Detection&tracking framework enables real-time object detection services on resource-limited mobile devices in edge networks, where mobile devices only offload few key frames to edge servers for detection and track the other frames locally. Following this framework, the offloading decisions become more important, as fewer frames can be offloaded and their results can directly affect tracking accuracy of the subsequent frames. To make wise frame offloading decisions, it is crucial to leverage the relationship between the detected frame and the tracked frame. However, the existing studies solely use the content-level information to reflect the above relationship, i.e., they tend to offload the frames with the significant pixel differences from the adjacent frames. Link impact on such relationship is completely ignored, thus leading to significant accuracy degradation. In this paper, we propose Kite, a link-adaptive and real-time object detection in dynamic edge networks, which integrates both the link-level and content-level impact into frame offloading. Kite exploits a novel performance metric, "frame-anchor" distance, to indicate the impact of dynamic wireless links. With this metric, we can incorporate both link and content information into the offloading process. The real-world experiment results show that Kite can improve the detection accuracy in dynamic edge networks.

Trustworthy oracles are an essential component of blockchain technology, providing a mechanism for blockchain applications to access data from the real world and verify it on-chain. With the continuous development of blockchain technology, trustworthy oracles have become one of the hot research topics in the blockchain industry. This article reviews the current research status and challenges of trustworthy oracles, summarizing the major research progress in the oracle field in recent years from aspects such as architecture and key technologies. Based on this, we propose a system architecture called SmartLLM that allows smart contracts to call large language models based on a trustworthy oracle mechanism. There are two types of architectures: one is chain-native, and the other is a combination of on-chain and off-chain. This architecture provides technical support for smart contracts to use large language models, enriching the use cases of smart contracts.

Data authorization is the basis for the orderly sharing of medical data. Most of the applied decentralized authorization mechanisms rely on blockchain, but face the problems of privacy leakage and low efficiency. To solve these problems, we design a policy-driven decentralized authorization system named BAA, which protects user's behavior privacy in medical data sharing and improves efficiency of both on-chain and off-chain. In order to achieve these goals, BAA uses authorization tokens to represent permissions, protects privacy of the authorization process through hiding user behaviors, realizes batch data accessing by proposing a two-tier Merkle tree, and saves authorization data in a two-tier blockchain to improve on-chain efficiency. Extensive experimental results show that the overhead of cryptographic operations in BAA is acceptable compared to that in traditional authorization systems. In addition, throughput and latency of each operation in BAA can meet the efficiency needs of medical data authorization. The results also show that the preset authorization in blockchain is effective for reducing data user's waiting time and improving the efficiency of authorization.