A5047295889- Parallel Computing and Optimization Techniques
- Advanced Data Storage Technologies
- Interconnection Networks and Systems
- Embedded Systems Design Techniques
- Distributed and Parallel Computing Systems
- Cloud Computing and Resource Management
- Caching and Content Delivery
- Cloud Data Security Solutions
- Cryptography and Data Security
- Distributed systems and fault tolerance
- Security and Verification in Computing
- Privacy-Preserving Technologies in Data
- Low-power high-performance VLSI design
- Advanced Memory and Neural Computing
- Cellular Automata and Applications
- Radiation Effects in Electronics
- Algorithms and Data Compression
- Insect Resistance and Genetics
- VLSI and Analog Circuit Testing
- Network Security and Intrusion Detection
- Advanced Computational Techniques and Applications
- Service-Oriented Architecture and Web Services
- Cooperative Communication and Network Coding
- Advanced Malware Detection Techniques
- Coding theory and cryptography
Jiangsu University of Science and Technology
2007-2025
Beihang University
2006-2025
Shanghai University of Traditional Chinese Medicine
2021-2025
Southwest University
2018-2025
Yueyang Hospital
2025
University of California, Irvine
2016-2024
Wuhan Institute of Virology
2016-2024
Chinese Academy of Sciences
2016-2024
National University of Defense Technology
2013-2023
Northwest A&F University
2022-2023
MDS array codes are widely used in storage systems to protect data against erasures. We address the \emph{rebuilding ratio} problem, namely, case of erasures, what is fraction remaining information that needs be accessed order rebuild \emph{exactly} lost information? It clear when number erasures equals maximum an code can correct then rebuilding ratio 1 (access all information). However, interesting and more practical smaller than erasure correcting capability code. For example, consider...
With the emergence of many-core architectures, it is quite likely that multiple applications will run concurrently on a system. Existing locally and globally adaptive routing algorithms largely overlook issues associated with workload consolidation. The shortsightedness limits performance due to poor network congestion avoidance. Globally attack this issue by introducing propagation obtain status information beyond neighboring nodes. However, they may suffer from intra- inter-application...
With the SIMT execution model, GPUs can hide memory latency through massive multithreading for many applications that have regular access patterns. To support with irregular patterns, cache hierarchies been introduced to GPU architectures capture temporal and spatial locality mitigate effect of accesses. However, caches exhibit poor efficiency due mismatch throughput-oriented model its hierarchy design, which limits system performance energy-efficiency. The amount requests generated by...
Liver cancer is one of the cancers with highest mortality. In order to help doctors diagnose and treat liver lesion, an automatic segmentation model urgently needed due manually time-consuming error-prone. this paper, we propose a nested attention-aware network, named Attention UNet++. Our proposed method has deep supervised encoder-decoder architecture redesigned dense skip connection. UNet++ introduces attention mechanism between convolutional blocks so that features extracted at different...
We consider the problem of T-Private Information Retrieval with private side information (TPIR-PSI). In this problem, N replicated databases store K independent messages, and a user, equipped local cache that holds M messages as information, wishes to retrieve one other - messages. The desired message index must remain jointly even if any T collude. show capacity TPIR-PSI is (1+ T/N + ⋯ +(T/N) <sup xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Abstract Motivation Since December 2019, the newly identified coronavirus SARS-CoV-2 has caused a massive health crisis worldwide and resulted in over 70 000 COVID-19 infections so far. Clinical drugs targeting are urgently needed to decrease high fatality rate of confirmed patients. Traditional de novo drug discovery needs more than 10 years, repurposing seems best option currently find potential for treating COVID-19. Results Compared with traditional non-covalent drugs, covalent have...
Maximum distance separable (MDS) codes are widely used in storage systems to protect against disk (node) failures. A node is said have capacity <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">l</i> over some field F, if it can store that amount of symbols the field. An ( xmlns:xlink="http://www.w3.org/1999/xlink">n, k, l</i> ) MDS code uses n nodes l k information nodes. The property guarantees resiliency any...
Network virtualization has been proposed as a powerful vehicle for running multiple customized networks on shared infrastructure. Virtual network embedding is critical step that deals with efficient mapping of virtual nodes and links onto the substrate resources. Previous work in primarily focused designing heuristic algorithms static networks. infrastructure should be reconfigured or redeployed response to growth. In this paper, we address problem optimally redeploying existing evolves....
Routing algorithms for networks-on-chip (NoCs) typically only have a small number of virtual channels (VCs) at their disposal. Limited VCs pose several challenges to the design fully adaptive routing algorithms. First, based on previous deadlock-avoidance theories require conservative VC re-allocation scheme: can be re-allocated when it is empty, which limits performance. We propose novel scheme, whole packet forwarding (WPF), allows non-empty re-allocated. WPF leverages observation that...
MDS (maximum distance separable) array codes are widely used in storage systems due to their computationally efficient encoding and decoding procedures. An code with r redundancy nodes can correct any erasures by accessing (reading) all the remaining information both systematic parity (redundancy) nodes. However, practice, a single erasure is most likely failure event; hence, natural question how much do we need access order rebuild node? We define rebuilding ratio as fraction of accessed...
MDS codes are erasure-correcting that can correct the maximum number of erasures given redundancy or parity symbols. If an code has r parities and no more than occur, then by transmitting all remaining data in one recover original information. However, it was shown order to a single symbol erasure, only fraction 1/r information needs be transmitted. This is called repair bandwidth (fraction). Explicit constructions were previous works. we view each as vector column, forms 2D array such...
In distributed storage, a file is stored in set of nodes and protected by erasure-correcting codes. Regenerating code type with two properties: first, it can reconstruct the entire presence any r node erasures for some specified integer r; second, efficiently repair an erased from subset remaining given size. process, amount information transmitted each normalized storage size per termed bandwidth (fraction). When minimized, lower bounded 1/r, where number parity nodes. A attaining this...
Performance, power, and energy (PPE) are critical aspects of modern computing. It is challenging to accurately predict, in real time, the effect dynamic voltage frequency scaling (DVFS) on PPE across a wide range voltages frequencies. This results use reactive, iterative, inefficient algorithms for dynamically finding good DVFS states. We propose PPEP, an online prediction framework that proactively rapidly searches space. PPEP uses hardware events implement both cycles-per-instruction (CPI)...
Health-promoting compounds, antioxidant and myrosinase activity in the sprouts of three broccoli cultivars under 40 mM, 80 mM 160 NaCl were investigated. LangYan (LY) had richest health-promoting compounds among evaluated cultivars. Treatment significantly decreased content ascorbic acid total phenolic as well activity, but did not affect glucoraphanin, sulforaphane compared to control. However, treatment enhanced level phenolic, sulforaphane, while content. The results suggest that would...
A secure multi-party batch matrix multiplication problem (SMBMM) is considered, where the goal to allow a master efficiently compute pairwise products of two batches massive matrices, by distributing computation across S servers. Any X colluding servers gain no information about input, and gains additional input beyond product. solution called Generalized Cross Subspace Alignment codes with Noise (GCSA- NA) proposed in this work, based on cross-subspace alignment codes. The state art SMBMM...
In distributed storage systems that use coding, the issue of minimizing communication required to rebuild a node after failure arises. We consider problem repairing an erased in system uses EVENODD code. codes are maximum distance separable (MDS) array used protect against erasures, and only require XOR operations for encoding decoding. show when there two redundancy nodes, one systematic node, 3/4 information needs be transmitted. Interestingly, many cases, disk I/O is also minimized.
MDS array codes are widely used in storage systems to protect data against erasures. We address the rebuilding ratio problem, namely, case of erasures, what is fraction remaining information that needs be accessed order rebuild exactly lost information? It clear when number erasures equals maximum an code can correct then 1 (access all information). However, interesting (and more practical) smaller than erasure correcting capability code. For example, consider two erasures: What smallest...
Motivated by mobile edge computing and wireless data centers, we study a distributed framework where the nodes exchange information over interference network. Our follows structure of MapReduce. This consists Map, Shuffle, Reduce phases, Map are computation phases Shuffle is transmission phase. In our setting, assume that operated We demonstrate that, duplicating work at cluster in phase, one can reduce amount load required for this work, characterize fundamental tradeoff between...
Purpose This study aims to explore the impact of information security knowledge on compliance behavior using neuroscience science. Design/methodology/approach uses electroencephalography (EEG) equipment collect brain cognitive data from participants under two specific conditions. The conditions are before and after interventions contextualized with “Wannacry” ransomware incident. Findings Firstly, lead differences in participants’ behavioral conditions, showing increased attention warnings...
<title>Abstract</title> <bold>Flexible photoelectric power sources are promising to Internet-of-things sensors, freeing the limitations of wired or batteries and enabling bio-integrated devices. Traditional photo have be integrated with energy storage modules since photovoltaic materials an order nanosecond lifetime for photogenerated charge carriers. Obtaining flexible, biocompatible that can perform light harvesting store simultaneously is still challenging. In this paper, inspired by...