arXiv subjects

Jemin Lee

Publications and source records attributed to Jemin Lee.

At least 19 recordsLinked to original sources

arXiv2026

EvoRIC: Reinforcement Learning Fine-Tuned LLM-empowered RAN Intelligent Control Toward Autonomous O-RAN

Despite recent advances in applying artificial intelligence (AI) techniques to radio access network (RAN), critical challenges remain: traditional machine learning (ML) algorithms suffer from limited generalization across varying network topologies, whereas general-purpose large language models (LLMs) face high computational demands and lack domain-specific knowledge. To address these gaps, this article introduces the evolving RAN intelligent controller (RIC) (EvoRIC) framework, a hierarchical architecture that enables continuous evolution by leveraging a non-real-time RIC (non-RT RIC) for global model updates and a near-real-time RIC (near-RT RIC) for local execution, dynamically empowering LLMs with domain-specific decision-making capabilities. Within this framework, we employ a reinforcement learning-based fine-tuning (RLFT) mechanism where an LLM operates as an actor within a proximal policy optimization (PPO) agent. By leveraging the interaction tuples collected from the wireless environment, the LLM's parameters are iteratively updated to align semantic reasoning with rigorous network performance objectives. We evaluate the generalization and efficacy of the proposed EvoRIC framework within integrated access and backhaul (IAB) networks, and finally, discuss the open challenges and future directions of the EvoRIC framework toward realizing autonomous O-RAN.

cs.NI↗

arXiv2026

QFlash: Bridging Quantization and Memory Efficiency in Vision Transformer Attention

FlashAttention improves efficiency through tiling, but its online softmax still relies on floating-point arithmetic for numerical stability, making full quantization difficult. We identify three main obstacles to integer-only FlashAttention: (1) scale explosion during tile-wise accumulation, (2) inefficient shift-based exponential operations on GPUs, and (3) quantization granularity constraints requiring uniform scales for integer comparison. To address these challenges, we propose \textit{QFlash}, an end-to-end integer FlashAttention design that performs softmax entirely in the integer domain and runs as a single Triton kernel. On seven attention workloads from ViT, DeiT, and Swin models, QFlash achieves up to 6.73$\times$ speedup over I-ViT and up to 8.69$\times$ speedup on Swin, while reducing energy consumption by 18.8\% compared to FP16 FlashAttention, without sacrificing Top-1 accuracy on ViT/DeiT and remaining competitive on Swin under per-tensor quantization. Our code is publicly available at https://github.com/EfficientCompLab/qflash.

cs.LG↗

arXiv2026

Age of Information under Source-Aware Truncated ARQ in Multi-Source Wireless Status Updating

This paper studies information timeliness in multi-source wireless Internet of Things (IoT) status updating systems under a truncated Automatic Repeat reQuest (ARQ) protocol. We propose a source-aware truncated ARQ (SATARQ) scheme that allows differentiated maximum transmission times (MTTs) tailored to different sources. This work focuses on a wireless system with preemptive update management. To study the statistical characteristics of the age of information (AoI) process for each source, a multi-dimensional age process (MDAP) is developed and modeled as a Markov chain, tracking both the AoI and the age of the concerned source's update currently in transmission. Via Markov analysis of the MDAP, we obtain analytical expressions for the distributions and averages of the AoI and peak AoI, as well as the average power consumption of IoT device. The timeliness-energy tradeoff is analyzed by examining the impact of the MTT, update generation probability (UGP), and wireless transmission power (TP). Moreover, this work explores the energy efficiency of the wireless status updating process and its relationship with the information timeliness and energy cost. Numerical results validate the theoretical analysis. Finally, it is demonstrated that the proposed SATARQ, combined with the optimization of MTTs, UGPs, and TPs, significantly improves the overall timeliness-energy tradeoff and energy efficiency across all sources.

cs.IT↗

arXiv2026

Secure Communications, Sensing, and Computing Towards Next-Generation Networks

Next-generation wireless networks are progressing beyond conventional connectivity to incorporate emerging sensing and computing capabilities. This convergence gives rise to integrated systems that enable not only uninterrupted communication, but also environmental awareness, intelligent decision-making, and novel applications that take advantage of these combined features. At the same time, this integration brings substantial security challenges. As computing, sensing, and communication become more tightly intertwined, the overall complexity of the system increases, creating new vulnerabilities and expanding the attack surface. The widespread deployment of data-heavy artificial intelligence applications further amplifies concerns regarding data security and privacy. This paper presents a comprehensive survey of security and privacy threats, along with potential countermeasures, in integrated wireless systems. We first review physical-layer security techniques for communication networks, and then investigate the security and privacy implications of semantic and pragmatic communications and their associated cross-layer design methodologies. For sensing functionalities, we pinpoint security and privacy risks at the levels of signal sources, propagation channels, and sensing targets, and summarize state-of-the-art defense strategies for each. The growing computational requirements of these applications drive the need for distributed computing over the network, which introduces additional risks such as data leakage, weak authentication, and multiple points of failure. We subsequently discuss secure coded computing approaches that can help overcome several of these challenges. Finally, we introduce unified security frameworks tailored to integrated communication-sensing-computing architectures, offering an end-to-end perspective on protecting future wireless systems.

cs.IT↗

arXiv2025

IPTQ-ViT: Post-Training Quantization of Non-linear Functions for Integer-only Vision Transformers

Previous Quantization-Aware Training (QAT) methods for vision transformers rely on expensive retraining to recover accuracy loss in non-linear layer quantization, limiting their use in resource-constrained environments. In contrast, existing Post-Training Quantization (PTQ) methods either partially quantize non-linear functions or adjust activation distributions to maintain accuracy but fail to achieve fully integer-only inference. In this paper, we introduce IPTQ-ViT, a novel PTQ framework for fully integer-only vision transformers without retraining. We present approximation functions: a polynomial-based GELU optimized for vision data and a bit-shifting-based Softmax designed to improve approximation accuracy in PTQ. In addition, we propose a unified metric integrating quantization sensitivity, perturbation, and computational cost to select the optimal approximation function per activation layer. IPTQ-ViT outperforms previous PTQ methods, achieving up to 6.44\%p (avg. 1.78\%p) top-1 accuracy improvement for image classification, 1.0 mAP for object detection. IPTQ-ViT outperforms partial floating-point PTQ methods under W8A8 and W4A8, and achieves accuracy and latency comparable to integer-only QAT methods. We plan to release our code https://github.com/gihwan-kim/IPTQ-ViT.git.

cs.CV↗

arXiv2025

A Survey on Inference Engines for Large Language Models: Perspectives on Optimization and Efficiency

Large language models (LLMs) are widely applied in chatbots, code generators, and search engines. Workload such as chain-of-throught, complex reasoning, agent services significantly increase the inference cost by invoke the model repeatedly. Optimization methods such as parallelism, compression, and caching have been adopted to reduce costs, but the diverse service requirements make it hard to select the right method. Recently, specialized LLM inference engines have emerged as a key component for integrating the optimization methods into service-oriented infrastructures. However, a systematic study on inference engines is still lacking.This paper provides a comprehensive evaluation of 25 open-source and commercial inference engines. We examine each inference engine in terms of ease-of-use, ease-of-deployment, general-purpose support, scalability, and suitability for throughput- and latency-aware computation. Furthermore, we explore the design goals of each inference engine by investigating the optimization techniques it supports. In addition, we assess the ecosystem maturity of open source inference engines and handle the performance and cost policy of commercial solutions.We outline future research directions that include support for complex LLM-based services, support of various hardware, and enhanced security, offering practical guidance to researchers and developers in selecting and designing optimized LLM inference engines. We also provide a public repository to continually track developments in this fast-evolving field: \href{https://github.com/sihyeong/Awesome-LLM-Inference-Engine}{https://github.com/sihyeong/Awesome-LLM-Inference-Engine}.

cs.CL↗

arXiv2025

LLM-hRIC: LLM-empowered Hierarchical RAN Intelligent Control for O-RAN

Despite recent advances in applying large language models (LLMs) and machine learning (ML) techniques to open radio access network (O-RAN), critical challenges remain, such as insufficient cooperation between radio access network (RAN) intelligent controllers (RICs), high computational demands hindering real-time decisions, and the lack of domain-specific finetuning. Therefore, this article introduces the LLM-empowered hierarchical RIC (LLM-hRIC) framework to improve the collaboration between RICs in O-RAN. The LLM-empowered non-real-time RIC (non-RT RIC) acts as a guider, offering a strategic guidance to the near-real-time RIC (near-RT RIC) using global network information. The RL-empowered near-RT RIC acts as an implementer, combining this guidance with local real-time data to make near-RT decisions. We evaluate the feasibility and performance of the LLM-hRIC framework in an integrated access and backhaul (IAB) network setting, and finally, discuss the open challenges of the LLM-hRIC framework for O-RAN.

cs.NI↗

arXiv2025

A Review on Proprietary Accelerators for Large Language Models

With the advancement of Large Language Models (LLMs), the importance of accelerators that efficiently process LLM computations has been increasing. This paper discusses the necessity of LLM accelerators and provides a comprehensive analysis of the hardware and software characteristics of the main commercial LLM accelerators. Based on this analysis, we propose considerations for the development of next-generation LLM accelerators and suggest future research directions.

cs.PF↗

arXiv2025

QuantuneV2: Compiler-Based Local Metric-Driven Mixed Precision Quantization for Practical Embedded AI Applications

Mixed-precision quantization methods have been proposed to reduce model size while minimizing accuracy degradation. However, existing studies require retraining and do not consider the computational overhead and intermediate representations (IR) generated during the compilation process, limiting their application at the compiler level. This computational overhead refers to the runtime latency caused by frequent quantization and dequantization operations during inference. Performing these operations at the individual operator level causes significant runtime delays. To address these issues, we propose QuantuneV2, a compiler-based mixed-precision quantization method designed for practical embedded AI applications. QuantuneV2 performs inference only twice, once before quantization and once after quantization, and operates with a computational complexity of O(n) that increases linearly with the number of model parameters. We also made the sensitivity analysis more stable by using local metrics like weights, activation values, the Signal to Quantization Noise Ratio, and the Mean Squared Error. We also cut down on computational overhead by choosing the best IR and using operator fusion. Experimental results show that QuantuneV2 achieved up to a 10.28 percent improvement in accuracy and a 12.52 percent increase in speed compared to existing methods across five models: ResNet18v1, ResNet50v1, SqueezeNetv1, VGGNet, and MobileNetv2. This demonstrates that QuantuneV2 enhances model performance while maintaining computational efficiency, making it suitable for deployment in embedded AI environments.

cs.AI↗

arXiv2024

ML$^2$Tuner: Efficient Code Tuning via Multi-Level Machine Learning Models

The increasing complexity of deep learning models necessitates specialized hardware and software optimizations, particularly for deep learning accelerators. Existing autotuning methods often suffer from prolonged tuning times due to profiling invalid configurations, which can cause runtime errors. We introduce ML$^2$Tuner, a multi-level machine learning tuning technique that enhances autotuning efficiency by incorporating a validity prediction model to filter out invalid configurations and an advanced performance prediction model utilizing hidden features from the compilation process. Experimental results on an extended VTA accelerator demonstrate that ML$^2$Tuner achieves equivalent performance improvements using only 12.3% of the samples required with a similar approach as TVM and reduces invalid profiling attempts by an average of 60.8%, Highlighting its potential to enhance autotuning performance by filtering out invalid configurations

cs.LG↗

arXiv2024

Exploring the Trade-Offs: Quantization Methods, Task Difficulty, and Model Size in Large Language Models From Edge to Giant

Quantization has gained attention as a promising solution for the cost-effective deployment of large and small language models. However, most prior work has been limited to perplexity or basic knowledge tasks and lacks a comprehensive evaluation of recent models like Llama-3.3. In this paper, we conduct a comprehensive evaluation of instruction-tuned models spanning 1B to 405B parameters, applying four quantization methods across 13 datasets. Our findings reveal that (1) quantized models generally surpass smaller FP16 baselines, yet they often struggle with instruction-following and hallucination detection; (2) FP8 consistently emerges as the most robust option across tasks, and AWQ tends to outperform GPTQ in weight-only quantization; (3) smaller models can suffer severe accuracy drops at 4-bit quantization, while 70B-scale models maintain stable performance; (4) notably, \textit{hard} tasks do not always experience the largest accuracy losses, indicating that quantization magnifies a model's inherent weaknesses rather than simply correlating with task difficulty; and (5) an LLM-based judge (MT-Bench) highlights significant performance declines in Coding and STEM tasks, though it occasionally reports improvements in reasoning.

cs.CL↗

arXiv2024

Mixed Non-linear Quantization for Vision Transformers

The majority of quantization methods have been proposed to reduce the model size of Vision Transformers, yet most of them have overlooked the quantization of non-linear operations. Only a few works have addressed quantization for non-linear operations, but they applied a single quantization method across all non-linear operations. We believe that this can be further improved by employing a different quantization method for each non-linear operation. Therefore, to assign the most error-minimizing quantization method from the known methods to each non-linear layer, we propose a mixed non-linear quantization that considers layer-wise quantization sensitivity measured by SQNR difference metric. The results show that our method outperforms I-BERT, FQ-ViT, and I-ViT in both 8-bit and 6-bit settings for ViT, DeiT, and Swin models by an average of 0.6%p and 19.6%p, respectively. Our method outperforms I-BERT and I-ViT by 0.6%p and 20.8%p, respectively, when training time is limited. We plan to release our code at https://gitlab.com/ones-ai/mixed-non-linear-quantization.

cs.CV↗

arXiv2024

Q-HyViT: Post-Training Quantization of Hybrid Vision Transformers with Bridge Block Reconstruction for IoT Systems

Recently, vision transformers (ViTs) have superseded convolutional neural networks in numerous applications, including classification, detection, and segmentation. However, the high computational requirements of ViTs hinder their widespread implementation. To address this issue, researchers have proposed efficient hybrid transformer architectures that combine convolutional and transformer layers with optimized attention computation of linear complexity. Additionally, post-training quantization has been proposed as a means of mitigating computational demands. For mobile devices, achieving optimal acceleration for ViTs necessitates the strategic integration of quantization techniques and efficient hybrid transformer structures. However, no prior investigation has applied quantization to efficient hybrid transformers. In this paper, we discover that applying existing post-training quantization (PTQ) methods for ViTs to efficient hybrid transformers leads to a drastic accuracy drop, attributed to the four following challenges: (i) highly dynamic ranges, (ii) zero-point overflow, (iii) diverse normalization, and (iv) limited model parameters ($<$5M). To overcome these challenges, we propose a new post-training quantization method, which is the first to quantize efficient hybrid ViTs (MobileViTv1, MobileViTv2, Mobile-Former, EfficientFormerV1, EfficientFormerV2). We achieve a significant improvement of 17.73% for 8-bit and 29.75% for 6-bit on average, respectively, compared with existing PTQ methods (EasyQuant, FQ-ViT, PTQ4ViT, and RepQ-ViT)}. We plan to release our code at https://gitlab.com/ones-ai/q-hyvit.

cs.CV↗

arXiv2024

Positioning Using Wireless Networks: Applications, Recent Progress and Future Challenges

Positioning has recently received considerable attention as a key enabler in emerging applications such as extended reality, unmanned aerial vehicles and smart environments. These applications require both data communication and high-precision positioning, and thus they are particularly well-suited to be offered in wireless networks (WNs). The purpose of this paper is to provide a comprehensive overview of existing works and new trends in the field of positioning techniques from both the academic and industrial perspectives. The paper provides a comprehensive overview of positioning in WNs, covering the background, applications, measurements, state-of-the-art technologies and future challenges. The paper outlines the applications of positioning from the perspectives of public facilities, enterprises and individual users. We investigate the key performance indicators and measurements of positioning systems, followed by the review of the key enabler techniques such as artificial intelligence/large models and adaptive systems. Next, we discuss a number of typical wireless positioning technologies. We extend our overview beyond the academic progress, to include the standardization efforts, and finally, we provide insight into the challenges that remain. The comprehensive overview of exisitng efforts and new trends in the field of positioning from both the academic and industrial communities would be a useful reference to researchers in the field.

eess.SP↗

arXiv2024

Visual Preference Inference: An Image Sequence-Based Preference Reasoning in Tabletop Object Manipulation

In robotic object manipulation, human preferences can often be influenced by the visual attributes of objects, such as color and shape. These properties play a crucial role in operating a robot to interact with objects and align with human intention. In this paper, we focus on the problem of inferring underlying human preferences from a sequence of raw visual observations in tabletop manipulation environments with a variety of object types, named Visual Preference Inference (VPI). To facilitate visual reasoning in the context of manipulation, we introduce the Chain-of-Visual-Residuals (CoVR) method. CoVR employs a prompting mechanism that describes the difference between the consecutive images (i.e., visual residuals) and incorporates such texts with a sequence of images to infer the user's preference. This approach significantly enhances the ability to understand and adapt to dynamic changes in its visual environment during manipulation tasks. Furthermore, we incorporate such texts along with a sequence of images to infer the user's preferences. Our method outperforms baseline methods in terms of extracting human preferences from visual sequences in both simulation and real-world environments. Code and videos are available at: \href{https://joonhyung-lee.github.io/vpi/}{https://joonhyung-lee.github.io/vpi/}

cs.RO↗

arXiv2024

Learning-based sensing and computing decision for data freshness in edge computing-enabled networks

As the demand on artificial intelligence (AI)-based applications increases, the freshness of sensed data becomes crucial in the wireless sensor networks. Since those applications require a large amount of computation for processing the sensed data, it is essential to offload the computation load to the edge computing (EC) server. In this paper, we propose the sensing and computing decision (SCD) algorithms for data freshness in the EC-enabled wireless sensor networks. We define the η-coverage probability to show the probability of maintaining fresh data for more than η ratio of the network, where the spatial-temporal correlation of information is considered. We then propose the probability-based SCD for the single pre-charged sensor case with providing the optimal point after deriving the η-coverage probability. We also propose the reinforcement learning (RL)- based SCD by training the SCD policy of sensors for both the single pre-charged and multiple energy harvesting (EH) sensor cases, to make a real-time decision based on its observation. Our simulation results verify the performance of the proposed algorithms under various environment settings, and show that the RL-based SCD algorithm achieves higher performance compared to baseline algorithms for both the single pre-charged sensor and multiple EH sensor cases.

eess.SY↗

arXiv2023

Ensuring End-to-End Security with Fine-grained Access Control for Connected and Autonomous Vehicles

As advanced V2X applications emerge in the connected and autonomous vehicle (CAV), the data communications between in-vehicle end-devices and outside nodes increase, which make the end-to-end (E2E) security to in-vehicle end-devices as the urgent issue to be handled. However, the E2E security with fine-grained access control still remains as a challenging issue for resource-constrained end-devices since the existing security solutions require complicated key management and high resource consumption. Therefore, we propose a practical and secure vehicular communication protocol for the E2E security based on a new attribute-based encryption (ABE) scheme. In our scheme, the outsourced computation is provided for encryption, and the computation cost for decryption constantly remains small, regardless of the number of attributes. The policy privacy can be ensured by the proposed ABE to support privacy-sensitive V2X applications, and the existing identity-based signature for outsourced signing is newly reconstructed. Our scheme achieves the confidentiality, message authentication, identity anonymity, unlinkability, traceability, and reconfigurable outsourced computation, and we also show the practical feasibility of our protocol via the performance evaluation.

cs.CR↗

arXiv2023

Joint Service Caching and Computing Resource Allocation for Edge Computing-Enabled Networks

In this paper, we consider the service caching and the computing resource allocation in edge computing (EC) enabled networks. We introduce a random service caching design considering multiple types of latency sensitive services and the base stations (BSs)' service caching storage. We then derive a successful service probability (SSP). We also formulate a SSP maximization problem subject to the service caching distribution and the computing resource allocation. Then, we show that the optimization problem is nonconvex and develop a novel algorithm to obtain the stationary point of the SSP maximization problem by adopting the parallel successive convex approximation (SCA). Moreover, to further reduce the computational complexity, we also provide a low complex algorithm that can obtain the near-optimal solution of the SSP maximization problem in high computing capability region. Finally, from numerical simulations, we show that proposed solutions achieve higher SSP than baseline schemes. Moreover, we show that the near-optimal solution achieves reliable performance in the high computing capability region. We also explore the impacts of target delays, a BSs' service cache size, and an EC servers' computing capability on the SSP.

cs.NI↗