In recent years, task offloading and its scheduling optimization have emerged as widely discussed and significant topics. The multi-objective optimization problems inherent in this domain, particularly those related to resource allocation, have been extensively investigated. However, existing studies predominantly focus on matching suitable computational resources for task offloading requests, often overlooking the optimization of the task data transmission process. This inefficiency in data transmission leads to delays in the arrival of task data at computational nodes within the edge network, resulting in increased service times due to elevated network transmission latencies and idle computational resources. To address this gap, we propose an Asynchronous Data Transmission Policy (ADTP) for optimizing data transmission for task offloading in edge-computing enabled ultra-dense IoT. ADTP dynamically generates data transmission scheduling strategies by jointly considering task offloading decisions and the fluctuating operational states of edge computing-enabled IoT networks. In contrast to existing methods, the Deep Deterministic Policy Gradient (DDPG) based task data transmission scheduling module works asynchronously with the Deep Q-Network (DQN) based Virtual Machine (VM) selection module in ADTP. This significantly reduces the computational space required for the scheduling algorithm. The continuous dynamic adjustment of data transmission bandwidth ensures timely delivery of task data and optimal utilization of network bandwidth resources. This reduces the task completion time and minimizes the failure rate caused by timeouts. Moreover, the VM selection module only performs the next inference step when a new task arrives or when a task finishes its computation. As a result, the wastage of computational resources is further reduced. The simulation results indicate that the proposed ADTP reduced average data transmission delay and service time by 7.11% and 8.09%, respectively. Furthermore, the task failure rate due to network congestion decreased by 68.73%.

The rapid development of Internet of Things (IoT) technology has led to a significant increase in the computational task load of Terminal Devices (TDs). TDs reduce response latency and energy consumption with the support of task-offloading in Multi-access Edge Computing (MEC). However, existing task-offloading optimization methods typically assume that MEC’s computing resources are unlimited, and there is a lack of research on the optimization of task-offloading when MEC resources are exhausted. In addition, existing solutions only decide whether to accept the offloaded task request based on the single decision result of the current time slot, but lack support for multiple retry in subsequent time slots. It is resulting in TD missing potential offloading opportunities in the future. To fill this gap, we propose a Two-Stage Offloading Decision-making Framework (TSODF) with request holding and dynamic eviction. Long Short-Term Memory (LSTM)-based task-offloading request prediction and MEC resource release estimation are integrated to infer the probability of a request being accepted in the subsequent time slot. The framework learns optimized decision-making experiences continuously to increase the success rate of task offloading based on deep learning technology. Simulation results show that TSODF reduces total TD’s energy consumption and delay for task execution and improves task offloading rate and system resource utilization compared to the benchmark method.