Searcharxiv⌕ Search

arXiv subjects

Bui Trong Duc

Publications and source records attributed to Bui Trong Duc.

6 recordsLinked to original sources

BECON: Belief-Conditioned Constrained Multi-Objective Reinforcement Learning under Drifting Preferences and Budgets

Constrained preference-conditioned MORL typically assumes full task observability, treating preferences and safety budgets as stationary task signals. In real-world deployment, however, these signals are often unreliable proxies: intent may drift, safety tolerances may shift with context, and exposed descriptors may be noisy, delayed, or incomplete. We therefore formulate constrained preference-conditioned MORL as a partially observed task-specification problem, where the agent must navigate a fundamental conflict between unreliable external signals and latent interaction dynamics. We propose \textbf{BECON}, a belief-conditioned framework for constrained MORL with drifting preferences and budgets, which infers latent task context from recent histories to estimate preferences and safety budgets. These estimates are fused with the exposed task signal via uncertainty-dependent gates, enabling the policy to defer to observations under high epistemic uncertainty while correcting them when history is informative. The fused context jointly conditions preference optimization and context-adaptive constraint enforcement. Experiments on diverse multi-objective control tasks demonstrate that BECON consistently improves over state-of-the-art MORL baselines, achieving stronger robustness and constraint satisfaction under task drift.

cs.LG↗

COMPASS: Composable Policy-Amortized Structured Search for LLM-Based Optimization Modeling

Large language models offer a promising technique for translating natural-language optimization problems into solver code, but their reliability remains constrained by the quality of template retrieval. While recent structured methods improve over flat prompting through hierarchical taxonomy search, their traversal rules remain fixed or LLM-driven at inference time, leaving retrieval unable to improve from past successes and failures. This motivates treating template retrieval as a policy that can be learned from modeling experience, rather than as a static inference-time rule. We propose COMPASS, a policy-amortized framework for retrieving optimization templates in LLM-based solver modeling, which represents the template library as a directed acyclic graph and formulates retrieval as a sequential decision process over template nodes. A deep Q-network learns a reusable traversal policy from template-level judgments and execution feedback, reducing reliance on fixed heuristic traversal through feedback-driven graph navigation. To support extensible retrieval, COMPASS embeds problem descriptions and template nodes in a shared semantic space, allowing newly inserted templates to be scored from their text descriptions without redefining the action space. Experiments across multiple LLM backbones and optimization-modeling benchmarks demonstrate that COMPASS consistently improves over heuristic structured retrieval, with the largest gains on complex and mixed-category problems.

cs.AI↗

Network-Assisted Full-Duplex Cell-Free Massive MIMO Systems Under Infeasible Circumstances

Cell-free massive multiple-input multiple-output is a potential candidate for future networks with pervasive connectivity by utilizing coherent joint transmission and distributed antenna arrays. This paper studies the exploitation of full-duplex communication for a distributed antenna array. Specifically, we derive a closed-form expression for the uplink and downlink ergodic spectral efficiency (SE) for a network where the APs can flexibly operate in either the full-duplex or half-duplex mode with linear processing and Rayleigh fading channels. A long-term total SE maximization problem is formulated subject to a network operation model and individual SE requirements with limited power budget. Due to the intrinsic nonconvexity and infeasible circumstances where some UEs might not be able to achieve the rate requirements, we adapt differential evolution to design a low computational complexity algorithm that can attain good power allocation and network operation mode in polynomial time. Numerical results demonstrate the effectiveness of our system design and proposed algorithm over state-of-the-art benchmarks with satisfactory service to the majority of UEs, although several ones may be unscheduled under harsh conditions.

eess.SP↗

Energy Efficiency Maximization for Integrated Sensing and Communications in Satellite-UAV MIMO Systems

This paper investigates energy efficiency maximization in an integrated sensing and communication framework for satellite-UAV MIMO systems, where a LEO satellite and a UAV simultaneously serve ground users and perform target sensing. Both the satellite and UAV are equipped with uniform planar arrays of transmit antennas, enabling a distributed multi-user and multi-target architecture. We derive the achievable downlink throughput by considering that the high-altitude satellite maintains a line-of-sight (LoS) link with users, while adopting a probabilistic model for the UAV that accounts for the likelihood of both LoS and non-line-of-sight conditions. The energy efficiency maximization problem is formulated as a complex non-convex optimization problem, subject to power constraints, quality of service (QoS) requirements, and beampattern gain constraints for accurate sensing. To tackle this challenge, we propose an efficient alternating optimization algorithm capable of handling the complex search space and QoS guarantees. Numerical results across diverse scenarios with multiple users demonstrate that the proposed method achieves high energy efficiency while meeting both communication and sensing performance targets.

cs.IT↗

Single- and Multi-Objective Stochastic Optimization for Next-Generation Networks in the Generative AI and Quantum Computing Era

Next Generation (NG) networks move beyond simply connecting devices to creating an ecosystem of connected intelligence, especially with the support of generative Artificial Intelligence (AI) and quantum computation. These systems are expected to handle large-scale deployments and high-density networks with diverse functionalities. As a result, there is an increasing demand for efficient and intelligent algorithms that can operate under uncertainty from both propagation environments and networking systems. Traditional optimization methods often depend on accurate theoretical models of data transmission, but in real-world NG scenarios, they suffer from high computational complexity in large-scale settings. Stochastic Optimization (SO) algorithms, designed to accommodate extremely high density and extensive network scalability, have emerged as a powerful solution for optimizing wireless networks. This includes various categories that range from model-based approaches to learning-based approaches. These techniques are capable of converging within a feasible time frame while addressing complex, large-scale optimization problems. However, there is currently limited research on SO applied for NG networks, especially the upcoming Sixth-Generation (6G). In this survey, we emphasize the relationship between NG systems and SO by eight open questions involving the background, key features, and lesson learned. Overall, our study starts by providing a detailed overview of both areas, covering fundamental and widely used SO techniques, spanning from single to multi-objective signal processing. Next, we explore how different algorithms can solve NG challenges, such as load balancing, optimizing energy efficiency, improving spectral efficiency, or handling multiple performance trade-offs. Lastly, we highlight the challenges in the current research and propose new directions for future studies.

cs.IT↗

Active and Passive Beamforming Designs for SER Minimization in RIS-Assisted MIMO Systems

This research exploits the applications of reconfigurable intelligent surface (RIS)-assisted multiple input multiple output (MIMO) systems, specifically addressing the enhancement of communication reliability with modulated signals. Specifically, we first derive the analytical downlink symbol error rate (SER) of each user as a multivariate function of both the phase-shift and beamforming vectors. The analytical SER enables us to obtain insights into the synergistic dynamics between the RIS and MIMO communication. We then introduce a novel average SER minimization problem subject to the practical constraints of the transmitted power budget and phase shift coefficients, which is NP-hard. By incorporating the differential evolution (DE) algorithm as a pivotal tool for optimizing the intricate active and passive beamforming variables in RIS-assisted communication systems, the non-convexity of the considered SER optimization problem can be effectively handled. Furthermore, an efficient local search is incorporated into the DE algorithm to overcome the local optimum, and hence offer low SER and high communication reliability. Monte Carlo simulations validate the analytical results and the proposed optimization framework, indicating that the joint active and passive beamforming design is superior to the other benchmarks.

cs.IT↗