arXiv Science⌕ Search

arXiv subjects

Kengo Urata

Publications and source records attributed to Kengo Urata.

2 recordsLinked to original sources

Sign Primal-Dual Gradient Dynamics for Supply-Demand Coordination under Binary Human Feedback

This paper addresses supply-demand coordination, in which a supplier sets prices to balance the demands of multiple buyers against limited resources. Existing studies typically solve the supply-demand coordination problem with a distributed algorithm called primal-dual gradient dynamics (PDGD), assuming that each buyer discloses the value of its own utility or marginal utility function. In contrast, we assume that each buyer discloses only binary information, namely whether it wants more or less of the current amount, and propose Sign PDGD. Sign PDGD is a PDGD algorithm in which the buyers' update depends only on the sign information. We analyze the convergence of Sign PDGD by interpreting it as a sliding mode control (SMC). For a quadratic cost and linear constraints, we show that under a gain condition, the primal and dual variables converge to the optimal solution. Numerical simulations on a toy resource-sharing problem confirm the two-phase behavior of SMC and show that the primal and dual variables approach the optimal solution.

math.OC↗

Chat-Driven Optimal Management for Virtual Network Services

This paper proposes a chat-driven network management framework that integrates natural language processing (NLP) with optimization-based virtual network allocation, enabling intuitive and reliable reconfiguration of virtual network services. Conventional intent-based networking (IBN) methods depend on statistical language models to interpret user intent but cannot guarantee the feasibility of generated configurations. To overcome this, we develop a two-stage framework consisting of an Interpreter, which extracts intent from natural language prompts using NLP, and an Optimizer, which computes feasible virtual machine (VM) placement and routing via an integer linear programming. In particular, the Interpreter translates user chats into update directions, i.e., whether to increase, decrease, or maintain parameters such as CPU demand and latency bounds, thereby enabling iterative refinement of the network configuration. In this paper, two intent extractors, which are a Sentence-BERT model with support vector machine (SVM) classifiers and a large language model (LLM), are introduced. Experiments in single-user and multi-user settings show that the framework dynamically updates VM placement and routing while preserving feasibility. The LLM-based extractor achieves higher accuracy with fewer labeled samples, whereas the Sentence-BERT with SVM classifiers provides significantly lower latency suitable for real-time operation. These results underscore the effectiveness of combining NLP-driven intent extraction with optimization-based allocation for safe, interpretable, and user-friendly virtual network management.

cs.NI↗