An information fusion-based multiobjective security system with a multiple-input/single-output sensor

In the framework of sensor fusion, multiple sensors corresponding to the number of physical variables that must be measured are used. In this paper, we propose a novel sensing approach that simultaneously deals with heterogeneous physical variables with a sensor. It is fundamentally different from sensor fusion. The proposed approach takes into consideration the fact that any sensor that detects a certain physical variable is influenced to a degree by other physical variables, which are designated as noise. The objective in conventional sensor design has been the minimization of noise. In contrast, the proposed approach takes advantage of sensors that are easily influenced by many physical variables and makes full use of the multisensing characteristics of these sensors. The system designed using this concept has advantages in terms of cost performance and system simplification compared to existing approaches. This concept can be realized by developing a novel multiple-input/single-output sensor that can detect various variables, including pressure, acceleration, temperature and incandescent light emission, by a single device. We apply the sensor to monitor the symptoms of fire, earthquakes, and break-ins for the purpose of home security. The proposed security system is realized through statistical signal processing and machine learning techniques.