The Research And Implementation Of RSSI-Based Indoor Positioning Algorithm For Mobile AD HOC Network

As a brand new technology of information acquisition and processing, Mobile Ad Hoc Network has accepted broad attentions from domestic and foreign scholars and can be utilized in many fields, such as national defense, environmental monitoring, medical applications, space explore, anti-terrorism, and so on. There are a lot of special characteristics in an indoor environment, such as existing obstacles and corners, attenuation of signal, multipath fading, non-line-of-sight. Meanwhile, the activities of persons affect the signal's propagation. As a result, how to get the object's position in a complicated indoor environment is an urgent problem to be solved.To address the problem above, RBIP, which is based on an actual medical project of ZigBee for the Internet of things, is proposed in this paper after a deep study of technology such as Mobile Ad Hoc Network, ranging between nodes. We first improved the model of signal propagation to get RSSI measure model which is suitable for indoor environment. Then, the collect signals are filtered by our signal filter algorithm. After position calculation in lower machine, we correct the position by using RSSI similarity based method.On the basis of above work, we implement RBIP and take some experiments by using ZigBee CC2431ZDK of TI Company. We use the algorithm in an Infant Protection System and the experiments. The application under certain scenarios indicates that RBIP is feasible and effective without bring more latency, and have improved the accuracy of position.Contrasted with the traditional position algorithm, this paper has the following characteristics: (1) A dynamic mechanism, which is based parameter-tuning, for calculating parameters of the RSSI-distance model is proposed. There isn't a suitable RSSI-distance model in an indoor environment for the variability of temperature, moisture and electromagnetic effect. The RSSI-distance model in this paper calculates and adjusts parameters periodically.(2) A filtering algorithm which is based on constant velocity is proposed in this paper. For there are some effects of electromagnetic noise and non-line-of-sight, we preprocessed the signals by constant velocity based filtering after collecting them. The experiments shows that the preprocess process reduces errors brought from measurement.(3) We use the average value of signal from single node. It can avoid fluctuation of position result, which is produced by single node's several wrong measurements, by using mean method.(4) For there are some barriers and corners in the room of actual environment, we propose an RSSI-similarity based algorithm to correct the position results. By this way, we can handle the static obstacles of the room.