A neural network approach towards reinforcing smart home security

In today's context, it is common to leave the house unattended as people are busy catching up with their tight daily schedule. Therefore, most people have chosen the home security system as the most reliable way to protect their home. However, the existing security mechanism provided by smart home is lack of intelligence for higher level decision making and action taking. Furthermore, the current security management of smart home is usually predefined by user. Thus, for consequences that are out of the defined scopes, chances to trigger false alarm are very high. Besides, the environment of each smart home is difference and the culture is also different from country to country, so the predefined security instructions may not work well or they may not suitable for all scenarios. Therefore, an intelligent algorithm with self learning ability must be adopted for smart home system in order to take adequate actions against the evaluated environment and decide for the necessary actions. Neural networks have been proposed in this paper where they are adaptive statistical models based on an analogy with the structure of the brain. The intelligence of the smart home should be able to analyze the information gathered by the sensors and detectors and to respond based on the analysis and the experiences gained via the neural networks. Neural network is chosen because they have the learning ability to estimate the parameters of some populations using a small number of examples at a time. When come to neural network learning process, there are several learning algorithms that could be applied, which can learn through pattern recognition, movement recognition, action recognition and so forth.