A Learning-Based Fast Uplink Grant for Massive IoT via Support Vector Machines and Long Short-Term Memory

The current random access (RA) allocation techniques suffer from congestion and high signaling overhead while serving massive machine-type communication (mMTC) applications. To this end, third-generation partnership project introduced the need to use fast uplink grant (FUG) in order reduce latency increase reliability for smart Internet of Things (IoT) applications with strict Quality-of-Service constraints. We propose a novel FUG based on support vector machine (SVM). First, (MTC) devices are prioritized using an SVM classifier. Second, long short-term memory architecture is used traffic prediction correction overcome errors. Both results achieve efficient resource scheduler terms average total throughput. A coupled Markov modulated Poisson process (CMMPP) model mixed alarm regular applied compare proposed other existing techniques. In addition, extended model-based CMMPP evaluate algorithm more dense network. test scheme real-time measurement data collected Numenta anomaly benchmark (NAB) database. Our simulation show outperforms RA schemes by achieving highest throughput lowest delay 1 ms accuracy 98 % when target critical MTC limited number resources.