculation time are high. Moreover, the performance is not effective from the fluctuant attack packet per second (pps). In order to effectively detect the flooding DDoS attack, we propose the lightweight detection using bloom filter against flooding DDoS attack. To detect the flooding DDoS attack and ensure the high accuracy, the high true positive rate, and the low false positive rate, the dec-all (decrement-all) operation and the checkpoint are flexibly changed from the fluctuant pps in the bloom filter. Since we only consider the IP address, all kinds of flooding attacks can be detected without the blacklist and whitelist. Moreover, there is no complexity to recognize the attack. By the computer simulation with the datasets, we show our scheme achieves an accuracy of 97.5%. True positive rate and false positive rate show 97.8% and 6.3%, respectively. The data size for processing is much small as 280bytes. Furthermore, our scheme can detect the flooding DDoS attack at once in 11.1sec calculation time.

We propose a nonphotorealistic rendering method for generating checkered pattern images from photographic images. The proposed method is executed by iterative calculation using a Prewitt filter with an expanded window size and can automatically generate checkered patterns according to changes in edges and shade of photographic images. To verify the effectiveness of the proposed method, an experiment was conducted using various photographic images. An additional experiment was conducted to visually confirm the checkered pattern images generated by changing the iteration number, window size, and parameter to emphasize the checkered patterns.

In this paper, we propose a radio frequency (RF) anti-aliasing filter design method considering the effect of a roll-off characteristic on a noise figure (NF) in the direct RF undersampling receiver. The proposed method is useful for broadband reception that a system bandwidth (BW) has nearly half of the sampling frequency (1/2 fs). When the system BW is extended nearly 1/2 fs, the roll-off band is out of the desired Nyquist zone and it affects NF additionally. The proposed method offers a design target regarding the roll-off characteristic not only the rejection ratio. The target is helpful as a design guide to meet the allowed NF. We design the filter based on the proposed method and it is applied to the direct RF undersampling on-board receiver for Ka-band high throughput satellite (HTS). The measured NF value of the implemented receiver almost matched the designed value. Moreover, the receiver achieved the reception bandwidth which is 90% of 1/2 fs.

Speech enhancement has been proposed to reduce the impulsive noise whose frequency characteristic is wideband. On the other hand, it is challenging to reduce the ringing sound, which is narrowband in impulsive noise. Therefore, we propose the modeling of the ringing sound and its estimation by a linear predictor (LP). However, it is difficult to estimate the ringing sound only in noisy speech due to the auto-correlation property of speech. The proposed system adopts the 4th order moment-based adaptive algorithm by noticing the difference between the 4th order statistics of speech and impulsive noise. The brief analysis and simulation results show that the proposed system has the potential to reduce ringing sound while keeping the quality of enhanced speech.

Hum noise such as power line interference is one of the critical problems in the biomedical signal acquisition. Various techniques have been proposed to suppress power line interference. However, some of the techniques require more components and power consumption. The notch depth in the conventional N-path notch filter circuits needs a higher number of paths and switches off-resistance. It makes the conventional N-path notch filter less of efficiency to suppress hum noise. This work proposed the new N-path notch filter to hum noise suppression in biomedical signal acquisition. The new N-path notch filter achieved notch depth above 40dB with sampling frequency 50Hz and 60Hz. Although the proposed circuits use less number of path and switches off-resistance. The proposed circuit has been verified using artificial ECG signal contaminated by hum noise at frequency 50Hz and 60Hz. The output of N-path notch filter achieved a noise-free signal even if the sampling frequency changes.

A polyphase sequence set with orthogonality consisting complex elements with unit magnitude, can be expressed by a unitary matrix corresponding to the complex Hadamard matrix or the discrete Fourier transform (DFT) matrix, whose rows are orthogonal to each other. Its matched filter bank (MFB), which can simultaneously output the correlation between a received symbol and any sequence in the set, is effective for constructing communication systems flexibly. This paper discusses the compact design of the MFB of a polyphase sequence set, which can be applied to any sequence set generated by the given logic function. It is primarily focused on a ZCZ code with q-phase or more elements expressed as A(N=qn+s, M=qn-1, Zcz=qs(q-1)), where q, N, M and Zcz respectively denote, a positive integer, sequence period, family size, and a zero correlation zone, since the compact design of the MFB becomes difficult when Zcz is large. It is shown that the given logic function on the ring of integers modulo q generating the ZCZ code gives the matrix representation of the MFB that M-dimensional output vector can be represented by the product of the unitary matrix of order M and an M-dimensional input vector whose elements are written as the sum of elements of an N-dimensional input vector. Since the unitary matrix (complex Hadamard matrix) can be factorized into n-1 unitary matrices of order M with qM nonzero elements corresponding to fast unitary transform, a compact MFB with a minimum number of circuit elements can be designed. Its hardware complexity is reduced from O(MN) to O(qM log q M+N).

In this paper, we propose a filter-and-forward relay scheme with energy harvesting for single-carrier transmission in frequency-selective channels. The relay node harvests energy from both the source node transmit signal and its own transmit signal by self-energy recycling. The signal received by the relay node is filtered to suppress the inter-symbol interference and then forwarded to the destination node using the harvested energy. We consider a filter design method based on the signal-to-interference-plus-noise power ratio maximization, subject to a constraint that limits the relay transmit power. In addition, we provide a golden-section search based algorithm to optimize the power splitting ratio of the power splitting protocol. The simulation results show that filtering and self-energy recycling of the proposed scheme are effective in improving performance. It is also shown that the proposed scheme is useful even when only partial channel state information is available.

To solve the area and power problems in Finite Impulse Response (FIR) implementations, a faithfully truncated adder-based FIR design is presented in this paper for significant area and power savings while the predefined output accuracy can still be obtained. As a solution to the accuracy loss caused by truncated adders, a static error analysis on the utilization of truncated adders in FIRs was performed. According to the mathematical analysis, we show that, with a given accuracy constraint, the optimal truncated adder configuration for an area-power efficient FIR design can be effortlessly determined. Evaluation results on various FIR implementations by using the proposed faithfully truncated adder designs showed that up to 35.4% and 27.9% savings in area and power consumption can be achieved with less than 1 ulp accuracy loss for uniformly distributed random inputs. Moreover, as a case study for normally distributed signals, a fixed 6-tap FIR is implemented for electrocardiogram (ECG) signal filtering was implemented, in which even with the increased truncated bits up to 10, the mean absolute error (Ē) can be guaranteed to be less than 1 ulp while up to 29.7% and 25.3% savings in area and power can be obtained.

This paper proposes a computational complexity-reduced algorithm for an adaptive peak-to-average power ratio (PAPR) reduction method previously developed by members of our research group that uses the null space in a multiple-input multiple-output (MIMO) channel for MIMO-orthogonal frequency division multiplexing (OFDM) signals. The proposed algorithm is an extension of the peak cancellation (PC) signal-based method that has been mainly investigated for per-antenna PAPR reduction. This method adds the PC signal, which is designed so that the out-of-band radiation is removed/reduced, directly to the time-domain transmission signal at each antenna. The proposed method, referred to as PCCNC (PC with channel-null constraint), performs vector-level signal processing in the PC signal generation so that the PC signal is transmitted only to the null space in the MIMO channel. We investigate three methods to control the beamforming (BF) vector in the PC signal, which is a key factor in determining the achievable PAPR performance of the algorithm. Computer simulation results show that the proposed PCCNC achieves approximately the same throughput-vs.-PAPR performance as the previous method while dramatically reducing the required computational cost.

Based on the concept of polynomial operators, this paper explores generalized direct-form II structure and its state-space realization for two-dimensional separable-denominator digital filters of order (m, n) where a structure with 3(m+n)+mn+1 fixed parameters plus m+n free parameters is introduced and analyzed. An l2-scaling method utilizing different coupling coefficients at different branch nodes to avoid overflow is presented. Expressions of evaluating the roundoff noise for the filter structure as well as its state-space realization are derived and investigated. The availability of the m+n free parameters is shown to be beneficial as the roundoff noise measures can be minimized with respect to these free parameters by means of an exhaustive search over a set with finite number of candidate elements. The important role these parameters can play in the endeavors of roundoff noise reduction is demonstrated by numerical experiments.

Collaborative filtering (CF) is one of the most popular approaches to building Recommender systems (RS) and has been extensively implemented in many online applications. But it still suffers from the new user cold start problem that users have only a small number of items interaction or purchase records in the system, resulting in poor recommendation performance. Thus, we design a new similarity model which can fully utilize the limited rating information of cold users. We first construct a new metric, Popularity-Mean Squared Difference, considering the influence of popular items, average difference between two user's common ratings and non-numerical information of ratings. Moreover, the second new metric, Singularity-Difference, presents the deviation degree of favor to items between two users. It considers the distribution of the similarity degree of co-ratings between two users as weight to adjust the deviation degree. Finally, we take account of user's personal rating preferences through introducing the mean and variance of user ratings. Experiment results based on three real-life datasets of MovieLens, Epinions and Netflix demonstrate that the proposed model outperforms seven popular similarity methods in terms of MAE, precision, recall and F1-Measure under new user cold start condition.

In this letter, an active complex filter with finite transmission zeros is proposed. In order to obtain a complex prototype ladder filter including no inductors, a new circuit transformation is proposed. This circuit is classified into the RiCR filter. It is shown that it includes no negative capacitors when it is obtained through a frequency transformation. The validity of the proposed method is confirmed through computer simulation.

We propose the design method of a compact long-wavelength-pass filter implemented in a two-dimensional metal-dielectric-metal (MDM) waveguide with three stubs using a transmission line model based on a low-pass prototype filter, and present the wavelength characteristics for filters in an MDM waveguide based on 0.5- and 3.0-dB equal-ripple low-pass prototype filters.