SmartDet: Context-Aware Dynamic Control of Edge Task Offloading for Mobile Object Detection
SmartDet: Context-Aware Dynamic Control of Edge Task Offloading for Mobile Object Detection
复制标题
DOI：
10.1109/wowmom54355.2022.00034
复制
发表时间：
2022-01
期刊：
影响因子：
--
通讯作者：
Davide Callegaro;Francesco Restuccia;M. Levorato
中科院分区：
文献类型：
--
作者：
Davide Callegaro;Francesco Restuccia;M. Levorato
无人机和自动驾驶汽车等移动设备越来越依赖通过深度神经网络 (DNN) 进行的对象检测 (OD) 来执行导航、目标跟踪和监视等关键任务。由于其高度复杂性，这些 DNN 的执行需要大量的时间和精力。因此，低复杂度对象跟踪 (OT) 与 OD 一起使用，其中定期应用 OD 来生成用于跟踪的“新”参考。但OD处理后的帧延迟较大，不符合实时应用的要求。将 OD 卸载到边缘服务器可以缓解这个问题，但现有的工作重点是无线通道容量非常大的系统中卸载过程的优化。在这里，我们考虑通道容量受限且不稳定的系统，并建立能够适应大 OD 延迟的并行 OT（在移动设备）和 OD（在边缘服务器）流程。我们提出了 Katch-Up，一种新颖的跟踪机制，可以提高系统对过度 OD 延迟的恢复能力。我们表明，这项技术极大地提高了可用于跟踪的参考的质量，并将性能提高了 33%。然而，Katch-Up 在显着提高性能的同时，也增加了移动设备的计算负载。因此，我们设计了 SmartDet，这是一种基于深度强化学习 (DRL) 的低复杂度控制器，可以学习在资源利用率和 OD 性能之间实现正确的权衡。 SmartDet 将与当前视频内容和当前网络状况相关的高度异构上下文相关信息作为输入，以优化 OD 卸载的频率和类型以及 Katch-Up 利用率。我们在由 JetSon Nano 作为移动设备和 GTX 980 Ti 作为边缘服务器组成的真实测试平台上对 SmartDet 进行了广泛评估，通过 Wi-Fi 链路连接，以收集多个与网络相关的跟踪以及能量测量结果。我们考虑最先进的视频数据集 (ILSVRC 2015 - VID) 和最先进的 OD 模型 (EfficientDet 0、2 和 4)。实验结果表明，SmartDet 在跟踪性能——平均召回率（mAR）和资源使用之间实现了最佳平衡。相对于完全 Katch-Up 使用和最大信道使用的基准，我们仍然将 mAR 增加 4%，同时使用与 Katch-Up 相关的 50% 的信道和 30% 的电力资源。相对于使用最少资源的固定策略，我们在 1/3 帧上使用 Katch-Up 时将 mAR 提高了 20%。
Mobile devices such as drones and autonomous vehicles increasingly rely on object detection (OD) through deep neural networks (DNNs) to perform critical tasks such as navigation, target-tracking and surveillance, just to name a few. Due to their high complexity, the execution of these DNNs requires excessive time and energy. Low-complexity object tracking (OT) is thus used along with OD, where the latter is periodically applied to generate "fresh" references for tracking. However, the frames processed with OD incur large delays, which does not comply with real-time applications requirements. Offloading OD to edge servers can mitigate this issue, but existing work focuses on the optimization of the offloading process in systems where the wireless channel has a very large capacity. Herein, we consider systems with constrained and erratic channel capacity, and establish parallel OT (at the mobile device) and OD (at the edge server) processes that are resilient to large OD latency. We propose Katch-Up, a novel tracking mechanism that improves the system resilience to excessive OD delay. We show that this technique greatly improves the quality of the reference available to tracking, and boosts performance up to 33%. However, while Katch-Up significantly improves performance, it also increases the computing load of the mobile device. Hence, we design SmartDet, a low-complexity controller based on deep reinforcement learning (DRL) that learns to achieve the right trade-off between resource utilization and OD performance. SmartDet takes as input highly-heterogeneous context-related information related to the current video content and the current network conditions to optimize frequency and type of OD offloading, as well as Katch-Up utilization. We extensively evaluate SmartDet on a real-world testbed composed by a JetSon Nano as mobile device and a GTX 980 Ti as edge server, connected through a Wi-Fi link, to collect several network-related traces, as well as energy measurements. We consider a state-of-the-art video dataset (ILSVRC 2015 - VID) and state-of-the-art OD models (EfficientDet 0, 2 and 4). Experimental results show that SmartDet achieves an optimal balance between tracking performance – mean Average Recall (mAR) and resource usage. With respect to a baseline with full Katch-Up usage and maximum channel usage, we still increase mAR by 4% while using 50% less of the channel and 30% power resources associated with Katch-Up. With respect to a fixed strategy using minimal resources, we increase mAR by 20% while using Katch-Up on 1/3 of the frames.