Searcharxiv⌕ Search

arXiv subjects

Junfeng Li

Publications and source records attributed to Junfeng Li.

At least 55 records · Page 3Linked to original sources

The space-time estimates for the Schrödinger equation

In this paper, we studied the space-time estimates for the solution to the Schrödinger equation. By polynomial partitioning, induction arguments, bilinear to linear arguments and broad norm estimates, we set up several maximal estimates for the Schrödinger equation with high-frequency input data. By these maximal estimates, we obtain the sharp global space-time estimate when $n=2$ and improve the known results in the critical cases when $n\geq 3$. The maximal estimate for $n=2$ is also used to extend the results of the local space-time estimates for the solution to the Schrödinger equation.

math.CA↗

$L_x^p\rightarrow L^q_{x,u}$ estimates for dilated averages over planar curves

In this paper, we consider the $L_x^p(\mathbb{R}^2)\rightarrow L_{x,u}^q(\mathbb{R}^2\times [1,2])$ estimate for the operator $T$ along a dilated plane curve $(ut,uγ(t))$, where $$Tf(x,u):=\int_{0}^{1}f(x_1-ut,x_2-u γ(t))\,\textrm{d}t,$$ $x:=(x_1,x_2)$ and $γ$ is a general plane curve satisfying some suitable smoothness and curvature conditions. We show that $T$ is $L_x^p(\mathbb{R}^2)$ to $L_{x,u}^q(\mathbb{R}^2\times [1,2])$ bounded whenever $(\frac{1}{p},\frac{1}{q})\in \square \cup \{(0,0)\}\cup \{(\frac{2}{3},\frac{1}{3})\}$ and $1+(1 +ω)(\frac{1}{q}-\frac{1}{p})>0$, where the trapezium $\square:=\{(\frac{1}{p},\frac{1}{q}):\ \frac{2}{p}-1\leq\frac{1}{q}\leq \frac{1}{p}, \frac{1}{q}>\frac{1}{3p}, \frac{1}{q}>\frac{1}{p}-\frac{1}{3}\}$ and $ω:=\limsup_{t\rightarrow 0^{+}}\frac{\ln|γ(t)|}{\ln t}$. This result is sharp except for some borderline cases. On the other hand, in a smaller $(\frac{1}{p},\frac{1}{q})$ region, we also obtain the almost sharp estimate $T : L_x^p(\mathbb{R}^2)\rightarrow L_{x}^q(\mathbb{R}^2)$ uniformly for $u\in [1,2]$. These results imply that the operator $T$ has the so called local smoothing phenomenon, i.e., the $L^q$ integral about $u$ on $[1,2]$ extends the region of $(\frac{1}{p},\frac{1}{q})$ in uniform estimate $T : L_x^p(\mathbb{R}^2)\rightarrow L_{x}^q(\mathbb{R}^2)$.

math.CA↗

YAYI 2: Multilingual Open-Source Large Language Models

As the latest advancements in natural language processing, large language models (LLMs) have achieved human-level language understanding and generation abilities in many real-world tasks, and even have been regarded as a potential path to the artificial general intelligence. To better facilitate research on LLMs, many open-source LLMs, such as Llama 2 and Falcon, have recently been proposed and gained comparable performances to proprietary models. However, these models are primarily designed for English scenarios and exhibit poor performances in Chinese contexts. In this technical report, we propose YAYI 2, including both base and chat models, with 30 billion parameters. YAYI 2 is pre-trained from scratch on a multilingual corpus which contains 2.65 trillion tokens filtered by our pre-training data processing pipeline. The base model is aligned with human values through supervised fine-tuning with millions of instructions and reinforcement learning from human feedback. Extensive experiments on multiple benchmarks, such as MMLU and CMMLU, consistently demonstrate that the proposed YAYI 2 outperforms other similar sized open-source models.

cs.CL↗

Traffic Flow Simulation for Autonomous Driving

A traffic system is a random and complex large system, which is difficult to conduct repeated modelling and control research in a real traffic environment. With the development of automatic driving technology, the requirements for testing and evaluating the development of automatic driving technology are getting higher and higher, so the application of computer technology for traffic simulation has become a very effective technical means. Based on the micro-traffic flow modelling, this paper adopts the vehicle motion model based on cellular automata and the theory of bicycle intelligence to build the simulation environment of autonomous vehicle flow. The architecture of autonomous vehicles is generally divided into a perception system, decision system and control system. The perception system is generally divided into many subsystems, responsible for autonomous vehicle positioning, obstacle recognition, traffic signal detection and recognition and other tasks. Decision systems are typically divided into many subsystems that are responsible for tasks such as path planning, path planning, behavior selection, motion planning, and control. The control system is the basis of the selfdriving car, and each control system of the vehicle needs to be connected with the decision-making system through the bus, and can accurately control the acceleration degree, braking degree, steering amplitude, lighting control and other driving actions according to the bus instructions issued by the decision-making system, so as to achieve the autonomous driving of the vehicle.

cs.RO↗

Undoped Strained Ge Quantum Well with Ultrahigh Mobility Grown by Reduce Pressure Chemical Vapor Deposition

We fabricate an undoped Ge quantum well under 30 nm Ge0.8Si0.2 shallow barrier with reverse grading technology. The under barrier is deposited by Ge0.8Si0.2 followed by Ge0.9Si0.1 so that the variation of Ge content forms a sharp interface which can suppress the threading dislocation density penetrating into undoped Ge quantum well. And the Ge0.8Si0.2 barrier introduces enough in-plane parallel strain -0.41% in the Ge quantum well. The heterostructure field-effect transistors with a shallow buried channel get a high two-dimensional hole gas (2DHG) mobility over 2E6 cm2/Vs at a low percolation density of 2.51 E-11 cm2. We also discover a tunable fractional quantum Hall effect at high densities and high magnetic fields. This approach defines strained germanium as providing the material basis for tuning the spin-orbit coupling strength for fast and coherent quantum computation.

cond-mat.mes-hall↗

Design and testing of LGAD sensor with shallow carbon implantation

The low gain avalanche detectors (LGADs) are thin sensors with fast charge collection which in combination with internal gain deliver an outstanding time resolution of about 30 ps. High collision rates and consequent large particle rates crossing the detectors at the upgraded Large Hadron Collider (LHC) in 2028 will lead to radiation damage and deteriorated performance of the LGADs. The main consequence of radiation damage is loss of gain layer doping (acceptor removal) which requires an increase of bias voltage to compensate for the loss of charge collection efficiency and consequently time resolution. The Institute of High Energy Physics (IHEP), Chinese Academy of Sciences (CAS) has developed a process based on the Institute of Microelectronics (IME), CAS capability to enrich the gain layer with carbon to reduce the acceptor removal effect by radiation. After 1 MeV neutron equivalent fluence of 2.5$\times$10$^{15}$ n$_{eq}$/cm$^{2}$, which is the maximum fluence to which sensors will be exposed at ATLAS High Granularity Timing Detector (HGTD), the IHEP-IME second version (IHEP-IMEv2) 50 $μ$m LGAD sensors already deliver adequate charge collection > 4 fC and time resolution < 50 ps at voltages < 400 V. The operation voltages of these 50 $μ$m devices are well below those at which single event burnout may occur.

physics.ins-det↗

Compressing LSTM Networks by Matrix Product Operators

Long Short Term Memory(LSTM) models are the building blocks of many state-of-the-art natural language processing(NLP) and speech enhancement(SE) algorithms. However, there are a large number of parameters in an LSTM model. This usually consumes a large number of resources to train the LSTM model. Also, LSTM models suffer from computational inefficiency in the inference phase. Existing model compression methods (e.g., model pruning) can only discriminate based on the magnitude of model parameters, ignoring the issue of importance distribution based on the model information. Here we introduce the MPO decomposition, which describes the local correlation of quantum states in quantum many-body physics and is used to represent the large model parameter matrix in a neural network, which can compress the neural network by truncating the unimportant information in the weight matrix. In this paper, we propose a matrix product operator(MPO) based neural network architecture to replace the LSTM model. The effective representation of neural networks by MPO can effectively reduce the computational consumption of training LSTM models on the one hand, and speed up the computation in the inference phase of the model on the other hand. We compare the MPO-LSTM model-based compression model with the traditional LSTM model with pruning methods on sequence classification, sequence prediction, and speech enhancement tasks in our experiments. The experimental results show that our proposed neural network architecture based on the MPO approach significantly outperforms the pruning approach.

cs.NI↗

A Novel Blind Source Separation Framework Towards Maximum Signal-To-Interference Ratio

This letter proposes a new blind source separation (BSS) framework termed minimum variance independent component analysis (MVICA), which can potentially achieve the maximum output signal-to-interference ratio (SIR) while also allowing more flexibility in real implementations. The statistical independence assumption has been the foundation of the most dominant BSS techniques in recent decades. However, this assumption does not always hold true and the accurate probabilistic modeling of source is inherently difficult. To overcome these limitations and improve the separation performance, the MVICA framework is rigorously derived by optimizing the design of these independence-based BSS algorithms with the maximum SIR criterion. A deep neural network-supported implementation of MVICA is subsequently described. Experimental results under various conditions show the superiority of MVICA over the state-of-the-art BSS algorithms, in terms of not only SIR but also signal-to-distortion ratio and automatic speech recognition rate.

cs.SD↗

Low Gain Avalanche Detectors with Good Time Resolution Developed by IHEP and IME for ATLAS HGTD project

This paper shows the simulation and test results of 50um thick Low Gain Avalanche Detectors (LGAD) sensors designed by the Institute of High Energy Physics (IHEP) and fabricated by the Institute of Microelectronics of the Chinese Academy of Sciences (IME). Three wafers have been produced with four different gain layer implant doses each. Different production processes, including variation in the n++ layer implant energy and carbon co-implantation were used. Test results show that the IHEP-IME sensors with the higher dose of gain layer have lower breakdown voltages and higher gain layer voltages from capacitance-voltage properties, which are consistent with the TCAD simulation. Beta test results show that the time resolution of IHEP-IME sensors is better than 35ps when operated at high voltage and the collected charges of IHEP-IME sensors are larger than 15fC before irradiation, which fulfill the required specifications of sensors before irradiations for the ATLAS HGTD project.

physics.ins-det↗

Analytical Shaping Method for Low-Thrust Rendezvous Trajectory Using Cubic Spline Functions

Preliminary mission design requires an efficient and accurate approximation to the low-thrust rendezvous trajectories, which might be generally three-dimensional and involve multiple revolutions. In this paper, a new shaping method using cubic spline functions is developed for the analytical approximation, which shows advantages in the optimality and computational efficiency. The rendezvous constraints on the boundary states and transfer time are all satisfied analytically, under the assumption that the boundary conditions and segment numbers of cubic spline functions are designated in advance. Two specific shapes are then formulated according to whether they have free optimization parameters. The shape without free parameters provides an efficient and robust estimation, while the other one allows a subsequent optimization for the satisfaction of additional constraints such as the constraint on the thrust magnitude. Applications of the proposed method in combination with the particle swarm optimization algorithm are discussed through two typical interplanetary rendezvous missions, that is, an inclined multi-revolution trajectory from the Earth to asteroid Dionysus and a multi-rendezvous trajectory of sample return. Simulation examples show that the proposed method is superior to existing methods in terms of providing good estimation for the global search and generating suitable initial guess for the subsequent trajectory optimization.

cs.RO↗

Effects of shallow carbon and deep N++ layer on the radiation hardness of IHEP-IME LGAD sensors

Low Gain Avalanche Diode (LGAD) is applied for the High-Granularity Timing Detector (HGTD), and it will be used to upgrade the ATLAS experiment. The first batch IHEP-IME LGAD sensors were designed by the Institute of High Energy Physics (IHEP) and fabricated by the Institute of Microelectronics (IME). Three IHEP-IME sensors (W1, W7 and W8) were irradiated by the neutrons up to the fluence of 2.5 x 10^15 n_eq/cm^2 to study the effect of the shallow carbon and deep N++ layer on the irradiation hardness. Taking W7 as a reference, W1 has an extra shallow carbon applied, and W8 has a deeper N++ layer.

physics.ins-det↗

Non-homogeneous initial boundary value problems for the biharmonic Schrödinger equation on an interval

In this paper we consider the initial boundary value problem (IBVP) for the nonlinear biharmonic Schrödinger equation posed on a bounded interval $(0,L)$ with non-homogeneous Navier or Dirichlet boundary conditions, respectively. For Navier boundary IBVP, we set up its local well-posedness if the initial data lies in $H^s(0, L)$ with $s\geq 0$ and $s\neq n+1/2, n\in \mathbb{N}$, and the boundary data are selected from the appropriate spaces with optimal regularities, i.e., the $j$-th order data are chosen in $H_{loc}^{(s+3-j)/4}(\mathbb {R}^+)$, for $j=0,2$. For Dirichlet boundary IBVP the corresponding local well-posedness is obtained when $s>10/7$ and $s\neq n+1/2, n\in \mathbb{N}$, and the boundary data are selected from the appropriate spaces with optimal regularities, i.e., the $j$-th order data are chosen in $H_{loc}^{(s+3-j)/4}(\mathbb {R}^+)$, for $j=0,1$.

math.FA↗

Analyzing Open-Source Serverless Platforms: Characteristics and Performance

Serverless computing is increasingly popular because of its lower cost and easier deployment. Several cloud service providers (CSPs) offer serverless computing on their public clouds, but it may bring the vendor lock-in risk. To avoid this limitation, many open-source serverless platforms come out to allow developers to freely deploy and manage functions on self-hosted clouds. However, building effective functions requires much expertise and thorough comprehension of platform frameworks and features that affect performance. It is a challenge for a service developer to differentiate and select the appropriate serverless platform for different demands and scenarios. Thus, we elaborate the frameworks and event processing models of four popular open-source serverless platforms and identify their salient idiosyncrasies. We analyze the root causes of performance differences between different service exporting and auto-scaling modes on those platforms. Further, we provide several insights for future work, such as auto-scaling and metric collection.

cs.DC↗

High Performance and Scalable NAT System on Commodity Platforms

Quick network address translation (NAT) is proposed to improve the network performance of the NAT system on the commodity server by three ways. First, the quick NAT search algorithm is designed to use the Hash search instead of the sequential search to reduce latency when looking up the NAT rule table. Second, to leverage the power of the multi-core central processing unit (CPU) and the multi-queue network interface card, Quick NAT enables multiple CPU cores to process in parallel. The localized connection tracking table and the compare-and-swap based lock-free NAT Hash tables are designed to eliminate the lock overhead. Third, Quick NAT uses the polling and zero-copy delivery to reduce the cost of interrupt and packet copies. The evaluation results show that Quick NAT obtains high scalability and line-rate throughput on the commodity server.

cs.NI↗

Quick NAT: High performance NAT system on commodity platforms

NAT gateway is an important network system in today's IPv4 network when translating a private IPv4 address to a public address. However, traditional NAT system based on Linux Netfilter cannot achieve high network throughput to meet modern requirements such as data centers. To address this challenge, we improve the network performance of NAT system by three ways. First, we leverage DPDK to enable polling and zero-copy delivery, so as to reduce the cost of interrupt and packet copies. Second, we enable multiple CPU cores to process in parallel and use lock-free hash table to minimize the contention between CPU cores. Third, we use hash search instead of sequential search when looking up the NAT rule table. Evaluation shows that our Quick NAT system significantly improves the performance of NAT on commodity platforms.

cs.NI↗

A Model Compression Method with Matrix Product Operators for Speech Enhancement

The deep neural network (DNN) based speech enhancement approaches have achieved promising performance. However, the number of parameters involved in these methods is usually enormous for the real applications of speech enhancement on the device with the limited resources. This seriously restricts the applications. To deal with this issue, model compression techniques are being widely studied. In this paper, we propose a model compression method based on matrix product operators (MPO) to substantially reduce the number of parameters in DNN models for speech enhancement. In this method, the weight matrices in the linear transformations of neural network model are replaced by the MPO decomposition format before training. In experiment, this process is applied to the causal neural network models, such as the feedforward multilayer perceptron (MLP) and long short-term memory (LSTM) models. Both MLP and LSTM models with/without compression are then utilized to estimate the ideal ratio mask for monaural speech enhancement. The experimental results show that our proposed MPO-based method outperforms the widely-used pruning method for speech enhancement under various compression rates, and further improvement can be achieved with respect to low compression rates. Our proposal provides an effective model compression method for speech enhancement, especially in cloud-free application.

cs.SD↗

Bilinear Hilbert Transforms and (Sub)Bilinear Maximal Functions along Convex Curves

In this paper, we determine the $L^p(\mathbb{R})\times L^q(\mathbb{R})\rightarrow L^r(\mathbb{R})$ boundedness of the bilinear Hilbert transform $H_γ(f,g)$ along a convex curve $γ$ $$H_γ(f,g)(x):=\mathrm{p.\,v.}\int_{-\infty}^{\infty}f(x-t)g(x-γ(t)) \,\frac{\textrm{d}t}{t},$$ where $p$, $q$, and $r$ satisfy $\frac{1}{p}+\frac{1}{q}=\frac{1}{r}$, and $r>\frac{1}{2}$, $p>1$, and $q>1$. Moreover, the same $L^p(\mathbb{R})\times L^q(\mathbb{R})\rightarrow L^r(\mathbb{R})$ boundedness property holds for the corresponding (sub)bilinear maximal function $M_γ(f,g)$ along a convex curve $γ$ $$M_γ(f,g)(x):=\sup_{\varepsilon>0}\frac{1}{2\varepsilon}\int_{-\varepsilon}^{\varepsilon}|f(x-t)g(x-γ(t))| \,\textrm{d}t.$$

math.CA↗

A Carleson problem for the Boussinesq operator

In this paper, Theorems 1.1- 1.2 show that the Boussinesq operator $\mathcal{B}_tf$ converges pointwise to its initial data $f\in H^s(\mathbb{R})$ as $t\to 0$ provided $s\geq\frac{1}{4}$ -- more precisely -- on the one hand, by constructing a counterexample in $\mathbb{R}$ we discover that the optimal convergence index $s_{c,1}=\frac14$; on the other hand, we find that the Hausdorff dimension of the disconvergence set for $\mathcal{B}_tf$ is \begin{align*} α_{1,\mathcal{B}}(s)&=\begin{cases} 1-2s&\ \ \text{as}\ \ \frac{1}{4}\leq s\leq\frac{1}{2};\\ 1 &\ \ \text{as}\ \ 0<s<\frac{1}{4}. \end{cases} \end{align*} Moreover, Theorem 1.3 presents a higher dimensional lift of Theorems 1.1- 1.2 under $f$ being radial.

math.CA↗