A5071840581- Advanced Data Storage Technologies
- Distributed systems and fault tolerance
- Cloud Computing and Resource Management
- Photodynamic Therapy Research Studies
- Data Management and Algorithms
- Luminescence Properties of Advanced Materials
- Nonlinear Optical Materials Studies
- Algorithms and Data Compression
- Parallel Computing and Optimization Techniques
- Photochemistry and Electron Transfer Studies
- Advanced biosensing and bioanalysis techniques
- Analytical Chemistry and Sensors
- Laser Applications in Dentistry and Medicine
- Computability, Logic, AI Algorithms
- Radiation Detection and Scintillator Technologies
- Data Stream Mining Techniques
- Gas Sensing Nanomaterials and Sensors
- Dental materials and restorations
- Error Correcting Code Techniques
- Advanced Image and Video Retrieval Techniques
- Metabolomics and Mass Spectrometry Studies
- Image Retrieval and Classification Techniques
Huawei Technologies (China)
2023-2025
Shenzhen University
2022-2024
University of Hong Kong
2018-2020
Key-value stores in Cloud environments can contain more than 2 45 unique elements and be larger 100 PByte. B + -Tree s are well suited for these larger-than-memory datasets seamlessly index data stored on thousands of secondary storage devices. Unfortunately, it is often uneconomical to even store all inner tree nodes memory dataset sizes. Therefore, lookup performance affected by the additional IOs reading nodes. This number reduced increasing size leaf We propose HLN-Tree s, which support...
The overlapping coalition structure generation problem (OCSGP) is a challenging computational in multi-agent systems. It focuses on selecting possibly coalitions from set of agents to maximize the social welfare all while containing agents. However, practical applications, may be formed selectively respond tasks pool potential assigned Consequently, this study considers OCSGP task-based setting, where each agent has finite resources and can only interest, take mutually disjoint subsets...
The recently proposed learned indexes have attracted much attention as they can adapt to the actual data and query distributions attain better search efficiency. Based on this technique, several existing works build up for multi-dimensional achieve improved performance. A common paradigm of these is (i) map points a one-dimensional space using fixed space-filling curve (SFC) or its variant (ii) then apply indexing techniques. We notice that first step typically uses SFC method, such...
Learned indexes use machine learning techniques to improve index construction. However, they often face a fundamental trade-off between performance and memory consumption, especially in dynamic environments with frequent insert delete operations. This stems from the construction approaches used learned indexes: The top-down approach increases at cost of significant overhead, while bottom-up focuses on efficiency but introduces issues due prediction errors. % A unified solution that...
Many NoSQL databases support quorum-based protocols, which require a subset of replicas (called quorum) to respond each write/read operation. These systems configure the quorum size tune operation latency and adopt multiple consistency levels. Some recent works illustrate that using probability models quantify chance reading last update is important because it could avoid returning stale values under eventual consistency. There are two challenging issues: (1) from inconsistent replicas, how...
While CPU architectures are incorporating many more cores to meet ever-bigger workloads, advance in fault-tolerance support is indispensable for sustaining system performance under reliability constraints. Emerging non-volatile memory technologies yielding fast, dense, and energy-efficient NVRAM that can dethrone SSD drives persisting data. Research on using enable fast in-memory data persistence ongoing. In this work, we design implement a persistent object framework, dubbed scalable...
Fault-tolerance capabilities attract increasing attention from existing data processing frameworks, such as Apache Spark. To avoid replaying costly distributed computation, like shuffle, local checkpoint and remote replication are two popular approaches. They incur significant runtime overhead, extra storage cost or network traffic. Erasure coding is another emerging technology, which also enables resilience. It perceived capable of replacing the mechanisms for its high efficiency. However,...
Photo bleaching induced destruction is one of the main challenge for fluorescence organic dyes. Especially, laser dyes are more fragile to photo due high photon density pumping. In this work, NaYF4:Yb3+/Er3+ upconversion micro-particles (UCMPs) hybrid with bis-squaraine (Bis-SQ) dye form a UCMPs/Bis-SQ composite. The UCMPs excited by near infrared 980 nm an emission band located at 656 nm. By controlling mass ratio and Bis-SQ, energy emitted from transferred Bis-SQ through förster resonance...
Many organics are dangerous to environment as well precise electronic devices. Detection of is a big challenge due huge amount species organic materials. This because specific fluorescence probe needed in order detect kind material. make it impossible develop for each As result, development an all-purpose which high sensitive different types crucial importance. In this work, we report up-conversion detection Surface functional groups –COOH, help alter electrons transition (alter emission...
Emerging byte-addressable non-volatile memory is considered an alternative storage device for persistence guarantee with high performance. In this work, we design NVCL (NVRAM Cache-Line Granularity Logging) and develop it in Redis. The proposed exploits the cache-line flush granularity of NVRAM to maintain write-ahead differential log able non-continuous dirty cache lines persistence. effectively reduces size persistent logs by only flushing updated data every page. We measured performance...