Research And Mplementation Of High Performance Of Indoor And Outdoor Scenario Identification Algorithm Using The Embedded Sensors In The Smartphones

With the rapid development of intelligent terminals and wireless communication technology, the location-based services are becoming an additional wireless service. The location and context switching, especially the indoor and outdoor scenario switching, provide basic and original information for various mobile applications. Diverse smart phone placements and limited battery supply pose challenge for the accurate and robust indoor-outdoor identification.In this paper, a pervasive indoor and outdoor scenario identification algorithm is proposed, which uses the basic and lightweight components on the Android platform to determine the indoor and outdoor environment. We require it to be more stable, accurate and efficient during the detection. Lightweight components, that is, the low energy consumption sensor resources on smart phones, including light sensors, magnetic sensors, pressure sensors, acceleration sensors, gyroscope sensors, etc. We do not require a prior knowledge of the environment, and only use common sensors on mainstream mobile phones. The system makes use of characteristics of each sensor to construct various of sub-detection-modules, and the detection result of each sub-module is fused by two strategies, one is Bayes voting strategy, which combines the confidence of each sub-modules to get the final result. In order to improve the detection accuracy of indoor-outdoor transition scenario, we introduce the GPGSV (the information of satellite number in the GPS protocol) as a correction factor for the final result. This module is called when necessary due to huge energy consumption and stay closed at ordinary times. Further more, we use Adaboost classifier to distinguish indoor and outdoor by training the raw sensor data. The two strategies perform stateless result. The stateless detection results are further used as the observation sequence of a hidden Markov model (HMM) to obtain the final stateful results. The adoption of the HMM filter can effectively eliminate the occasional noises and improve detection accuracy.Extensive experimental results confirm that the proposed pervasive indoor-outdoor identification algorithm outperforms the state-of-the-art IODetector with more than 95% detection accuracy and less than 5mw power consumption under various weather condition and smart phone placements, and thus the validity of this algorithm has been verified.