In general, rich media applications demand high bandwidth to transfer large files over wide-area connections. The arrival of 5G will overcome the limitations of existing networks in support of these network-intensive applications because they provide high bandwidth reaching up to multiple Gbps for large file transfer. One of the main challenges to maximize and stabilize goodput is explicitly managing the randomness inherent in high-speed mobile networks. The AIMD-based TCP cannot make full utilization of bandwidth over links with high Bandwidth Delay Product. The inefficiency of transport control protocol can not satisfy the media applications transport requirements even over a high bandwidth end-to-end connection. We conducted an extensive analytical study of the design and implementation issues of high-speed data transfer methods, especially the impact of application-level receive buffer and the background workloads on the data transfer performance. We developed an optimized large file transfer solution based on this analysis to achieve the maximal bottleneck goodput. We showed that the proposed transport protocol achieves better performance compared to state-of-the-art transport methods.