Cluster-based pathloss exponential modeling for indoor positioning in wireless sensor network

GPS (Global Positioning System) cannot be used for localization object in indoor environments. Even though, estimation of localization is an important requirement in wireless sensor networks to determine the position of the object. Localization based on RSSI (Received Signal Strength Indicator) has low accuracy because of attenuation of the transmitted signal as a function of the distance between nodes. In indoor environment, the path loss exponent values are varied depending on the environment characteristic. This value is required to estimate the distance between nodes and it obtained from the measurement process. However, the fluctuating value based on measurement, it is strongly influencing the output of estimating distance based on RSSI. There for, we proposed a new model for cluster-based PLE in three conditions. After this measurement phase, the variation values of PLE using purposed method and linear regression can be used to calculate distance estimated from RSSI data. On the simulation result shows that cluster-based PLE model condition 1 (anchor node area) have the capability in decreasing MSE estimation distance up to 14.64 %, meanwhile MSE estimation distance PLE using regression liner has 23.55%, cluster-based PLE model condition 3 (cluster area) has 26.27 % and cluster-based PLE model condition 2 (sub cluster area) has 35.54 %.