Emerging network applications, such as virtual reality technology, autonomous driving, and large-scale Internet of Things, usually require a significant amount of bandwidth and have high requirements for data delivery time. However, the existing Quick UDP Internet Connection (QUIC), TCP, and other protocols cannot fully meet these needs. A scheduling algorithm based on a time-sensitive protocol can meet the above requirements at the transport layer and reduce the development difficulty of the application layer. Therefore, this paper is based on Deadline-aware Transport Protocol (DTP) and implemented in a Python 3.0 environment. It focuses on the scenario where video data blocks are sent to the transport layer. In this paper, we study the decision-making method and congestion control method for data block priority and deadline, and optimize the delay-sensitive transmission protocol. We propose a scheduling algorithm to meet the service's delay demands and enhance the overall service quality. Additionally, we introduce an improved scheduling scheme for network scenarios involving multiple senders or receivers. Our research aims to effectively address the service delay requirements under different network conditions.