No abstract

Key-value stores based on a log-structured merge (LSM) tree have emerged in big data systems because of their scalability and reliability. An LSM-tree offers a multilevel data structure with a simple interface. However, it performs file rewrites at the disk level, which causes write amplification. This study is concerned with this problem in relation to an embedded board environment, which can be used in edge computing. Addressing the major problems associated with an LSM-tree, we propose a new key-value store named CaseDB, which aggressively separates keys and bloom filters on the non-volatile memory express (NVMe) drive and stores the values on the SSD. Our solution reduces the I/O cost and enhances the overall performance in a cost-efficient manner. CaseDB employs a memory component, CBuffer, to avoid small write operations, and a delayed value compaction technique that guarantees the sorted order for both keys and values. CaseDB also utilizes deduction-based data deduplication to prevent space amplification in the values layer. The experiments show that CaseDB outperforms LevelDB and WiscKey 5.7 and 1.8 times, respectively, with respect to data writes, and additionally improves the read performance by 1.5 times. CaseDB also avoids the space amplification of WiscKey.

No abstract

As the utilization of complex and heavy applications increases in autonomous driving, research on using mobile edge computing and task offloading for autonomous driving is being actively conducted. Recently, researchers have been studying task offloading algorithms using artificial intelligence, such as reinforcement learning or partial offloading. However, these methods require a lot of training data and critical deadlines and are weakly adaptive to complex and dynamically changing environments. To overcome this weakness, in this paper, we propose a novel task offloading algorithm based on Lyapunov optimization to maintain the system stability and minimize task processing delay. First, a real-time monitoring system is built to utilize distributed computing resources in an autonomous driving environment efficiently. Second, the computational complexity and memory access rate are analyzed to reflect the characteristics of the deep learning applications to the task offloading algorithm. Third, Lyapunov and Lagrange optimization solves the trade-off issues between system stability and user requirements. The experimental results show that the system queue backlog remains stable, and the tasks are completed within an average of 0.4231 s, 0.7095 s, and 0.9017 s for object detection, driver profiling, and image recognition, respectively. Therefore, we ensure that the proposed task offloading algorithm enables the deep learning application to be processed within the deadline and keeps the system stable.

No abstract