Abstract—The Internet of Things (IoT) is increasing and encompasses various areas such as smart homes, smart cars, and e-healthcare. Identifying the source of the transmitted data is important because information without a known source is meaningless. The running applications must be able to determine their position without relying on the Global Positioning System (GPS), as the signals are attenuated indoors and in difficult environments. Wireless Sensor Networks (WSNs) play an important role in positioning IoT devices. The Distance Vector-Hop (DV-Hop) algorithm can be utilized to localize unknown sensor nodes. DV-Hop is used due to its simplicity and low cost. It is used to locate a node several hops away from anchor nodes. However, the accuracy achieved is not satisfactory. On the other hand, the Received Signal Strength Indicator (RSSI)algorithm is employed to approximate the positions of the sensor nodes, but its effectiveness is limited to a single hop from the anchor nodes. In this paper, the Kalman Filter Multilayer Perceptron-Distance Vector Received Signal Strength Indicator (KF-MLP-DVRSSI) algorithm is presented to improve the accuracy and reliability of the DV-Hop algorithm without the need for additional hardware. The proposed algorithm adjusts the RSSI values of connections between one-hop neighbors using the Kalman Filter (KF). The Kalman filter predicts the variables and estimates the states of the future system based on the prior predictions. A Multilayer Perceptron (MLP) is then used to learn from the actual data and adjust the weights to produce accurate output data. The simulation results demonstrate the performance of the proposed approach compared to three existing models.