arXiv subjects

Jinkun Zhang

Publications and source records attributed to Jinkun Zhang.

12 recordsLinked to original sources

arXiv2026

Design Insights into Partition Placement and Routing for DNN Inference in Multi-Hop Edge Networks

Partitioned DNN inference is a promising approach for latency-sensitive intelligent services in edge networks, since it allows different parts of a model to be executed across end devices, edge servers, and the cloud. However, in a multi-hop edge network, partition placement and inference traffic routing are inherently coupled: raw inputs, intermediate features, and final outputs may have very different sizes, while candidate nodes also differ in computation capability. In addition, both communication and computation delays can become congestion-dependent under load. In this paper, we study joint partition placement and routing for fixed-partition DNN inference over heterogeneous multi-hop edge networks. We consider a small number of DNN partitions, each placed at exactly one node without replication, and formulate a congestion-aware mixed discrete--continuous optimization problem that captures both routing and execution costs. To solve it, we develop a practical alternating framework that couples partition placement with congestion-aware forwarding updates. Through numerical evaluation on hierarchical, regular, synthetic irregular, and real backbone-inspired topologies, we show that split flexibility is particularly important in IoT--edge--cloud settings, while congestion-aware refinement becomes increasingly beneficial as the offered load grows. We further illustrate how the preferred operating point depends on the communication--computation tradeoff.

cs.NI↗

arXiv2025

Decentralized AI Service Placement, Selection and Routing in Mobile Networks

The rapid development and usage of large-scale AI models by mobile users will dominate the traffic load in future communication networks. The advent of AI technology also facilitates a decentralized AI ecosystem where small organizations or even individuals can host AI services. In such scenarios, AI service (models) placement, selection, and request routing decisions are tightly coupled, posing a challenging yet fundamental trade-off between service quality and service latency, especially when considering user mobility. Existing solutions for related problems in mobile edge computing (MEC) and data-intensive networks fall short due to restrictive assumptions about network structure or user mobility. To bridge this gap, we propose a decentralized framework that jointly optimizes AI service placement, selection, and request routing. In the proposed framework, we use traffic tunneling to support user mobility without costly AI service migrations. To account for nonlinear queuing delays, we formulate a nonconvex problem to optimize the trade-off between service quality and end-to-end latency. We derive the node-level KKT conditions and develop a decentralized Frank--Wolfe algorithm with a novel messaging protocol. Numerical evaluations validate the proposed approach and show substantial performance improvements over existing methods.

cs.NI↗

arXiv2025

Delay-optimal Congestion-aware Routing and Computation Offloading in Arbitrary Network

Emerging edge computing paradigms enable heterogeneous devices to collaborate on complex computation applications. However, for arbitrary heterogeneous edge networks, delay-optimal forwarding and computation offloading remains an open problem. In this paper, we jointly optimize data/result routing and computation placement in arbitrary networks with heterogeneous node capabilities, and congestion-dependent nonlinear transmission and processing delay. Despite the non-convexity of the formulated problem, based on analyzing the KKT condition, we provide a set of sufficient optimality conditions that solve the problem globally. To provide the insights for such global optimality, we show that the proposed non-convex problem is geodesic-convex with mild assumptions. We also show that the proposed sufficient optimality condition leads to a lower hemicontinuous solution set, providing stability against user-input perturbation. We then extend the framework to incorporate utility-based congestion control and fairness. A fully distributed algorithm is developed to converge to the global optimum. Numerical results demonstrate significant improvements over multiple baselines algorithms.

cs.NI↗

arXiv2025

Teleology-Driven Affective Computing: A Causal Framework for Sustained Well-Being

Affective computing has made significant strides in emotion recognition and generation, yet current approaches mainly focus on short-term pattern recognition and lack a comprehensive framework to guide affective agents toward long-term human well-being. To address this, we propose a teleology-driven affective computing framework that unifies major emotion theories (basic emotion, appraisal, and constructivist approaches) under the premise that affect is an adaptive, goal-directed process that facilitates survival and development. Our framework emphasizes aligning agent responses with both personal/individual and group/collective well-being over extended timescales. We advocate for creating a "dataverse" of personal affective events, capturing the interplay between beliefs, goals, actions, and outcomes through real-world experience sampling and immersive virtual reality. By leveraging causal modeling, this "dataverse" enables AI systems to infer individuals' unique affective concerns and provide tailored interventions for sustained well-being. Additionally, we introduce a meta-reinforcement learning paradigm to train agents in simulated environments, allowing them to adapt to evolving affective concerns and balance hierarchical goals - from immediate emotional needs to long-term self-actualization. This framework shifts the focus from statistical correlations to causal reasoning, enhancing agents' ability to predict and respond proactively to emotional challenges, and offers a foundation for developing personalized, ethically aligned affective systems that promote meaningful human-AI interactions and societal well-being.

cs.HC↗

arXiv2024

LOAM: Low-latency Communication, Caching, and Computation Placement in Data-Intensive Computing Networks

Deploying data- and computation-intensive applications such as large-scale AI into heterogeneous dispersed computing networks can significantly enhance application performance by mitigating bottlenecks caused by limited network resources, including bandwidth, storage, and computing power. However, current resource allocation methods in dispersed computing do not provide a comprehensive solution that considers arbitrary topology, elastic resource amount, reuse of computation results, and nonlinear congestion-dependent optimization objectives. In this paper, we propose LOAM, a low-latency joint communication, caching, and computation placement framework with a rigorous analytical foundation that incorporates the above aspects. We tackle the NP-hard aggregated cost minimization problem with two methods: an offline method with a 1/2 approximation and an online adaptive method with a bounded gap from the optimum. Through extensive simulation, the proposed framework outperforms multiple baselines in both synthesis and real-world network scenarios.

cs.NI↗

arXiv2024

Delay-Optimal Forwarding and Computation Offloading for Service Chain Tasks

Emerging edge computing paradigms enable heterogeneous devices to collaborate on complex computation applications. However, for congestible links and computing units, delay-optimal forwarding and offloading for service chain tasks (e.g., DNN with vertical split) in edge computing networks remains an open problem. In this paper, we formulate the service chain forwarding and offloading problem with arbitrary topology and heterogeneous transmission/computation capability, and aim to minimize the aggregated network cost. We consider congestion-aware nonlinear cost functions that cover various performance metrics and constraints, such as average queueing delay with limited processor capacity. We solve the non-convex optimization problem globally by analyzing the KKT condition and proposing a sufficient condition for optimality. We then propose a distributed algorithm that converges to the global optimum. The algorithm adapts to changes in input rates and network topology, and can be implemented as an online algorithm. Numerical evaluation shows that our method significantly outperforms baselines in multiple network instances, especially in congested scenarios.

cs.NI↗

arXiv2024

Robot Body Schema Learning from Full-body Extero/Proprioception Sensors

For a robot, its body structure is an a-prior knowledge when it is designed. However, when such information is not available, can a robot recognize it by itself? In this paper, we aim to grant a robot such ability to learn its body structure from exteroception and proprioception data collected from on-body sensors. By a novel machine learning method, the robot can learn a binary Heterogeneous Dependency Matrix from its sensor readings. We showed such matrix is equivalent to a Heterogeneous out-tree structure which can uniquely represent the robot body topology. We explored the properties of such matrix and the out-tree, and proposed a remedy to fix them when they are contaminated by partial observability or data noise. We ran our algorithm on 6 different robots with different body structures in simulation and 1 real robot. Our algorithm correctly recognized their body structures with only on-body sensor readings but no topology prior knowledge.

cs.RO↗

arXiv2023

Delay-Optimal Service Chain Forwarding and Offloading in Collaborative Edge Computing

Collaborative edge computing (CEC) is an emerging paradigm for heterogeneous devices to collaborate on edge computation jobs. For congestible links and computing units, delay-optimal forwarding and offloading for service chain tasks (e.g., DNN with vertical split) in CEC remains an open problem. In this paper, we formulate the service chain forwarding and offloading in CEC with arbitrary topology and heterogeneous transmission/computation capability, and aim to minimize the network aggregated cost. We consider congestion-aware nonlinear cost functions that cover various performance metrics and constraints, such as average queueing delay with limited processor capacity. We solve the non-convex optimization problem globally by analyzing the KKT condition and proposing a sufficiency optimality condition. We propose a polynomial-time distributed algorithm that converges to the global optimum. The algorithm adapts to changes in input rates and network topology, and can be implemented as an online algorithm. Numerical evaluation shows that our method significantly outperforms baselines in multiple network instances, especially in congested scenarios.

cs.DC↗

arXiv2023

Congestion-aware routing and content placement in elastic cache networks

Caching can be leveraged to significantly improve network performance and mitigate congestion. However, characterizing the optimal tradeoff between routing cost and cache deployment cost remains an open problem. In this paper, for a network with arbitrary topology and congestion-dependent nonlinear cost functions, we aim to jointly determine the cache deployment, content placement, and hop-by-hop routing strategies, so that the sum of routing cost and cache deployment cost is minimized. We tackle this NP-hard problem starting with a fixed-routing setting, and then to a general dynamic-routing setting. For the fixed-routing setting, a Gradient-combining Frank-Wolfe algorithm with $(\frac{1}{2},1)$-approximation is presented. For the general dynamic-routing setting, we obtain a set of KKT necessary optimal conditions, and devise a distributed and adaptive online algorithm based on the conditions. We demonstrate via extensive simulation that our algorithms significantly outperform a number of baseline techniques.

cs.NI↗

arXiv2022

Optimal Congestion-aware Routing and Offloading in Collaborative Edge Computing

Collaborative edge computing (CEC) is an emerging paradigm where heterogeneous edge devices collaborate to fulfill computation tasks, such as model training or video processing, by sharing communication and computation resources. Nevertheless, the optimal data/result routing and computation offloading strategy in CEC with arbitrary topology still remains an open problem. In this paper, we formulate the flow model of partial-offloading and multi-hop routing for arbitrarily divisible tasks, where each node individually decides its routing/offloading strategy. In contrast to most existing works, our model applies to tasks with non-negligible result size, and allows data sources to be distinct from the result destination. We propose a network-wide cost minimization problem with congestion-aware convex cost functions for communication and computation. Such convex cost covers various performance metrics and constraints, such as average queueing delay with limited processor capacity. Although the problem is non-convex, we provide necessary conditions and sufficient conditions for the global-optimal solution, and devise a fully distributed algorithm that converges to the optimum in polynomial time, allows asynchronous individual updating, and is adaptive to changes in task pattern. Numerical evaluation shows that our proposed method significantly outperforms other baseline algorithms in multiple network instances, especially in congested scenarios.

cs.DC↗

arXiv2022

Result and Congestion Aware Optimal Routing and Partial Offloading in Collaborative Edge Computing

Collaborative edge computing (CEC) is an emerging paradigm where heterogeneous edge devices (stakeholders) collaborate to fulfill computation tasks, such as model training or video processing, by sharing communication and computation resources. Nevertheless, the optimal data/result routing and computation offloading strategy in CEC with arbitrary topology still remains an open problem. In this paper, we formulate a partial-offloading and multi-hop routing model for arbitrarily divisible tasks. Each node individually decides the computation of the received data and the forwarding of data/result traffic. In contrast to most existing works, our model applies to tasks with non-negligible result size, and enables separable data sources and result destinations. We propose a network-wide cost minimization problem with congestion-aware cost to jointly optimize routing and computation offloading. This problem covers various performance metrics and constraints, such as average queueing delay with limited processor capacity. Although the problem is non-convex, we provide non-trivial necessary and sufficient conditions for the global-optimal solution, and devise a fully distributed algorithm that converges to the optimum in polynomial time, allows asynchronous individual updating, and is adaptive to changes in network topology or task pattern. Numerical evaluation shows that our proposed method significantly outperforms other baseline algorithms in multiple network instances, especially in congested scenarios.

cs.NI↗

arXiv2021

Transmission Delay Minimization via Joint Power Control and Caching in Wireless HetNets

A fundamental challenge in wireless heterogeneous networks (HetNets) is to effectively utilize the limited transmission and storage resources in the presence of increasing deployment density and backhaul capacity constraints. To alleviate bottlenecks and reduce resource consumption, we design optimal caching and power control algorithms for multi-hop wireless HetNets. We formulate a joint optimization framework to minimize the average transmission delay as a function of the caching variables and the signal-to-interference-plus-noise ratios (SINR) which are determined by the transmission powers, while explicitly accounting for backhaul connection costs and the power constraints. Using convex relaxation and rounding, we obtain a reduced-complexity formulation (RCF) of the joint optimization problem, which can provide a constant factor approximation to the globally optimal solution. We then solve RCF in two ways: 1) alternating optimization of the power and caching variables by leveraging biconvexity, and 2) joint optimization of power control and caching. We characterize the necessary (KKT) conditions for an optimal solution to RCF, and use strict quasi-convexity to show that the KKT points are Pareto optimal for RCF. We then devise a subgradient projection algorithm to jointly update the caching and power variables, and show that under appropriate conditions, the algorithm converges at a linear rate to the local minima of RCF, under general SINR conditions. We support our analytical findings with results from extensive numerical experiments.

cs.IT↗