Masaki Takanashi

Header reduction to increase the throughput in decentralized TDMA-based vehicular networks

In the decentralized-TDMA (D-TDMA) protocol the terminals select a free slot based on the frame information (FI) included in each packet to be transmitted. However, the FI constitutes a large portion of the packet which seriously compromised the throughput of the D-TDMA. We are proposing a technique for increasing the throughput by reducing the length of the FI without jeopardizing the packet success rate. Our technique, which is based on FI redundancies, lets the terminals transmit the relevant information when necessary. A computer simulation was conducted for an urban environment in which vehicles are moving. The simulation results show that the proposed technique significantly increases the throughput without degrading the quality of the D-TDMA protocol.

Channel measurement and modeling of high-voltage power line communication in a hybrid vehicle

Recently, as hybrid vehicles (HVs) and electric vehicles (EVs) have become widespread, a large number of electronic devices have been incorporated into the vehicles. Thus, there is concern regarding the increasing number of lines and connectors. In HVs and EVs, a high-voltage (200 V) battery is employed. We have studied power line communication (PLC) that uses a high-voltage power line to address this concern. This line comprises two parallel wires (a positive (P) wire and a negative (N) wire) and a shielded sheath. Therefore, noise emission and noise immunity, which are significant problems for vehicle PLC using a 12 V power supply line, can be improved by using this high-voltage power line. In this paper, we clarify the transmission channel between a power control unit (PCU) and the high-voltage battery in a commercial HV through experiment and simulation. We also clarify the carrier frequency band that is suitable for high-voltage PLC.

Linear response analysis of vibrational resonance in over-damped systems

The phenomenon whereby the response to a weak low-frequency input signal is maximized in an oscillatory system with additional high-frequency input signals of appropriate amplitude is known as vibrational resonance (VR). In this study, the occurrence of VR in over-damped systems is analytically and numerically explained in the framework of linear response theory. The origin of the large response of VR is shown to be the dynamical bifurcation caused by tuning the input signals.

High-voltage power line communication system for hybrid vehicle

Recently, hybrid vehicles (HVs) have become widespread. In the current HVs, a high-voltage (200 V) battery supplies power to the Power Control Unit (PCU) for driving a motor, generator and air conditioner compressor. It is expected that such kinds of power electronic devices will increase in number in the future. We previously proposed a power line communication (PLC) that uses a high-voltage power line. In this paper, we present the development of a high-voltage PLC system that can simultaneously control multiple power electronic devices in real time and demonstrate the system operation. First, in a system of a commercially available HV, the results of transmission characteristics and electromagnetic noise characteristics that are superimposed on the high-voltage power line are shown. Then the bit error rate (BER) performance under the above noise environment is clarified by computer simulations. Finally, we construct a prototype of the high-voltage PLC system and demonstrate that the system can simultaneously control the rotational speed of two motors.

Reduction of frequency shift error on periodical noise suppression system

The artificially generated noise from the appliances has a periodicity in frequency domain. A noise suppression system, which is based on the PI(Power Inversion) algorithm and using frequency shift, has already been proposed. However, the system has a significant problem that the suppression effect can not be obtained when the frequency shift is not accurate. In this paper, reduction method of frequency shift error is proposed. Numerical results show that frequency shift error is reduced by the proposed method when the number of snapshots and the interval of sampling are set appropriately.

Compressive sensing based image transmission for multiuser V2X communications

For the practical implementation of the compressive sensing (CS) it is very important to achieve a good balance between the system stability and the processing simplicity. This paper proposes simple and robust measurement matrices for CS aiming for V2X image transmission. To introduce the random factors into array codes, which is known error correcting codes, we can achieve both simplicity and robustness. Furthermore, because the proposed measurement matrices can be easily generated, they are easily applied to multiuser system without retransmission control. The reconstruction stability of the proposed method is quite high even in case of small row weight. This leads to faster reconstruction compared to the conventional low density matrix approaches.

A sampling frequency and observation time analysis for efficient stochastic resonance in digital signal processing

The noise-related phenomenon, stochastic resonance (SR), may improve the performance of signal detection devices. In digital signal processing, the effect of SR depends on the observation time and sampling frequency. To apply it to digital signal processing, these two parameters must be determined. This paper investigates the observation time and sampling frequency for efficient SR in a digital system.

Deterministic fluctuation-response relation

Efficient detection of a weak signal in a noisy environment is one of the most intensely studied topics in signal processing. In such research, the concept of a noise-induced giant response, such as occurs in stochastic resonance, has attracted a great deal of attention. In the present study, by analyzing a system subjected to a deterministic fluctuation, we show that the occurrence of a giant response is strongly related to instability in the system. We derive the response function of the system, and obtain its eigenfunctions which determine the response behavior of the system. We find that the maximum eigenvalue of the response function diverges when instability arises in the system, and the infinite eigenvalue gives rise to a giant response. Our approach is analogous to the well-established fluctuation-response theory, and can be easily extended to a system with purely stochastic fluctuations.

Channel Tracking with Zero Padding Scheme for STBC-OFDM System under Fast Fading Environment

An inter-vehicle communication system is highly advantageous for driving safety support systems, such as collision avoidance systems at intersections. In an inter-vehicle communication system, a channel-tracking scheme is essential to achieve reliable wireless communication under the fast fading environment caused by surrounding vehicles. For coherent detection, decision feedback is a well-known channel-tracking scheme. In this scheme, replica signals are generated and channel estimation is performed by using the replica signals. However, because channel estimation performance depends on the signal-to-noise (SNR) at symbol time, decision feedback errors occur by the lower SNR caused by fast fading. The feedback errors degrade the transmission performance significantly. In an orthogonal frequency division multiplexing (OFDM) system, the time domain 0-padding scheme is a well-known channel estimation technique. Although better channel tracking performance is expected by applying the 0-padding scheme, the tracking performance is degraded when employing guard-bands in the OFDM system. In this paper, we propose an improved time domain 0-padding scheme for an STBC-OFDM system employing guard-bands based on the IEEE802.11p standard for inter-vehicle communication, and show the improved performance through computer simulations.