The accelerated integration of edge computing and artificial intelligence has promoted the rise of edge intelligence, which is regarded as the key to solving the last-mile delivery problem of artificial intelligence technology. Although previous research has made many efforts in this area, they have struggled to serve the scenario containing multiple heterogeneous task requests. This challenge is further exacerbated when the number of terminal devices increases and multiple edge servers are required to collaborate to handle task requests. To this end, this paper proposes a fine-grained Collaborative DNN Model Selection scheme for heterogeneous edge computing systems (CoMS), aiming to promote cooperation between edge servers and achieve more effective model selection. Specifically, we first design a reinforcement learning scheme with real-time dynamic normalization strategy, aiming to accelerate model convergence and improve the efficiency of model selection. Next, we introduce a model selection strategy based on greedy algorithm and an efficient fine-grained collaborative model selection strategy respectively to promote cooperation between edge servers, thereby further achieving a balance between inference accuracy and overhead. Extensive experimental results show that compared with baselines, our CoMS reduces the average trade-off overhead by 4.2% to 12.8% and improves the success ratio by 3% to 16%.

Most existing synthetic aperture radar (SAR) ship classification technologies heavily rely on correctly labeled data, ignoring the discriminative features of unlabeled SAR ship images. Even though researchers try to enrich CNN-based features by introducing traditional handcrafted features, existing methods easily cause information redundancy and fail to capture the interaction between them. To address these issues, we propose a novel dual-stream contrastive predictive network (DCPNet), which consists of two asymmetric task designs and the false negative sample elimination module. The first task is to construct positive sample pairs, guiding the core encoder to learn more general representations. The second task is to encourage adaptive capture of the correspondence between deep features and handcrated features, achieving knowledge transfer within the model, and effectively improving the redundancy caused by the feature fusion. To increase the separability between clusters, we also design a cluster-level tasks. The experimental results on OpenSARShip and FUSAR-Ship datasets demonstrate the improvement in classification accuracy of supervised models and confirm the capability of learning effective representations of DCPNet.

In recent years, since edge computing has improved the performance of transportation systems, research on edge-computing-enabled transportation systems has received widespread attention. However, most previous studies overlooked that task requests in transportation systems are unevenly distributed in time and space, which easily causes the overloading of edge servers, resulting in high response latency. To this end, we present a novel task offloading scheme based on graph neural network (GNN) and deep reinforcement learning (DRL) in edge-computing-enabled transportation systems (TransEdge). Specifically, we first propose an adaptive node placement algorithm to assign Internet of Things sensors to appropriate edge servers, thereby minimizing transmission latency. Then, an improved DRL scheme based on GNN is designed to capture the spatial features between sensors, aiming to improve the accuracy of task offloading decisions. Finally, we introduce a task forwarding strategy based on the greedy algorithm to achieve collaborative task offloading between different edge servers and overcome the system instability caused by a sudden surge in task requests. We conduct extensive experiments on two real-world traffic data sets. The results show that TransEdge reduces the response latency by at least 3.7% compared to four baselines while achieving a success rate of 99%.