Dawei Chen

We present GreenAuto, an end-to-end automated platform designed for sustainable AI model exploration, generation, deployment, and evaluation. GreenAuto employs a Pareto front-based search method within an expanded neural architecture search (NAS) space, guided by gradient descent to optimize model exploration. Pre-trained kernel-level energy predictors estimate energy consumption across all models, providing a global view that directs the search toward more sustainable solutions. By automating performance measurements and iteratively refining the search process, GreenAuto demonstrates the efficient identification of sustainable AI models without the need for human intervention.

High-definition (HD) maps will serve as the key enabler to achieve autonomous driving. To maintain up-todate information in HD maps, we explore crowdsourcing from connected and automated vehicles (CAVs). It is, however, challenging to achieve live HD maps under complex dynamics in automotive edge computing networks. In this paper, we propose a new crowdsourcing HD map (EdgeMap) to enable collaborative computation between CAVs and edge servers. We formulate the optimization problem to minimize the total usage of transmission and computing resources under the constraint of end-to-end latency. To solve the problem, we design a new DATE algorithm to adaptively partition vehicular offloadings and reserve multiple domain resources, based on asynchronous multiagent reinforcement learning and expected gradient descents. We develop an end-to-end network simulator, including vehicle and edge computation and radio transmission in the 5G network, to evaluate the performance of EdgeMap. The results show that EdgeMap can achieve more than 30% reduction in total resource usage as compared to state-of-the-art solutions.

Video instance segmentation has emerged as a critical component in enabling connected vehicles to comprehend complex driving scenes, thereby facilitating navigation under various driving conditions. Recent advances focus on video-based solutions, which leverage temporal and spatial information to achieve superior performance compared to the traditional image-based approaches. However, these video-based solutions present challenges for efficient deployment at the edge due to their high computational and memory demands, making them inefficient for deployment on edge devices, such as intelligent vehicles. Furthermore, the large size of video data makes it impractical to upload to cloud servers. To address the latency challenge during on-device inference, we propose to incorporate early exits into the model. While the early exit strategy has been successful in image classification and natural language processing tasks, our study is the first to explore its application in video instance segmentation. Specifically, we incorporate early exits into the transformer-based video instance segmentation model, VisTR. Our experimental results on the YouTube-VIS dataset demonstrate that early exit can significantly speed up the inference by up to 4.83× with a minimal trade-off of only 3% in the averaged precision scores. Furthermore, our qualitative analysis confirms the satisfactory quality of the generated segmentation masks.

Cooperative perception is the key approach to augment the perception of connected and automated vehicles (CAVs) toward safe autonomous driving. However, it is challenging to achieve real-time perception sharing for hundreds of CAVs in large-scale deployment scenarios. In this paper, we propose AdaMap, a new high-scalable real-time cooperative perception system, which achieves assured percentile end-to-end latency under time-varying network dynamics. To achieve AdaMap, we design a tightly coupled data plane and control plane. In the data plane, we design a new hybrid localization module to dynamically switch between object detection and tracking, and a novel point cloud representation module to adaptively compress and reconstruct the point cloud of detected objects. In the control plane, we design a new graph-based object selection method to un-select excessive multi-viewed point clouds of objects, and a novel approximated gradient descent algorithm to optimize the representation of point clouds. We implement AdaMap on an emulation platform, including realistic vehicle and server computation and a simulated 5G network, under a 150-CAV trace collected from the CARLA simulator. The evaluation results show that, AdaMap reduces up to 49x average transmission data size at the cost of 0.37 reconstruction loss, as compared to state-of-the-art solutions, which verifies its high scalability, adaptability, and computation efficiency.

Today, deep learning optimization is primarily driven by research focused on achieving high inference accuracy and reducing latency. However, the energy efficiency aspect is often overlooked, possibly due to a lack of sustainability mindset in the field and the absence of a holistic energy dataset. In this paper, we conduct a threefold study, including energy measurement, prediction, and efficiency scoring, with an objective to foster transparency in power and energy consumption within deep learning across various edge devices. Firstly, we present a detailed, first-of-its-kind measurement study that uncovers the energy consumption characteristics of on-device deep learning. This study results in the creation of three extensive energy datasets for edge devices, covering a wide range of kernels, state-of-the-art DNN models, and popular AI applications. Secondly, we design and implement the first kernel-level energy predictors for edge devices based on our kernel-level energy dataset. Evaluation results demonstrate the ability of our predictors to provide consistent and accurate energy estimations on unseen DNN models. Lastly, we introduce two scoring metrics, PCS and IECS, developed to convert complex power and energy consumption data of an edge device into an easily understandable manner for edge device end-users. We hope our work can help shift the mindset of both end-users and the research community towards sustainability in edge computing, a principle that drives our research. Find data, code, and more up-to-date information at https://amai-gsu.github.io/DeepEn2023.

Highly-dynamic (HD) map is an indispensable building block in the future of autonomous driving, allowing for fine-grained environmental awareness, precise localization, and route planning. However, since HD maps include rich, multidimensional information, the volume of HD map data is substantial and cannot be transmitted frequently by several vehicles over vehicular networks in real-time. Therefore, in this paper, we propose a data source selection scheme for effective HD map transmissions in vehicular named data networking (NDN) scenarios. To achieve our goal, we created a vehicular NDN environment for data collection, processing, and transmission using the CARLA simulator and robot operating system 2 (ROS2). Next, due to our vehicular NDN’s dynamic and complex nature, we formulate the data source selection problem as a Markov decision process (MDP) and solve it using a reinforcement learning approach. For simplicity, we termed our proposed scheme data source optimization with reinforcement learning (DSORL), which selects suitable vehicles for HD map data transmission to MEC servers. The experiment results indicate that our suggested method outperformed existing baseline schemes, such as RLSS, Pro-RTT, and HDM-RTT, across all performance criteria in the evaluation. For instance, the system throughput increases by $65\%-72.68\%$ compared to other baseline systems. Similarly, the proposed approach can minimize packet loss rate, data size, and transmission time by up to $60.6\%$ , $77.5\%$ , and $54.1\%$ , respectively.

The metaverse is a perpetual, immersive, and shared digital universe that is linked to but is beyond the physical reality, and this emerging technology is attracting enormous attention from different industries. In this article, we define the first holistic realization of the metaverse in the mobility domain, coined as metamobility. We present our vision of what metamobility will be and describe its basic architecture. We also propose two use cases, tactile live maps and metaempowered advanced driver-assistance systems (ADASs), to demonstrate how metamobility will benefit and reshape future mobility systems. Each use case is discussed from the perspective of the technology evolution, future vision, and critical research challenges. Finally, we identify multiple concrete open research issues.

High definition (HD) map is a key technology that enables autonomous driving, which has the characteristic of frequent updates and low delay requirements. In order to minimize the HD map updating latency, this paper jointly considers the generation and transmission processes of HD map. The generation process of HD map relies heavily on the data captured by different types of sensors, which is modeled by the federated analytics process. Also, in order to quantify the diversity of utilized sensors, the concept of love of variety is introduced in this paper. Then, considering HD map is composed of dynamic and static layers that have different latency requirements during the transmission process, this paper proposes a method to allocate edge server capacity to each HD map layer such that the overall information staleness can be minimized. Firstly, the deterministic edge capacity case is discussed and the solution is derived by obtaining the Karush—Kuhn—Tucker conditions. Then, considering the practice that an edge server provides services to multiple attached devices simultaneously, contentions among these devices make available capacity for the autonomous vehicle variational. Therefore, an uncertain edge capacity case is discussed as well, where the uncertainty is described by Wasserstein metrics and the problem is reformulated into distributional robust chance constrained optimization problem. And for the low latency purpose, we utilize an inner approximation to reduce the complexity of the original problem and find a suboptimal solution. Finally, simulation results demonstrate the effectiveness of our proposed methods that achieves the lowest latency for HD map updating.