A significant study area in cloud computing that still requires attention is how to distribute the workload among virtual machines and resources. Main goal of this research is to develop an efficient cloud load balancing approach, improve response time, decrease readiness time, maximise source utilisation, and decrease activity rejection time. This research propose novel technique in load balancing based network optimization using routing protocol for 5G wireless communication networks. the network load balancing has been carried out using cloud based software defined multi-objective optimization routing protocol. then the network security has been enhanced by data classification utilizing deep belief Boltzmann NN. Experimental analysis has been carried out based on load balancing and security data classification in terms of throughput, packet delivery ratio, energy efficiency, latency, accuracy, precision, recall.