Guohong Cao

Guohong Cao

Auto-generated profile, based on publicly available information

Guohong Cao

School of Electrical Engineering and Computer Science, Pennsylvania State University, University Park, PA, USA

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

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

Representative Publications

View All Publications
3 July 2024 in IEEE Network
Verified
Journal Article
Closed access
KKKhen Bo Kan
Khen Bo Kan
HMHyunsu Mun
Hyunsu Mun
GCGuohong Cao
Guohong Cao
YLYoungseok Lee
Youngseok Lee
In the evolving landscape of 5G networks, Network Data Analytics Function (NWDAF) emerges as a key component, interacting with core network elements to enhance data collection, model training, and analytical outcomes. Language Models (LLMs), with their state-of-the-art capabilities in natural language processing, have been successful in numerous fields. In particular, LLMs enhanced through instruction fine-tuning have demonstrated their effectiveness by employing sets of instructions to precisely tailor the model’s responses and behavior. However, it requires collecting a large pool of high-quality training data regarding the precise domain knowledge and the corresponding programming codes. In this paper, we present an open-source mobile network-specialized LLM, Mobile-LLaMA, which is an instruction-finetuned variant of the LLaMA 2 13B model. We build Mobile-LLaMA by instruction fine-tuning LLaMA 2 13B with our own network analysis data collected from publicly available, real-world 5G network datasets, and expanded its capabilities through a self-instruct framework utilizing OpenAI’s pre-trained models (PMs). Mobile-LLaMA has three main functions: packet analysis, IP routing analysis, and performance analysis, enabling it to provide network analysis and contribute to the automation and artificial intelligence (AI) required for 5G network management and data analysis. Our evaluation demonstrates Mobile-LLaMA’s proficiency in network analysis code generation, achieving a score of 247 out of 300, surpassing GPT-3.5’s score of 209.

2 November 2021 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
HLHeting Liu
Heting Liu
GCGuohong Cao
Guohong Cao
With Mobile Edge Computing (MEC), computational intensive applications can be offloaded to the nearby edge servers to support latency-sensitive applications on mobile devices. Different from the cloud, edge servers usually have limited resources, and then selecting which edge server to run the offloaded computation becomes an important issue. Although server selection has received considerable attention, not much work has been done to consider the limited coverage of the edge server and the frequent user movement, which introduce many dynamic changing factors affecting the workload of the edge server and making it hard to achieve long-term optimum in the edge server selection. To deal with these challenges, we model the problem of continuous server selection as a Markov Decision Process (MDP). The difficulty of this problem is that achieving long-term optimum requires future knowledge, such as user mobility, server workload, etc, which is not known a priori. We do not have such knowledge and thus cannot find the optimal policy through traditional methods. To address this problem, we propose a Deep Reinforcement Learning (DRL) based algorithm to learn the selection policy based on the observed performance of past server selections. Specifically, a Long Short-Term Memory (LSTM) based neural network is exploited to encode the historical information which helps infer future knowledge of the dynamically changing factors. Then the DRL model selects the optimal server automatically based on the extracted system states. Extensive trace-driven evaluations demonstrate that the proposed DRL-based algorithm has the lowest overall cost compared to existing solutions.

29 May 2023
Verified
Conference Paper
Open access
CGChongqi Guan
Chongqi Guan
HLHeting Liu
Heting Liu
GCGuohong Cao
Guohong Cao
SZSencun Zhu
Sencun Zhu
TPThomas La Porta
Thomas La Porta
As IoT devices are becoming widely deployed, there exist many threats to IoT-based systems due to their inherent vulnerabilities. One effective approach to improving IoT security is to deploy IoT honeypot systems, which can collect attack information and reveal the methods and strategies used by attackers. However, building high-interaction IoT honeypots is challenging due to the heterogeneity of IoT devices. Vulnerabilities in IoT devices typically depend on specific device types or firmware versions, which encourages attackers to perform pre-attack checks to gather device information before launching attacks. Moreover, conventional honeypots are easily detected because their replying logic differs from that of the IoT devices they try to mimic.To address these problems, we develop an adaptive high-interaction honeypot for IoT devices, called em HoneyIoT. We first build a real device based attack trace collection system to learn how attackers interact with IoT devices. We then model the attack behavior through markov decision process and leverage reinforcement learning techniques to learn the best responses to engage attackers based on the attack trace. We also use differential analysis techniques to mutate response values in some fields to generate high-fidelity responses.HoneyIoT has been deployed on the public Internet. Experimental results show that HoneyIoT can effectively bypass the pre-attack checks and mislead the attackers into uploading malware. Furthermore, HoneyIoT is covert against widely used reconnaissance and honeypot detection tools.

25 August 2020 in IEEE Transactions on Mobile Computing
Verified
Journal Article
Closed access
XCXianda Chen
Xianda Chen
TTTianxiang Tan
Tianxiang Tan
GCGuohong Cao
Guohong Cao
TPThomas F. La Porta
Thomas F. La Porta
High quality video streaming for mobile devices implies high energy consumption due to the transmitted data and the variation of wireless signals. As an example, transmissions in mobile scenarios (e.g., inside a moving bus) consumes more energy for devices than when accessing from a static environment (e.g., at home). The QoE for the user does not substantially increase when watching high bitrate videos in a vibrating environment (i.e., a moving vehicle), as the context, in this case vehicle’s vibration, affects the perceived QoE. To address this problem, we propose to save energy by considering the context (environment) of video streaming. To model the impact of context, we exploit the embedded accelerometer in smartphones to record the vibration level during video streaming. Based on quality assessment experiments, we collect traces and model the impact of video bitrate and vibration level on QoE, and model the impact of video bitrate and signal strength on power consumption. Based on the QoE model and the power model, we formulate the context-aware and energy-aware video streaming problem as an optimization problem. We present an optimal algorithm which can maximize QoE and minimize energy. Since the optimal algorithm requires perfect knowledge of future tasks, we propose an online bitrate selection algorithm. To further improve the performance of the online algorithm, we propose a crowdsourcing based bitrate selection algorithm. Through real measurements and trace-driven simulations, we demonstrate that the proposed algorithms can significantly outperform existing approaches when considering both energy and QoE.

25 March 2024 in IEEE Transactions on Mobile Computing
Verified
Journal Article
Closed access
TTTianxiang Tan
Tianxiang Tan
GCGuohong Cao
Guohong Cao
As Deep Neural Networks (DNNs) have been successfully applied to various fields, there is a tremendous demand for running DNNs on mobile devices. Although mobile GPU can be leveraged to improve performance, it consumes a large amount of energy. After a short period of time, the mobile device may become overheated and the processors are forced to reduce the clock speed, significantly reducing the processing speed. A different approach to support DNNs on mobile device is to leverage the Neural Processing Units (NPUs). Compared to GPU, NPU is much faster and more energy efficient, but with lower accuracy due to the use of low precision floating-point numbers. We propose to combine these two approaches to improve the performance of running DNNs on mobile devices by studying the thermal-aware scheduling problem, where the goal is to achieve a better tradeoff between processing time and accuracy while ensuring that the mobile device is not overheated. To solve the problem, we propose a heuristic-based scheduling algorithm to determine when to run DNNs on GPU and when to run DNNs on NPU based on the current states of the mobile device. The heuristic-based algorithm makes scheduling decisions greedily and ignores their future impacts. Thus, we propose a deep reinforcement learning based scheduling algorithm to further improve performance. Extensive evaluation results show that the proposed algorithms can significantly improve the performance of running DNNs on mobile devices while avoiding overheating.

5 June 2023
Verified
Conference Paper
Open access
HLHeting Liu
Heting Liu
ZLZhichao Li
Zhichao Li
CTCheng Tan
Cheng Tan
RYRongqiu Yang
Rongqiu Yang
GCGuohong Cao
Guohong Cao
CGChuanxiong Guo
Chuanxiong Guo
Graphics processing units (GPUs) are the de facto standard for processing deep learning (DL) tasks. In large-scale GPU clusters, GPU failures are inevitable and may cause severe consequences. For example, GPU failures disrupt distributed training, crash inference services, and result in service level agreement violations. In this paper, we study the problem of predicting GPU failures using machine learning (ML) models to mitigate their damages. We train prediction models on a four-month production dataset with 350 million entries at ByteDance. We observe that classic prediction models (GBDT, MLP, LSTM, and 1D-CNN) do not perform well---they are inaccurate for predictions and unstable over time. We propose several techniques to improve the precision and stability of predictions, including parallel and cascade model-ensemble mechanisms and a sliding training method. We evaluate the performance of our proposed techniques. The results show that our proposed techniques improve the prediction precision from 46.3% to 85.4% on production workloads.