Qing Li’s research while affiliated with Peng Cheng Laboratory and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (322)


Unified and efficient multi-view clustering with tensorized bipartite graph
  • Article

April 2025

·

1 Read

Expert Systems with Applications

Lei Cao

·

Zhenzhu Chen

·

Chuanqing Tang

·

[...]

·


Synthetic User Behavior Sequence Generation with Large Language Models for Smart Homes

January 2025

In recent years, as smart home systems have become more widespread, security concerns within these environments have become a growing threat. Currently, most smart home security solutions, such as anomaly detection and behavior prediction models, are trained using fixed datasets that are precollected. However, the process of dataset collection is time-consuming and lacks the flexibility needed to adapt to the constantly evolving smart home environment. Additionally, the collection of personal data raises significant privacy concerns for users. Lately, large language models (LLMs) have emerged as a powerful tool for a wide range of tasks across diverse application domains, thanks to their strong capabilities in natural language processing, reasoning, and problem-solving. In this paper, we propose an LLM-based synthetic dataset generation IoTGen framework to enhance the generalization of downstream smart home intelligent models. By generating new synthetic datasets that reflect changes in the environment, smart home intelligent models can be retrained to overcome the limitations of fixed and outdated data, allowing them to better align with the dynamic nature of real-world home environments. Specifically, we first propose a Structure Pattern Perception Compression (SPPC) method tailored for IoT behavior data, which preserves the most informative content in the data while significantly reducing token consumption. Then, we propose a systematic approach to create prompts and implement data generation to automatically generate IoT synthetic data with normative and reasonable properties, assisting task models in adaptive training to improve generalization and real-world performance.


Hyper-parameters
Experimental Results on the CrossPlat- form(Android) Dataset w.r.t. Flow-level Traffic Clas- sification Task. (BOLD indicates the best score, and UNDERLINE denotes the second best one. Avg. Rank indicates the average ranking of each model's results across all datasets and metrics.)
Revolutionizing Encrypted Traffic Classification with MH-Net: A Multi-View Heterogeneous Graph Model
  • Preprint
  • File available

January 2025

·

6 Reads

With the growing significance of network security, the classification of encrypted traffic has emerged as an urgent challenge. Traditional byte-based traffic analysis methods are constrained by the rigid granularity of information and fail to fully exploit the diverse correlations between bytes. To address these limitations, this paper introduces MH-Net, a novel approach for classifying network traffic that leverages multi-view heterogeneous traffic graphs to model the intricate relationships between traffic bytes. The essence of MH-Net lies in aggregating varying numbers of traffic bits into multiple types of traffic units, thereby constructing multi-view traffic graphs with diverse information granularities. By accounting for different types of byte correlations, such as header-payload relationships, MH-Net further endows the traffic graph with heterogeneity, significantly enhancing model performance. Notably, we employ contrastive learning in a multi-task manner to strengthen the robustness of the learned traffic unit representations. Experiments conducted on the ISCX and CIC-IoT datasets for both the packet-level and flow-level traffic classification tasks demonstrate that MH-Net achieves the best overall performance compared to dozens of SOTA methods.

Download

Defying Multi-model Forgetting in One-shot Neural Architecture Search Using Orthogonal Gradient Learning

January 2025

IEEE Transactions on Computers

One-shot neural architecture search (NAS) trains an over-parameterized network (termed as supernet) that assembles all the architectures as its subnets by using weight sharing for computational budget reduction. However, there is an issue of multi-model forgetting during supernet training that some weights of the previously well-trained architecture will be overwritten by that of the newly sampled architecture which has overlapped structures with the old one. To overcome the issue, we propose an orthogonal gradient learning (OGL) guided supernet training paradigm, where the novelty lies in the fact that the weights of the overlapped structures of current architecture are updated in the orthogonal direction to the gradient space of these overlapped structures of all previously trained architectures. Moreover, a new approach of calculating the projection is designed to effectively find the base vectors of the gradient space to acquire the orthogonal direction. We have theoretically and experimentally proved the effectiveness of the proposed paradigm in overcoming the multi-model forgetting. Besides, we apply the proposed paradigm to two one-shot NAS baselines, and experimental results demonstrate that our approach is able to mitigate the multi-model forgetting and enhance the predictive ability of the supernet with remarkable efficiency on popular test datasets.


Adaptive and Low-cost Traffic Engineering: A Traffic Matrix Clustering Perspective

January 2025

·

6 Reads

IEEE Journal on Selected Areas in Communications

Traffic engineering (TE) has attracted extensive attention over the years. Operators expect to design a TE scheme that accommodates traffic dynamics well and achieves good TE performance with little overhead. Some approaches like oblivious routing compute an optimal static routing based on a large traffic matrix (TM) range, which usually leads to much performance loss. Many approaches compute routing solutions based on one or a few representative TMs obtained from observed historical TMs. However, they may suffer from performance degradation for unexpected TMs and usually induce much overhead of system operating. In this paper, we propose ALTE, an adaptive and low-cost TE scheme based on TM classification. We develop a novel clustering algorithm to properly group a set of historical TMs into several clusters and compute a candidate routing solution for each TM cluster. A machine learning classifier is trained to infer the proper candidate routing solution online based on the features extracted from some easily measured statistics. We implement a system prototype of ALTE and do extensive simulations and experiments using both real and synthetic traffic traces. The results show that ALTE achieves near-optimal performance for dynamic traffic and introduces little overhead of routing updates.



ParaLoupe: Real-Time Video Analytics on Edge Cluster via Mini Model Parallelization

December 2024

·

23 Reads

IEEE Transactions on Mobile Computing

Real-time video analytics on edge devices has gained increasing attention across a wide range of business areas. However, edge devices usually have limited computing resources. Consequently, conventional approaches to video analytics either deploy simplified models on the edge (resulting in low accuracy) or transmit video content to the cloud (resulting in high latency and network overheads) to enable deep learning inference ( e.g. object detection). In this paper, we introduce ParaLoupe, a novel real-time video analytics system that parallelizes deep learning inference in the edge cluster with task-oriented mini models. These mini models do not attain state-of-the-art accuracy individually, but collectively can achieve much better accuracy-latency tradeoff than state-of-the-art models. To achieve this, ParaLoupe crops multiple single-object patches from a given video frame. These single-object patches are then sent to multiple edge devices for parallel inference with specifically designed mini models. A patch-based task scheduling algorithm is further proposed to leverage the computing resources of the edge cluster to meet the service-level objectives. Our experimental results on real-world datasets show that ParaLoupe significantly outperforms baseline methods, achieving up to 14.1× inference speedup with accuracy on par with state-of-the-art models, or improving accuracy up to 45.1% under the same latency constraints.


IoTGemini: Modeling IoT Network Behaviors for Synthetic Traffic Generation

December 2024

·

19 Reads

·

1 Citation

IEEE Transactions on Mobile Computing

Synthetic traffic generation can produce sufficient data for model training of various traffic analysis tasks for IoT networks with few costs and ethical concerns. However, with the increasing functionalities of the latest smart devices, existing approaches can neither customize the traffic generation of various device functions nor generate traffic that preserves the sequentiality among packets as the real traffic. To address these limitations, this paper proposes IoTGemini, a novel framework for high-quality IoT traffic generation, which consists of a Device Modeling Module and a Traffic Generation Module. In the Device Modeling Module, we propose a method to obtain the profiles of the device functions and network behaviors, enabling IoTGemini to customize the traffic generation like using a real IoT device. In the Traffic Generation Module, we design a Packet Sequence Generative Adversarial Network (PS-GAN), which can generate synthetic traffic with high fidelity of both per-packet fields and sequential relationships. We set up a real-world IoT testbed to evaluate IoTGemini. The experiment result shows that IoTGemini can achieve great effectiveness in device modeling, high fidelity of synthetic traffic generation, and remarkable usability to downstream tasks on different traffic datasets and downstream traffic analysis tasks.




Citations (32)


... Multimedia recommendation systems (MRS) utilize user-item interactions and multimodal features such as text, images, audio, and videos to provide content recommendations based on user preferences [4][5][6][7]. They play a crucial role in platforms like e-commerce [8], social media [9], and video sharing [10], enhancing recommendation accuracy by capturing user preferences at the fine-grained level [4,11,12]. Early methods, such as VBPR [13] and DeepStyle [14], integrated multimodal information into traditional collaborative filtering paradigms but overlooked high-order user-item interaction connectivity [15]. Recent approaches, including MMGCN [16], GRCN [17], LATTICE [5], and DualGNN [18], employ graph convolution network (GCN) to better represent user-item interactions and improve recommendation performance [19,20]. ...

Reference:

Invariant Representation Learning in Multimedia Recommendation with Modality Alignment and Model Fusion
Mitigating Sample Selection Bias with Robust Domain Adaption in Multimedia Recommendation
  • Citing Conference Paper
  • October 2024

... However, the substantial size and high computational requirements of LLMs present significant challenges, particularly in resource-constrained environments and with the increasing reliance on MCE. The deployment of LLMs on remote cloud servers often results in time-consuming and resource-intensive interactions, which significantly increases storage and communication costs [11]. Additionally, the use of commercial LLMs can incur charges based on usage frequency, further adding to the economic burden. ...

Integration of LLM and Human-AI Coordination for Power Dispatching with Connected Electric Vehicles under SAGVNs
  • Citing Article
  • January 2024

IEEE Transactions on Vehicular Technology

... This optimization allows for maintaining performance levels while significantly reducing the demand for computing resources and storage space. These models are particularly wellsuited for resource-constrained real-world applications, such as mobile devices, edge computing, and real-time systems, while ensuring a higher level of accuracy [15], [16]. Typically, lightweight LLMs exhibit fewer parameters than 7 billion [17]. ...

Deep Customized Network Slicing and Efficient Routing for IoT Applications in B5G-Enabled Edge Computing Networks
  • Citing Article
  • January 2024

IEEE Internet of Things Journal

... The QLoRA techinque uses 4-bit quantized models and Low Rank Adapters, achieving 99.3% of ChatGPT's performance on the Vicuna benchmark with significant innovations for memory efficiency 35 . Many research have demonstrated success in fine-tuning pretrained LLMs using LoRA/QLoRA techniques [36][37][38][39][40][41][42][43] . Yet, few demonstrated the potential of QLoRA method over small volume of text data scraped from internet. ...

Quantized Side Tuning: Fast and Memory-Efficient Tuning of Quantized Large Language Models
  • Citing Conference Paper
  • January 2024

... They migrate to the site of damaged tissues to support their structure through the secretion of several factors that promote the healing process [62]. Recently, MSCs were shown to promote the regeneration of lung tissue's damaged mitochondria post-acute respiratory distress syndrome (ARDS) through mitochondria transfer [63]. MSCs also played a key role in repairing tissues of low regenerative capacities, such as knee cartilage by balancing the metabolic activity and the differentiation of chondrocytes [64]. ...

MSC-Mediated Mitochondrial Transfer Promotes Metabolic Reprogramming in Endothelial Cells and Vascular Regeneration in ARDS

... To enhance user convenience and safety in smart homes, these systems have integrated intelligent models to detect harmful or abnormal behaviors. They can also make recommendations or automatically take actions based on user behavior or contextual information, making this approach a common solution paradigm for home IoT [3], [4]. ...

Make Your Home Safe: Time-aware Unsupervised User Behavior Anomaly Detection in Smart Homes via Loss-guided Mask
  • Citing Conference Paper
  • August 2024

... Combining characteristics of video popularity and billing scenarios, such as peak billing that needs to consider popularity of videos during peak time, we have designed a file deployment solution based on popularity [81], and further designed file deployment push solution that considers load balancing of edge nodes [82]. ...

Smart Data-Driven Proactive Push to Edge Network for User-Generated Videos
  • Citing Conference Paper
  • May 2024

... This section shows the experimental results and demonstrates the comparison of the ECCS policy with NOSF of a task priority-based policy, PSO algorithm in searching for optimal solutions of resource allocation, and EVMS algorithm for task deadlines-awareness. Latency and throughput [34,35], resource utilization and server load, cache performance, and the task completion rate metrics are investigated respectively according to different scheduling policies under varying numbers of edge servers in an edge-cloud computing environment. ...

Genos: General In-Network Unsupervised Intrusion Detection by Rule Extraction
  • Citing Conference Paper
  • May 2024

... This optimization allows for maintaining performance levels while significantly reducing the demand for computing resources and storage space. These models are particularly wellsuited for resource-constrained real-world applications, such as mobile devices, edge computing, and real-time systems, while ensuring a higher level of accuracy [15], [16]. Typically, lightweight LLMs exhibit fewer parameters than 7 billion [17]. ...

SeIoT: Detecting Anomalous Semantics in Smart Homes via Knowledge Graph
  • Citing Article
  • January 2024

IEEE Transactions on Information Forensics and Security