EATSA: An MPTCP Scheduling Algorithm Based on Expected Arrival Time and Its Application in Bandwidth Compensation

The Internet of Vehicles (IoV) is developing rapidly, and user demands are increasing. Modern vehicular networks need high bandwidth, low latency, and stable links. A single network link often cannot meet these needs in complex mobile environments. The Multipath Transmission Control Protocol (MPTCP) can use multiple heterogeneous network links at the same time. It improves bandwidth utilization and communication robustness. In IoV heterogeneous networks, links still have very different bandwidths and delays. These differences reduce the performance of existing MPTCP scheduling algorithms. They cause low resource utilization and serious packet reordering. To solve these problems, this paper proposes the Expected Arrival Time-based Scheduling Algorithm (EATSA). EATSA analyzes the transmission characteristics of each link and estimates the expected arrival time of packets. It then allocates data to links based on these estimates and updates the allocation dynamically. This improves transmission efficiency and reduces packet reordering. We also design an intelligent bandwidth compensation framework based on MPTCP. The framework aggregates links from in-vehicle devices and user devices and adjusts bandwidth in real time. Experimental results show that EATSA improves network performance in complex heterogeneous environments and supports next-generation IoV communications.