چكيده به لاتين
localization of devices and users outdoors using satellites on a global scale, But in the building and in some open environments, including the forests, satellite waves are severely damaged and its effectiveness decreases. Because of the weakening and dispersion of satellite waves in contact with the ceiling and walls of the building, it is not possible to use GPS to localization the user inside the building. Therefore, only outside use of the GPS system is used, which is why indoor positioning systems (IPS) were introduced to address this problem. By the way, indoor positioning system using the Received signal strength (RSS), also called a fingerprint, is more popular among locating methods due to lower cost, infrastructure and accuracy.
But there are large amount of fingerprints in the database whose processing and optimal use for positioning is a challenge. Data mining and classification are an attempt to obtain useful information from these data. In this research, there are some ways to reduce the amount of data collected (fingerprinting), which results in faster positioning with and sutable accuracy, called the CMAR algorithm, which uses a reinforced FP-TREE, in which the distribution of the label Classes in the middle of the tuples that cover the collection of repetitive items. In this way, the algorithm is able to double-examine the set of repeated items and generate rules in one step. We tried to simulate the research in MATLAB software and, by giving several different values to the support and confidence indexes and generating association rules, we implemented the of time and cost of positioning with sutable errors for several times and localization them with clustering algorithms to compare.
