The knowledge of the future capacity variations in wireless networks using smartphones becomes more and more possible by exploiting the rich contextual information from smartphone sensors through mobile applications and services. It is entirely likely that such contextual information, which may include the traffic, mobility and radio conditions, could lead to a novel agile resource management not yet thought of. Inspired by the attractive features and potential advantages of this agile resource management, several approaches have been proposed during the last period. However, agile resource management also comes with its own challenges, and there are significant technical issues that still need to be addressed for successful rollout and operation of this technique. In this paper, we propose an approach (called NEWCAST) for anticipating throughput variation for mobile video streaming services. The solution of the optimization problem realizes a fundamental trade-off among critical metrics that impact the user's perceptual quality of the experience (QoE) and system utilization. Both simulated and real-world traces collected from [1] are carried out to evaluate the performance of NEWCAST. In particular, it is shown that NEWCAST provides the efficiency, computational complexity and robustness that the new 5G architectures require.