Wireless communication aims at providing a reliable and high quality communication. The evolution of cellular communication has been a step in the right direction from 1979 to date. However, each generation of cellular network has some requirements like delay, throughput and QoS that must be considered to provide an effective communication. The growth in the consumption of mobile services has resulted to an overload in the cellular networks. This has opened up challenges for resource management in future mobile networks. Therefore, there is need for an effective resource scheduling and sharing schemes to cope with the available bandwidth. This paper provides a review of the different generations of networks to date with reference to their efficient communication resource sharing and scheduling schemes. We will explore possible deployment of effective data driven AI and machine learning algorithms for Radio Access Network (RAN) slicing, which provides reduced latency and an overload reduction in 5G networks. We will also proffer solutions to the new problems encountered by the 5G RAN slices.