Resource Aware Quadratic Discriminative Gentle Steepest Boost Classification for D2D Communication in 5G

Abstract

D2D communication for 5G networks enables mobile devices to communicate directly with other devices. To effectively utilize available resources, 5G cellular networks that enable high-speed network communication are the major challenging issues. A novel Resource Aware Quadratic discriminative gentle steepest boost classification (RAQDGDBC) technique for D2D communications is introduced. For each device in the 5G network, energy, bandwidth, and connection speed is measured to improve the continuous flow of data and minimize data loss and latency during the communication. The RAQDGDBC technique uses the ensemble method called gentle adaptive steepest boost classification technique to identify the higher bandwidth, energy-efficient, and speed aware devices by using the set of the weak learner, i.e., Quadratic discriminative classifier. Weak learner measures the bandwidth, energy, and connection speed of the device with the threshold value. An optimal node is selected for efficient communication based on the likelihood measure. The ensemble method offers accurate outcomes with minimum error with the help of the steepest descent function. Therefore, the device with higher bandwidth, energy efficiency, and connection speed is chosen for data communication. These advanced features of the 5G network provide better data transmission and reduce loss. A comparison of RAQDGDBC and existing techniques indicates that the RAQDGDBC technique achieves a higher Data delivery rate (DDR), throughput, energy efficiency, and lesser Data loss rate (DLR) latency than the conventional methods.