The Internet of Things is defined by two specific words: data security and resource allocation (IoT). The fast proliferation of 5G technology necessitates the development of novel network resource management methods. Through connecting all of the devices, this current technological advancement has had an impact on industrial uses, making the network highly flexible and computation efficient. The architecture of Industrial IoT, as a subset of IoT, is built on a large number of nodes with a continual flow of several functions at once. As a result, in a multi-objective network, interference in the path always causes a loss of network resources and makes data security susceptible. Reinforcement and Existing Heuristics Learning-based methods lack generalizability and flexibility, making them inadequate for addressing this issue. Recently, data-driven Deep Learning (DL) enabled techniques to achieve optimal performance with low computing complexity have been created. Future intelligent systems are likely to use Long- and Short- Term Memory (LSTM). Proposed techniques for radio resource allocation in 5G networks are discussed in this paper. We examine the performance of the suggested learning approach and compare it to other techniques. The examination demonstrates the principles' wide applications. As a consequence, a fast system with optimal resource use emerges, which is confirmed by mathematical analysis and simulations