This paper presents the design and implementation of a novel electromyography (EMG) device capable of wirelessly transmitting data via a 5 GHz Wi-Fi network. A one channel electronic device and sensor containing three electrodes was developed and produced to capture EMG signals from hand gestures of open and closed positions of the participants. These signals were real-time transmitted through wireless network connection of two RTL8720DN modules to contribute to the creation of a custom dataset for classification purposes. The model of recurrent Trend Predictive Neural Network (rTPNN), which strengths the temporal dynamics with a deep learning classifier, was employed for gesture classification. Experimental results demonstrate the success of the wireless transmission system, which uses a total support of almost 10,000 to classify hand gestures on the test set, giving an accuracy of around 89% and an F1 score of around 88%. These results suggest that the proposed system is both efficient in real-time data transmission with 5 GHz band and effective in accurate gesture recognition, offering a potential tool for various applications, such as prosthetics control and rehabilitation.