arXiv subjects

Miao Du

Publications and source records attributed to Miao Du.

15 recordsLinked to original sources

arXiv2025

Decentralized Semantic Federated Learning for Real-Time Public Safety Tasks: Challenges, Methods, and Directions

Public safety tasks rely on the collaborative functioning of multiple edge devices (MEDs) and base stations (BSs) in different regions, consuming significant communication energy and computational resources to execute critical operations like fire monitoring and rescue missions. Traditional federated edge computing (EC) methods require frequent central communication, consuming substantial energy and struggling with resource heterogeneity across devices, networks, and data. To this end, this paper introduces a decentralized semantic federated learning (DSFL) framework tailored for large-scale wireless communication systems and heterogeneous MEDs. The framework incorporates a hierarchical semantic communication (SC) scheme to extend EC coverage and reduce communication overhead. Specifically, the lower layer optimizes intra-BS communication through task-specific encoding and selective transmission under constrained networks, while the upper layer ensures robust inter-BS communication via semantic aggregation and distributed consensus across different regions. To further balance communication costs and semantic accuracy, an energy-efficient aggregation scheme is developed for both intra-BS and inter-BS communication. The effectiveness of the DSFL framework is demonstrated through a case study using the BoWFire dataset, showcasing its potential in real-time fire detection scenarios. Finally, we outlines open issues for edge intelligence and SC in public safety tasks.

cs.DC↗

arXiv2024

In-depth Understanding of the Band Alignment and Interface States Scenario in Bi$_2$O$_2$Se/SrTiO$_3$ Ultrathin Heterojunction

Bismuth oxyselenide (Bi$_2$O$_2$Se), a novel quasi-2D charge-carrying semiconductor, is hailed as one of the best emerging platforms for the next generation semiconductor devices. Recent efforts on developing diverse Bi$_2$O$_2$Se heterojunctions have produced extensive potential applications in electronics and optoelectronics. In-depth understanding of the band alignment and especially interface dynamics is, however, still challenging. In this work, a comprehensive experimental investigation on the band alignment is performed by a high-resolution X-ray photoelectron spectrometer (HRXPS), and the properties of interface states are also fully discussed. The results show that the ultrathin film Bi$_2$O$_2$Se grown on SrTiO$_3$ (TiO$_2$ (001) termination) exhibits Type-I (straddling gap) band alignment with a valence band offset (VBO) of about 1.77\pm0.04 eV and conduction band offset (CBO) of about 0.68\pm0.04 eV. However, further considering the contribution of the interface states, the bands on the interface present a herringbone configuration due to sizable build-in electric fields, which is significantly different from the conventional band alignment. In this sense, our results provide an insightful guidance to the development of high-efficiency electronic and optoelectronic devices, specifically of the devices where the charge transfer is highly sensitive to interface states.

cond-mat.mtrl-sci↗

arXiv2024

On a class of planar Schrödinger-Poisson system with a bounded potential well

In this paper, we deal with the planar Schrödinger-Poisson system \begin{equation*}\begin{cases} -Δu + V(x) u + ϕu = b|u|^{p-2} u \ &\text{in}\ \mathbb{R}^{2},\\Δϕ= u^{2} &\text{in}\ \mathbb{R}^{2},\end{cases} \end{equation*} where $b \geq 0$, $p > 2 $ and $V \in C(\mathbb{R}^2, \mathbb{R})$ is a potential function with $\inf_{\mathbb{R}^2} V >0$. Suppose moreover that $V$ exhibits a bounded potential well in the sense that $\lim_{|x|\rightarrow \infty} V(x)$ exists and is equal to $\sup_{\mathbb{R}^2} V$. By using variational methods, we obtain the existence of ground state solutions for this system in the case where $p \geq 3$. Furthermore, we also present a minimax characterization of ground state solutions. The main feature of this work is that we do not assume any periodicity or symmetry condition on the external potential $V$, which is essential to establish the compactness condition of Cerami sequences.

math.AP↗

arXiv2024

Ground states of planar Schrödinger-Poisson systems with an unbounded potential

In this paper, we deal with a class of planar Schrödinger-Poisson systems, namely, $-Δu+V(x)u+\fracγ{2π}\bigl(\log(|\cdot|)\ast|u|^{2}\bigr)u=b|u|^{p-2}u\ \text{in}\ \mathbb{R}^{2}$, where $γ> 0$, $b \geq 0$, $p>2$ and $V \in C(\mathbb{R}^2, \mathbb{R})$ is an unbounded potential function with $\inf_{\mathbb{R}^2} V >0$. Suppose moreover that the potential $V$ satisfies $\left|\{x \in \mathbb{R}^2:\: V(x)\leq M\}\right| < \infty$ for every $M>0$, we establish the existence of ground state solutions for this system via variational methods. Furthermore, we also explore the minimax characterization of ground state solutions. Our main results can be viewed as a counterpart of the result from Molle and Sardilli (Proc. Edinb. Math. Soc. 65:1133-1146, 2022), where the authors studied the existence of ground state solutions for the above planar Schrödinger-Poisson system in the case where $b>0$ and $p >4$.

math.AP↗

arXiv2022

Organic metallic epsilon-near-zero materials with large ultrafast optical nonlinearity

Epsilon-near-zero (ENZ) materials have shown significant potential for nonlinear optical applications due to their ultrafast hot carriers and consequent optical nonlinearity enhancement. Modified poly(3,4-ethylenedioxythiophene) (PEDOT) films show metallic characteristics and a resultant ENZ wavelength near 1550nm through polar solvent treatment and annealing. The metallic PEDOT film exhibits an intrinsic optical nonlinear response that is comparable to gold and 100-fold higher than typical inorganic semiconductor ENZ materials due to π-conjugated delocalized electrons. Hot carriers generate a 22-fold increase in the optical nonlinearity coefficient of metallic PEDOT films at 1550 nm. Hot holes in metallic PEDOT films have a smaller enhancement multiple of carrier temperature and a longer relaxation time than hot electrons in inorganic ENZ materials due to the larger imaginary permittivity and hot-phonon bottleneck for carrier cooling. Our findings suggest that π-conjugated ENZ polymer may have unique ultrafast and nonlinear optical properties compared to inorganic ENZ materials, enabling new possibilities in on-chip nanophotonic devices, nonlinear optics, and plasmonics.

physics.optics↗

arXiv2020

Positive solutions for the Schrödinger-Poisson system with steep potential well

In this paper, we consider the following Schrödinger-Poisson system \begin{equation*} \begin{cases} - Δu+λV(x)u+ μϕu=|u|^{p-2}u &\text{in $\mathbb{R}^3$},\cr -Δϕ=u^{2} &\text{in $\mathbb{R}^3$}, \end{cases} \end{equation*} where $λ,\:μ>0$ are real parameters and $2<p<6$. Suppose that $V(x)$ represents a potential well with the bottom $V^{-1}(0)$, the system has been widely studied in the case $4\leq p<6$. In contrast, no existence result of solutions is available for the case $2<p<4$ due to the presence of the nonlocal term $ϕu$. With the aid of the truncation technique and the parameter-dependent compactness lemma, we first prove the existence of positive solutions for $λ$ large and $μ$ small in the case $2<p<4$. Then we obtain the nonexistence of nontrivial solutions for $λ$ large and $μ$ large in the case $2<p\leq3$. Finally, we explore the decay rate of the positive solutions as $|x| \rightarrow \infty$ as well as their asymptotic behavior as $λ\rightarrow \infty$ and $μ\rightarrow 0$.

math.AP↗

arXiv2019

Fully Dense Neural Network for the Automatic Modulation Recognition

Nowadays, we mainly use various convolution neural network (CNN) structures to extract features from radio data or spectrogram in AMR. Based on expert experience and spectrograms, they not only increase the difficulty of preprocessing, but also consume a lot of memory. In order to directly use in-phase and quadrature (IQ) data obtained by the receiver and enhance the efficiency of network extraction features to improve the recognition rate of modulation mode, this paper proposes a new network structure called Fully Dense Neural Network (FDNN). This network uses residual blocks to extract features, dense connect to reduce model size, and adds attentions mechanism to recalibrate. Experiments on RML2016.10a show that this network has a higher recognition rate and lower model complexity. And it shows that the FDNN model with dense connections can not only extract features effectively but also greatly reduce model parameters, which also provides a significant contribution for the application of deep learning to the intelligent radio system.

eess.SP↗

arXiv2019

Multi-layer Attention Mechanism for Speech Keyword Recognition

As an important part of speech recognition technology, automatic speech keyword recognition has been intensively studied in recent years. Such technology becomes especially pivotal under situations with limited infrastructures and computational resources, such as voice command recognition in vehicles and robot interaction. At present, the mainstream methods in automatic speech keyword recognition are based on long short-term memory (LSTM) networks with attention mechanism. However, due to inevitable information losses for the LSTM layer caused during feature extraction, the calculated attention weights are biased. In this paper, a novel approach, namely Multi-layer Attention Mechanism, is proposed to handle the inaccurate attention weights problem. The key idea is that, in addition to the conventional attention mechanism, information of layers prior to feature extraction and LSTM are introduced into attention weights calculations. Therefore, the attention weights are more accurate because the overall model can have more precise and focused areas. We conduct a comprehensive comparison and analysis on the keyword spotting performances on convolution neural network, bi-directional LSTM cyclic neural network, and cyclic neural network with the proposed attention mechanism on Google Speech Command datasets V2 datasets. Experimental results indicate favorable results for the proposed method and demonstrate the validity of the proposed method. The proposed multi-layer attention methods can be useful for other researches related to object spotting.

cs.LG↗

arXiv2018

Scalar Quantization as Sparse Least Square Optimization

Quantization can be used to form new vectors/matrices with shared values close to the original. In recent years, the popularity of scalar quantization for value-sharing applications has been soaring as it has been found huge utilities in reducing the complexity of neural networks. Existing clustering-based quantization techniques, while being well-developed, have multiple drawbacks including the dependency of the random seed, empty or out-of-the-range clusters, and high time complexity for a large number of clusters. To overcome these problems, in this paper, the problem of scalar quantization is examined from a new perspective, namely sparse least square optimization. Specifically, inspired by the property of sparse least square regression, several quantization algorithms based on $l_1$ least square are proposed. In addition, similar schemes with $l_1 + l_2$ and $l_0$ regularization are proposed. Furthermore, to compute quantization results with a given amount of values/clusters, this paper designed an iterative method and a clustering-based method, and both of them are built on sparse least square. The paper shows that the latter method is mathematically equivalent to an improved version of k-means clustering-based quantization algorithm, although the two algorithms originated from different intuitions. The algorithms proposed were tested with three types of data and their computational performances, including information loss, time consumption, and the distribution of the values of the sparse vectors, were compared and analyzed. The paper offers a new perspective to probe the area of quantization, and the algorithms proposed can outperform existing methods especially under some bit-width reduction scenarios, when the required post-quantization resolution (number of values) is not significantly lower than the original number.

cs.LG↗

arXiv2017

Ground states and high energy solutions of the planar Schrödinger-Poisson system

In this paper, we are concerned with the Schrödinger-Poisson system \begin{equation} (0.1)\qquad -Δu + u +ϕu = |u|^{p-2}u \quad \text{in}\ \mathbb{R}^{d},\qquad Δϕ= u^{2} \quad \text{in}\ \mathbb{R}^{d}. \end{equation} Due to its relevance in physics, the system has been extensively studied and is quite well understood in the case $d \ge 3$. In contrast, much less information is available in the planar case $d=2$ which is the focus of the present paper. It has been observed by Cingolani and the second author \cite{Cingolani-Weth-2016} that the variational structure of $(0.1)$ differs substantially in the case $d=2$ and leads to a richer structure of the set of solutions. However, the variational approach of \cite{Cingolani-Weth-2016} is restricted to the case $p \ge 4$ which excludes some physically relevant exponents. In the present paper, we remove this unpleasant restriction and explore the more complicated underlying functional geometry in the case $2<p<4$ with a different variational approach.

math.AP↗

arXiv2016

Opaque Response Generation Enabling Automatic Creation of Virtual Services for Service Virtualisation

Service virtualisation is a method to create virtual service models that can mimic interaction behaviors between a system under test and the target system. With service virtualisation, the development team can get access to the production-like conditions whenever and however many times they need, enabling frequent and comprehensive testing. Previous techniques for service virtualisation have relied on explicitly modelling the target services by a service expert and require detailed knowledge of message protocol and structure. However, neither of these are necessarily available. In this thesis, we introduce our novel opaque response generation approach. This approach enables services to be virtualised automatically without any expert knowledge or documentation of system protocol and interaction behaviours. Given a collection of interactions exchanged between a system under test and a target real service, our approach can 1) organise the same type of interactions into the same cluster and derive a cluster prototype for each cluster; 2) search a given incoming request for its the most similar request in the interaction library; 3) learn knowledge from the incoming request and the recorded interaction; and 4) generate a response. A framework and proof-of-concept implementation of our opaque response generation approach is described. Experimental results show our opaque response generation approach is able to automatically generate accurate responses in real time with an accuracy rate over 99% on average.

cs.SE↗

arXiv2016

Opaque Service Virtualisation: A Practical Tool for Emulating Endpoint Systems

Large enterprise software systems make many complex interactions with other services in their environment. Developing and testing for production-like conditions is therefore a very challenging task. Current approaches include emulation of dependent services using either explicit modelling or record-and-replay approaches. Models require deep knowledge of the target services while record-and-replay is limited in accuracy. Both face developmental and scaling issues. We present a new technique that improves the accuracy of record-and-replay approaches, without requiring prior knowledge of the service protocols. The approach uses Multiple Sequence Alignment to derive message prototypes from recorded system interactions and a scheme to match incoming request messages against prototypes to generate response messages. We use a modified Needleman-Wunsch algorithm for distance calculation during message matching. Our approach has shown greater than 99% accuracy for four evaluated enterprise system messaging protocols. The approach has been successfully integrated into the CA Service Virtualization commercial product to complement its existing techniques.

cs.SE↗

arXiv2016

Enhanced Playback of Automated Service Emulation Models Using Entropy Analysis

Service virtualisation is a supporting tool for DevOps to generate interactive service models of dependency systems on which a system-under-test relies. These service models allow applications under development to be continuously tested against production-like conditions. Generating these virtual service models requires expert knowledge of the service protocol, which may not always be available. However, service models may be generated automatically from network traces. Previous work has used the Needleman-Wunsch algorithm to select a response from the service model to play back for a live request. We propose an extension of the Needleman-Wunsch algorithm, which uses entropy analysis to automatically detect the critical matching fields for selecting a response. Empirical tests against four enterprise protocols demonstrate that entropy weighted matching can improve response accuracy.

cs.SE↗

arXiv2016

From Network Traces to System Responses: Opaquely Emulating Software Services

Enterprise software systems make complex interactions with other services in their environment. Developing and testing for production-like conditions is therefore a challenging task. Prior approaches include emulations of the dependency services using either explicit modelling or record-and-replay approaches. Models require deep knowledge of the target services while record-and-replay is limited in accuracy. We present a new technique that improves the accuracy of record-and-replay approaches, without requiring prior knowledge of the services. The approach uses multiple sequence alignment to derive message prototypes from recorded system interactions and a scheme to match incoming request messages against message prototypes to generate response messages. We introduce a modified Needleman-Wunsch algorithm for distance calculation during message matching, wildcards in message prototypes for high variability sections, and entropy-based weightings in distance calculations for increased accuracy. Combined, our new approach has shown greater than 99% accuracy for four evaluated enterprise system messaging protocols.

cs.SE↗

arXiv2015

Adaptive Automation: Leveraging Machine Learning to Support Uninterrupted Automated Testing of Software Applications

Checking software application suitability using automated software tools has become a vital element for most organisations irrespective of whether they produce in-house software or simply customise off-the-shelf software applications for internal use. As software solutions become ever more complex, the industry becomes increasingly dependent on software automation tools, yet the brittle nature of the available software automation tools limits their effectiveness. Companies invest significantly in obtaining and implementing automation software but most of the tools fail to deliver when the cost of maintaining an effective automation test suite exceeds the cost and time that would have otherwise been spent on manual testing. A failing in the current generation of software automation tools is they do not adapt to unexpected modifications and obstructions without frequent (and time expensive) manual interference. Such issues are commonly acknowledged amongst industry practitioners, yet none of the current generation of tools have leveraged the advances in machine learning and artificial intelligence to address these problems. This paper proposes a framework solution that utilises machine learning concepts, namely fuzzy matching and error recovery. The suggested solution applies adaptive techniques to recover from unexpected obstructions that would otherwise have prevented the script from proceeding. Recovery details are presented to the user in a report which can be analysed to determine if the recovery procedure was acceptable and the framework will adapt future runs based on the decisions of the user. Using this framework, a practitioner can run the automated suits without human intervention while minimising the risk of schedule delays.

cs.SE↗