Wireless Communication with high data rate applications attains greater significance recent years. The better quality of services in wireless network directs to massive energy utilization in wireless network. To attain better efficiency for resource distribution in cellular organization, a device-to-device (D2D) communication is considered as a vital factor. Several research technologies have been intended in conventional work to give the better D2D connection but, achieving maximum throughput is a still difficult issue. In addition, D2D communication expresses few confronts i.e. Energy usage and data delivery rate. Resource efficient device selection is the major area of center in D2D communication. Though, latency and loss rate involved during the communication was failed to be minimized. In order to overcome the above limitation, three proposed techniques are implemented in this research work as follows. Initially, a Bio-inspired Conic Optimized and Distributed Latency Q Learning (BCO-DLQL) is intended with the goal of attaining effective D2D communication in 5G. At the starting, a Bio-inspired Conic Particle Swarm Optimization (BCPSO) concept is implemented in this work with the aim of providing the computationally effective solutions for better energy usage in 5G. Subsequently, the Distributed Latency Managed Q Learning (DLMQL) concept is implemented with the objective of decreasing the latency is observed during the D2D communication and thereby obtaining better connectivity in network. Then, nearby device and communication range for the consequent continuous flow of information is also analyzed in order to diminish information loss.