Caijun Sun

Caijun Sun

Auto-generated profile, based on publicly available information

Caijun Sun

Zhejiang Lab, Hangzhou, China

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
21 November 2023 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
Guangsheng YuGuangsheng Yu
Guangsheng Yu
Xu WangXu Wang
Xu Wang
CSCaijun Sun
Caijun Sun
QWQin Wang
Qin Wang
PYPing Yu
Ping Yu
Ren Ping LiuRen Ping Liu
Ren Ping Liu
Federated learning (FL) offers an effective learning architecture to protect data privacy in a distributed manner. However, the inevitable network asynchrony, overdependence on a central coordinator, and lack of an open and fair incentive mechanism collectively hinder FL’s further development. We propose IronForge, a new generation of FL framework, that features a directed acyclic graph (DAG)-based structure, where nodes represent uploaded models, and referencing relationships between models form the DAG that guides the aggregation process. This design eliminates the need for central coordinators to achieve fully decentralized operations. IronForge runs in a public and open network and launches a fair incentive mechanism by enabling state consistency in the DAG. Hence, the system fits in networks where training resources are unevenly distributed. In addition, dedicated defense strategies against prevalent FL attacks on incentive fairness and data privacy are presented to ensure the security of IronForge. Experimental results based on a newly developed test bed FLSim highlight the superiority of IronForge to the existing prevalent FL frameworks under various specifications in performance, fairness, and security. To the best of our knowledge, IronForge is the first secure and fully decentralized FL (DFL) framework that can be applied in open networks with realistic network and training settings.

18 March 2024 in Computer Networks
Verified
Journal Article
Open access
ZZZixu Zhang
Zixu Zhang
Guangsheng YuGuangsheng Yu
Guangsheng Yu
CSCaijun Sun
Caijun Sun
XWXu Wang
Xu Wang
YWYing Wang
Ying Wang
NGNektarios Georgalas
Nektarios Georgalas

17 February 2025 in IEEE Transactions on Computational Social Systems
Verified
Journal Article
Open access
QWQin Wang
Qin Wang
Guangsheng YuGuangsheng Yu
Guangsheng Yu
YSYilin Sai
Yilin Sai
CSCaijun Sun
Caijun Sun
LNLam Duc Nguyen
Lam Duc Nguyen
SCShiping Chen
Shiping Chen
As a typical instance of human–computer interaction, the notion of decentralized autonomous organization (DAO) represents an organization constructed by automatically executed rules, such as via smart contracts, incorporating features of the permissionless committee, transparent proposals, and fair contributions by stakeholders. As of May 2023, DAO has impacted over $24.3B market caps. However, there are limited studies focused on this emerging field. To fill the gap, we start from the ground truth by empirically studying the breadth and depth of the DAO markets in mainstream public chain ecosystems in this article. We dive into the most widely adoptable DAO launchpad, Snapshot, which covers 95% of the wild DAO projects for data collection and analysis. By integrating extensively enrolled DAOs and corresponding data measurements, we explore statistical resources from Snapshot and analyze data from 581 DAO projects, encompassing 16 246 proposals over the course of 3+ years. Our empirical research has uncovered a multitude of previously unknown facts about DAOs, spanning topics such as their status, features, performance, threats, and ways of improvement. We have distilled these findings into a series of key insights and takeaway messages, emphasizing their significance. Notably, our study is the first of its kind to comprehensively examine the DAO ecosystem with a focus on scale and scope of data, real-time relevance, practical implementations, and comprehensive metrics, addressing critical gaps in the current literature.

20 July 2020 in IEEE Access
Verified
Journal Article
Open access
CSCaijun Sun
Caijun Sun
HZHua Zhang
Hua Zhang
SQSujuan Qin
Sujuan Qin
JQJiawei Qin
Jiawei Qin
YSYijie Shi
Yijie Shi
QWQiaoyan Wen
Qiaoyan Wen
The Android packing techniques were originally used to pack and conceal important information of the apps to prevent malicious developers from deconstructing the software logic. However, due to the lack of supervision, packing techniques have become the common methods for Android malwares to harden apps and circumvent virus detection engines in recent years. With obfuscation and encryption techniques, packing engines can alter the code structure of malwares and hide the malicious code to deceive and bypass the detection mechanisms, such as signature matching. For packing techniques, the packers are the agents created by packing engines and used to protect the softwares. Hence the packers pose a major challenge to the automated malware detection when researchers analyze a large collection of Android apps statically. It is necessary to identify packed samples in advance so that researchers can adopt different process procedures. To address this problem, we propose an intelligent AnDroid Packer Detection Framework called DroidPDF. It adopts a concise feature set that is resilient to obfuscation techniques. It also introduces weighted entropy to improve the detection effectiveness and achieves an average F1 Score of 0.9870.

12 October 2018 in IEEE Access
Verified
Journal Article
Open access
CSCaijun Sun
Caijun Sun
HZHua Zhang
Hua Zhang
SQSujuan Qin
Sujuan Qin
NHNengqiang He
Nengqiang He
JQJiawei Qin
Jiawei Qin
HPHongwei Pan
Hongwei Pan
In recent years, many packing services emerge and have been used to protect Android applications by concealing the executable files. However, it also brings some severe problems. For example, Android malwares use packers to escape detection from the anti-virus engine, which makes it harder to filter out the malicious applications. At present, existing Android unpacking exploits are designed complicated and not adaptive for new packers, which makes the unpackers always failed to keep up with the new packing techniques. In this paper, we propose a universal unpacking framework named DexX to extract dex files protected by these packing services.We apply DexX to packed Android applications, the experiment results show that our DexX can extract and recover original executable files (dex files) packed by most well-known commercial packers effectively and accurately.

19 February 2019 in Wireless Personal Communications
Verified
Journal Article
Closed access
PDF file
PYPing Yu
Ping Yu
QWQiaoyan Wen
Qiaoyan Wen
WNWei Ni
Wei Ni
WLWenmin Li
Wenmin Li
CSCaijun Sun
Caijun Sun
ZJZhengping Jin
Zhengping Jin