Despite considerable advancements in specialized hardware, the majority of IoT edge devices still rely on CPUs. The burgeoning number of IoT users amplifies the challenges associated with performing multiple Deep Neural Network inferences on these resource-limited, CPU-only edge devices. Existing strategies, including model compression, hardware acceleration, and model partitioning, often involve a trade-off in inference accuracy, are unsuitable due to hardware specificity, or lead to inefficient resource utilization. In response to these challenges, this paper introduces L-PIC (Latency Minimized Parallel Inference on CPU)—a framework expressly devised to optimize resource allocation, decrease inference latency, and maintain result accuracy on CPU-only edge devices. A series of comprehensive experiments have verified the superior efficiency and effectiveness of the L-PIC framework in comparison to the state-of-the-art method. Remarkably, compared to the state-of-the-art method, L-PIC can reduce the inference latency of multi-DNN by an average of approximately 30% across all tested scenarios.

This work aims to design a real-time inference scheduler that delivers accurate and timely edge inference ser-vices for dynamic inference arrivals by leveraging dynamic batching and early-exiting techniques. Specifically, we consider an edge inference server that is preinstalled with multiple early-exit Deep Neural Networks (DNNs) that support batch processing. The in-ference tasks with strict deadline requirements arrive at the edge server randomly, and the utility of each timely processed task depends on the achieved accuracy. Therefore, we aim to design an edge inference scheduler that maximizes the system's total utility subject to resource and deadline constraints. We present this problem's mixed integer linear programming formulation. This problem is challenging due to high computational complexity, coupled-decision making, and task randomness. We propose to decompose the original problem into two sub-problems: the task assignment problem and the DNN configuration problem. For the task assignment problem, we develop a greedy task assignment algorithm. For the DNN configuration problem, we propose a Deep Reinforcement Learning-based solution. Simulation results show that the proposed algorithms outperform the state-of-the-art baselines.

Multi-agent reinforcement learning (MARL) has proven effective in training multi-robot confrontation, such as StarCraft and robot soccer games. However, the current joint action policies utilized in MARL have been unsuccessful in recognizing and preventing actions that often lead to failures on our side. This exacerbates the cooperation dilemma, ultimately resulting in our agents acting independently and being defeated individually by their opponents. To tackle this challenge, we propose a novel joint action policy, referred to as the consensus action policy (CAP). Specifically, CAP records the number of times each joint action has caused our side to fail in the past and computes a cooperation tendency, which is integrated with each agent’sQ-value and Nash bargaining solution to determine a joint action. The cooperation tendency promotes team cooperation by selecting joint actions that have a high tendency of cooperation and avoiding actions that may lead to team failure. Moreover, the proposed CAP policy can be extended to partially observable scenarios by combining it with DeepQnetwork or actor-critic–based methods. We conducted extensive experiments to compare the proposed method with seven existing joint action policies, including four commonly used methods and three state-of-the-art methods, in terms of episode rewards, winning rates, and other metrics. Our results demonstrate that this approach holds great promise for multi-robot confrontation scenarios.

To meet increasing demands for machine-learning-based applications, pushing inference services to the network edge has been a trend. This work aims to design an on-demand edge inference scheduler with accuracy and deadline guarantee for repetitive tasks. Specifically, we consider an edge server that is preinstalled with multiple early-exit Deep Neural Networks (DNNs), and each DNN-exit pair can provide inference service of different quality. We also consider tasks' diversity in quality of service requirements and related utility. We aim to maximize the system's total utility by optimizing service assignment and time scheduling subject to resource, accuracy, and deadline constraints. We present this problem's integer linear problem formulation and show this problem is NP-hard even for the offline case. This problem is challenging due to the coupled effect of service assignment and time scheduling. To derive low-complexity scheduling solutions, we introduce a task-service graph and convert this problem into a service assignment selection problem with schedulability constraints. Then, we design a polynomial complexity algorithm with$\frac{\rho}{\delta}$-approximation ratio for the offline problem, with$\rho$referring to the task-wise utility ratio,$\delta$referring to the maximum number of concurrent tasks. To handle the online problem, we propose an online heuristic algorithm. Simulation results show that the proposed algorithms outperform the state-of-the-art baseline algorithms.

In Shared Autonomous Vehicle (SAV) systems, real-time vehicle repositioning plays a crucial role in meeting time-varying traffic demand, which is normally designed by taking advantage of user demand prediction. Nonetheless, most existing studies only predict traffic demand and schedule SAVs separately, ignoring the tight interaction between the two components, e.g. the potential impact of repositioning results on demand prediction. Such a design lacks a deeply integrated design for both and may lead to inaccurate demand prediction and impaired repositioning performance. To tackle this challenge, we propose DRiVe, a coupling approach to Demand prediction and Repositioning for shared autonomous Vehicle system. Specifically, we consider electric SAVs and adopt model predictive control (MPC) to develop the repositioning strategy with the goal of minimizing the operator’s repositioning costs and passenger dissatisfaction. An online prediction is then introduced which not only implements the traditional demand prediction but also integrates the additional traffic demand generated by repositioning action. The numerical results demonstrate that the proposed DRiVe method achieves better performance in reducing passenger waiting time and idle distance compared to the state-of-the-art repositioning methods.

Image-based Multi-Agent Coverage Path Planning (MACPP) utilizes images as input to control multiple agents touring all nodes in a map, minimizing task duration and node revisiting. State-of-the-art (SOTA) studies have applied Multi-Agent Deep Reinforcement Learning (MADRL) to automate MACPP, primarily focusing on minimizing task duration. However, these approaches overlook the issue of repeated node visits, resulting in longer task durations and limited real-world applicability. To tackle this challenge, we develop a novel MADRL solution, referred to as MADRL with Mask Soft Attention, to minimize task duration and node re-visiting simultaneously. Our method uses mask soft attention to extract key features from raw image observations while masking task-independent features, reducing computational complexity and improving sample efficiency. We also cascade a multi-actor-critic architecture to accommodate even more agents with ease. Each agent is equipped with an actor to learn an action policy, and a shared critic evaluates a state value. To validate our approach, we implement seven SOTA MADRL methods in the MACPP area as baselines. Simulation results show that our method significantly outperforms the baselines regarding task duration and the number of times the node is repeatedly visited.

In Deep Reinforcement Learning (DRL) domain, a compound learning task is often decomposed into several sub-tasks in a divide-and-conquer manner, each trained separately and then fused concurrently to achieve the original task, referred to as policy fusion. However, the state-of-the-art (SOTA) policy fusion methods treat the importance of sub-tasks equally throughout the task process, eliminating the possibility of the agent relying on different sub-tasks at various stages. To address this limitation, we propose a generic policy fusion approach, referred to as Policy Fusion Learning with Dynamic Weights and Prior Reward (PFLDWPR), to automate the time-varying selection of sub-tasks. Specifically, PFLDWPR produces a time-varying one-hot vector for sub-tasks to dynamically select a suitable sub-task and mask the rest throughout the entire task process, enabling the fused strategy to optimally guide the agent in executing the compound task. The sub-tasks with the dynamic one-hot vector are then aggregated to obtain the action policy for the original task. Moreover, we collect sub-tasks’s rewards at the pre-training stage as a prior reward, which, along with the current reward, is used to train the policy fusion network. Thus, this approach reduces fusion bias by leveraging prior experience. Experimental results under three popular learning tasks demonstrate that the proposed method significantly improves three SOTA policy fusion methods in terms of task duration, episode reward, and score difference.

In this paper, the identification of multiple leaks in pipes based on transient waves is studied, which is, however, quite challenging due to its nonconvex nature with lots of local optima. Existing approaches need the number of leaks and suffer from a huge computational complexity that is increased exponentially with the number of leaks. To provide a scalable solution, we propose a compressive sensing (CS) framework to solve the multileaks identification. We first exploit the sparseness nature of leak locations through spatial sampling. Then, we formulate the multileak identification as a CS problem, where the spatial sample-dependent components form a basis matrix and the leak sizes are viewed as a sparse signal. We establish the convergence of spatial sampling mismatch and the two-restricted isometry property of basis matrix to justify the proposed CS framework. The proposed CS framework renders a superior-performance solution to multileak identification and its computational complexity is linear with the number of leaks, which is a significant technical improvement over existing approaches. In addition, a closed-form Cramer-Rao lower bound (CRLB) on the leak localization errors is derived. A geometric insight of CRLB evolution is presented to give us an intuitive understanding of the contribution of new measurements to leak localization performance.