Wi-CL: Low-Cost WiFi-Based Detection System for Nonmotorized Traffic Travel Mode Classification

Traffic travel mode identification and classification are crucial for the development of intelligent transportation systems (ITSs). At present, scholars have investigated motorized nonmotorized traffic in various road environments; however, walking bicycle modes has been largely ignored. Therefore, this paper, we investigate propose a new low-cost system, known as Wi-Fi (Wi-CL) system that uses signal detectors refined characteristics modes. The Wi-CL includes four modules: data acquisition module, processing feature extraction module. In proposed detects signals participants environments. addition, received strength indicator (RSSI) filtering algorithm hybrid networks effectively addresses surrounding obstacles environmental noise. extract relevant features to construct model. Finally, recurrent neural network (RNN) framework based on long short-term memory (LSTM) is successfully implemented module identification. To validate effectiveness extensive experiments were conducted using field collected by installed at South China University Technology (SCUT). experimental results show RSSI achieves excellent real constructed speed estimation outperforms other baseline models different scenarios (flat-peak walking, midday peak flat-peak cycling, cycling), achieving an overall accuracy 97.92%. summary, our feasible approach classification.