Learning and Batch-Processing Based Coded Computation With Mobility Awareness for Networked Airborne Computing
Learning and Batch-Processing Based Coded Computation With Mobility Awareness for Networked Airborne Computing
复制标题
DOI：
10.1109/tvt.2022.3231179
复制
发表时间：
2023-05
影响因子：
6.8
通讯作者：
Baoqian Wang;Junfei Xie;K. Lu;Yan Wan;Shengli Fu
中科院分区：
文献类型：
--
作者：
Baoqian Wang;Junfei Xie;K. Lu;Yan Wan;Shengli Fu
许多无人机（UAV）应用的实现（例如，火灾探测、监视和包裹递送）需要大量的计算资源来实现可靠的性能。现有的将计算任务转移到地面的解决方案可能会遭受长时间的通信延迟。为了解决这一问题，网络机载计算（NAC）是一种很有前途的技术，它通过直接飞行到飞行链路在无人机之间共享资源，提供先进的机载机载计算能力。然而，NAC还不存在，并且使其能够需要克服许多技术挑战，例如无人机的高机动性和不确定、异构和动态空域。本文通过1)开发基于动态批处理的编码计算（D-BPCC）框架来实现鲁棒性和适应性协同机载计算，以及2)设计基于深度强化学习（DRL）的负载分配和无人机移动控制策略来优化系统性能来解决这些挑战。作为第一个系统研究NAC的研究，据我们所知，我们通过设计NAC模拟器并与四种最先进的分布式计算方案进行比较研究来评估所提出的方法。结果表明，所提出的方法具有良好的性能。
The implementation of many Unmanned Aerial Vehicle (UAV) applications (e.g., fire detection, surveillance, and package delivery) requires extensive computing resources to achieve reliable performance. Existing solutions that offload computation tasks to the ground may suffer from long communication delays. To address this issue, the Networked Airborne Computing (NAC) is a promising technique, which offers advanced onboard airborne computing capabilities by sharing resources among the UAVs via direct flight-to-flight links. However, NAC does not exist yet and enabling it requires overcoming many technical challenges, such as the high UAV mobility, and the uncertain, heterogeneous, and dynamic airspace. This paper addresses these challenges by 1) developing a Dynamic Batch-Processing based Coded Computation (D-BPCC) framework for achieving robust and adaptable cooperative airborne computing, and 2) designing deep reinforcement learning (DRL) based load allocation and UAV mobility control strategies for optimizing the system performance. As the first study to systematically investigate NAC, to the best of our knowledge, we evaluate the proposed methods through designing a NAC simulator and conducting comparative studies with four state-of-the-art distributed computing schemes. The results demonstrate the promising performance of the proposed methods.