In this paper, we report a computational autonomous design of a 24 GHz one-stage microwave amplifier, which is a key device for high-speed communication. A novel fitness function for particle swarm optimization (PSO) is introduced and a microwave amplifier is optimized by considering stability, center frequency, gain, and bandwidth. Moreover, we consider the parameters of PSO such as inertia coefficient, the number of particles, and topology to ensure immunity against a local optimum and accelerate the convergence. As a result, a 12 mW, 24 GHz amplifier, which has 10.8 dB gain and 3.8 GHz bandwidth, is designed autonomously in an hour.

View full abstract

This paper focuses on room access management based on the received signal strength indicator (RSSI) at two different points using two monitoring devices from smartphones. A server with a wireless LAN access point (AP) continuously sends an echo request packet to every smartphone connected to the AP, and the RSSI of the echo reply packet from the smartphones is monitored by two monitoring devices at two different points. The RSSI characteristics are that the RSSI becomes higher as the user approaches the monitoring devices and lower as the move away from them. Applying machine learning using the RSSI characteristics, we estimate the room access information of users concerning entering, staying in, or leaving the room. The proposed method does not require any special application software and user operations. Because the RSSI is monitored at two different points, our proposed method can handle various user behaviors. As a result, our proposed method achieves a high estimation accuracy of 94.44%.

View full abstract

Ad hoc on-demand distance vector (AODV) is an often used routing protocol in multihop communication. In AODV, there is a tendency for the first established route to be repeatedly used. This causes a problem that the remaining battery levels become uneven between the nodes used in the communication route and the other nodes. In this paper, we propose a simple route reconstruction method to ease the problem. The proposed method makes use of the sequence number that is used in the conventional AODV. We conduct a fundamental investigation into the lifetime of the first-dead node using a network simulator. The result shows that the proposed method with multiple thresholds can prolong the lifetime of the first-dead node by 11%. This means that the proposed method can contribute toward easing the problem of uneven battery consumption among the nodes.

View full abstract

In this paper, a spectrum-sensing method is proposed which exploits the combination of a comb filter and autocorrelation function. The proposed method improves the detection performance in severe noise environments, where a cyclic prefix (CP)-based orthogonal frequency division multiplexing (OFDM)-transmitted signal is considered. In our proposed method, the primary user information is not required for spectrum sensing. The detection performance is measured over additive white Gaussian noise (AWGN) and multipath Rayleigh fading channels for different digital modulations. A comprehensive evaluation by simulation shows that the proposed method significantly outperforms the conventional schemes at a low signal-to-noise ratio (SNR).

View full abstract

This paper proposes an approach to pitch-synchronous linear prediction (LP) for bone-conducted (BC) voiced speech. A combination of the spectral compensation (SC) method with pitch extension LP is used to obtain a more accurate power spectrum of the BC speech signal. Simulation experiments show that the proposed method provides better performance than the conventional autocorrelation and original SC methods.

View full abstract

A simple and effective denoising method for a spectral subtractive (SS)-type parametric Wiener filter (PWF) for a blind condition is proposed. A simple noise estimation method is used to estimate the noise variance directly from a noisy image. Preliminary experiments with trained images are conducted to find the best parameters for the PWF. The PWF gives the highest performance with the best parameter setting. However, in practice, it is difficult to know the best parameters because they depend on the characteristics of the image. To estimate the best parameters for the PWF, therefore, a novel tool named image power spectrum sparsity, which is not influenced by the noise level, is derived. The parameters for the PWF are set according to the power spectrum sparsity. To demonstrate the effectiveness of the PWF, untrained images are used. The experimental results show that the proposed method gives a good performance with the shortest computational time among the WF methods to restore an image under a blind condition.

View full abstract

This paper proposes texture-based features for clothing category classification based on graph representation. Recently, graph-based representation has been used for texture characterization to aid texture analysis. In this work, graph-based theory is applied to characterize the local image structure. The rotation invariance uniformity (riu2) of local binary pattern mapping is adopted to represent feature descriptors. The proposed approach is evaluated by using the Brodatz and UIUC standard texture databases, and a clothing dataset. The proposed method is shown to be more effective for clothing classification than other methods.

View full abstract

In this paper, we first present a theorem on the relationship between the traditional model and a model based on an orthonormal system in experimental design. Using the theorem, the former model can be converted to the latter, and vice versa. Next, we introduce prior distributions over the hyperparameters for experimental design models to consider fully Bayesian predictions. Combining the conversion and a previous result, we show that we can make an approximation in which we set the hyperparameters to specific values determined by maximizing the evidence function.

View full abstract

The quantum-inspired evolutionary algorithm (QEA) and QEA with a pair-swap strategy (QEAPS) have quantum-inspired individuals, where each gene is represented by a quantum bit (qubit). QEA and QEAPS iterate the evolution using the unitary transformation of probability amplitudes in each qubit, and can automatically shift the evolution from a global search to a local search. Here, the convergence speed depends on the rotation angle of each qubit toward either the |0〉 or |1〉 state vector, and the probability amplitude diverges or the population is likely to fall into a local solution if the rotation angle is too large. If the rotation angle is too small, the convergence speed may become low and the search performance will degrade. In this study, we introduce nonuniform convergence speeds into the quantum-inspired individuals and regard the convergence speed as the individual feature (individuality) of each quantum-inspired individual. Introducing the proposed individuality can eliminate the cumbersome process required to design a rotation angle while ensuring the quality of the obtained solution.

View full abstract

The shepherding problem is to control and guide a flock of multiple autonomous agents by means of one or more external controllable agents. A critical example is a sheepdog/shepherd guiding a flock of sheep/agents and herding them to a target destination. Solving this problem is expected to lead to the development of robots for herding livestock and for guiding people who need evacuation. Strömbom et al. modeled a shepherd's behavior mathematically, in which a single shepherd can herd a flock of agents to a target. They evaluated the performance of the herding algorithm (HA) with a constant difference between the shepherd's and agents' moving speeds. In this study, we evaluate simulated HA proposed by Strömbom et al. for various differences between the shepherd's and agents' moving speeds caused by the agents' stride length and analyze the robustness of this algorithm regarding the agents' moving speeds. The experimental results show that the success of herding is mostly guaranteed when the moving speeds of all the agents in the flock are lower than the shepherd's moving speed. Also, the results show that the herding succeeds even if the flock consists of agents having various moving speeds. From these results, we have clarified that HA with a single shepherd is mostly robust regarding agents' moving speeds.

View full abstract

We investigate the fundamental tradeoff among the identification, secrecy, and compression rates under the condition that the prior distribution of an identified individual is unknown. We combine the techniques developed by Tuncel (2009) and Ignatenko and Willems (2010) to analyze the capacity region of the three rate tuples mentioned above. We show that when the secrecy rate is zero, the capacity region of the identification and compression rates is characterized as a form similar to the one given by Tuncel.

View full abstract