Accurate navigation is of paramount importance to ensure flight safety and
efficiency for autonomous drones. Recent research starts to use Deep Neural
Networks to enhance drone navigation given their remarkable predictive
capability for visual perception. However, existing solutions either run DNN
inference tasks on drones in situ, impeded by the limited onboard resource, or
offload the computation to external servers which may incur large network
latency. Few works consider jointly optimizing the offloading decisions along
with image transmission configurations and adapting them on the fly. In this
paper, we propose A3D, an edge server assisted drone navigation framework that
can dynamically adjust task execution location, input resolution, and image
compression ratio in order to achieve low inference latency, high prediction
accuracy, and long flight distances. Specifically, we first augment
state-of-the-art convolutional neural networks for drone navigation and define
a novel metric called Quality of Navigation as our optimization objective which
can effectively capture the above goals. We then design a deep reinforcement
learning based neural scheduler at the drone side for which an information
encoder is devised to reshape the state features and thus improve its learning
ability. To further support simultaneous multi-drone serving, we extend the
edge server design by developing a network-aware resource allocation algorithm,
which allows provisioning containerized resources aligned with drones' demand.
We finally implement a proof-of-concept prototype with realistic devices and
validate its performance in a real-world campus scene, as well as a simulation
environment for thorough evaluation upon AirSim. Extensive experimental results
show that A3D can reduce end-to-end latency by 28.06% and extend the flight
distance by up to 27.28% compared with non-adaptive solutions.

利用深度神经网络改进自主无人机导航，提出 A3D 边缘服务器辅助的无人机导航框架，可动态调整任务执行位置、输入分辨率和图像压缩比，以实现低推理延迟、高预测准确性和长飞行距离。证实 A3D 相比非自适应方案可将端到端延迟减少 28.06％，飞行距离延长高达 27.28％。

A3D：边缘辅助飞行器的自适应、准确、自主导航

A3D: Adaptive, Accurate, and Autonomous Navigation for Edge-Assisted  Drones

Federated Learning is a modern decentralized machine learning technique where
user equipments perform machine learning tasks locally and then upload the
model parameters to a central server. In this paper, we consider a 3-layer
hierarchical federated learning system which involves model parameter exchanges
between the cloud and edge servers, and the edge servers and user equipment. In
a hierarchical federated learning model, delay in communication and computation
of model parameters has a great impact on achieving a predefined global model
accuracy. Therefore, we formulate a joint learning and communication
optimization problem to minimize total model parameter communication and
computation delay, by optimizing local iteration counts and edge iteration
counts. To solve the problem, an iterative algorithm is proposed. After that, a
time-minimized UE-to-edge association algorithm is presented where the maximum
latency of the system is reduced. Simulation results show that the global model
converges faster under optimal edge server and local iteration counts. The
hierarchical federated learning latency is minimized with the proposed
UE-to-edge association strategy.

本文提出了一种层次化联邦学习系统，结合联邦学习中的通信优化问题，通过优化本地迭代计数和边缘迭代计数，减少模型参数通信和计算延迟，并提出一种时限最小化的 UE-to-edge 关联算法。仿真结果表明，最优的边缘服务器和本地迭代计数可以使全局模型更快地收敛，利用 UE-to-edge 关联的策略可以最小化层次联邦学习延迟。

分层联邦学习中的时间最小化

Time Minimization in Hierarchical Federated Learning

Task scheduling is a critical problem when one user offloads multiple
different tasks to the edge server. When a user has multiple tasks to offload
and only one task can be transmitted to server at a time, while server
processes tasks according to the transmission order, the problem is NP-hard.
However, it is difficult for traditional optimization methods to quickly obtain
the optimal solution, while approaches based on reinforcement learning face
with the challenge of excessively large action space and slow convergence. In
this paper, we propose a Digital Twin (DT)-assisted RL-based task scheduling
method in order to improve the performance and convergence of the RL. We use DT
to simulate the results of different decisions made by the agent, so that one
agent can try multiple actions at a time, or, similarly, multiple agents can
interact with environment in parallel in DT. In this way, the exploration
efficiency of RL can be significantly improved via DT, and thus RL can
converges faster and local optimality is less likely to happen. Particularly,
two algorithms are designed to made task scheduling decisions, i.e.,
DT-assisted asynchronous Q-learning (DTAQL) and DT-assisted exploring
Q-learning (DTEQL). Simulation results show that both algorithms significantly
improve the convergence speed of Q-learning by increasing the exploration
efficiency.

本文提出了一种 Digital Twin 辅助的 RL 任务调度方法，在探索效率上通过 DT 显著提高 RL 的收敛速度，使用 DT 模拟智能体做出的不同决策，从而实现探索多种行动并行交互的效果，提高了更快的数据分析能力和收敛效果。