The most recent modern communication and the next level of wireless transmission are the 5G/6G network. To achieve continuous packet transmission to a destination, a feature of Internet of Things (IoT) systems, 5G/6G in practice, uses the IoT to operate in high-traffic systems with numerous nodes/sensors. As a result, 5G/6G delivers enormous bandwidth, minimal delay, and high-speed data transfer rates. Therefore, using next-generation technologies, particularly the media access control protocol, is made possible and motivated by 5G. However, the typical congestion control actions have a cumulativeeffect on overall efficiency.A Deep Learning-based Congestion Control Algorithm (DL-CCA) is suggested in this article to enhance performance. The proposed system is generated using Non-Orthogonal Multiple Access (NOMA), Orthogonal Frequency Division Multiplexing (OFDM), and Orthogonal Quadrature Amplitude Multiplexing (OQAM). Current processes hamper the performance of 5G/6G and IoT.To forecast the best enhancement of collision control in the remote monitoring of 5G/6G IoT environments, andusinga deep learning system relying on a decision tree (DT) method is suggested in this paper. The model was applied to a training sample to find the ideal parametric configuration in a 5G/6G context. The information has been used to build deep learning and make it possible to identify the best alternatives that could improve the effectiveness of the proposed approach to network congestion. Forecasting and classification are two more tasks for which the DT technique uses. Any consumer could use charts that DT techniques produce to learn the prediction methodology. With a greater than 91.5 % recognition rate, the DT C4.5 had excellent results.