arXiv subjects

Haijun Zhang

Publications and source records attributed to Haijun Zhang.

At least 91 records · Page 5Linked to original sources

arXiv2024

Symmetry-driven anisotropic coupling effect in antiferromagnetic topological insulator: Mechanism for high-Chern-number quantum anomalous Hall state

Antiferromagnetic (AFM) topological insulators (TIs), which host magnetically gapped Dirac-cone surface states and exhibit many exotic physical phenomena, have attracted great attention. Here, we find that the coupled surface states can be intertwined to give birth to a set of $2n$ unique new Dirac cones, dubbed intertwined Dirac cones, through the anisotropic coupling enforced by crystalline $n$-fold ($n=2, 3, 4, 6$) rotation symmetry $C_{nz}$ in the presence of a $PT$-symmetry breaking potential, for example, an electric field. Interestingly, we also find that the warping effect further drives the intertwined Dirac-cone state into a quantum anomalous Hall phase with a high Chern number ($C=n$). Then, based on first-principles calculations, we have explicitly demonstrated six intertwined Dirac cones and a Chern insulating phase with a high Chern number ($C=3$) in MnBi$_2$Te$_4$$/$(Bi$_2$Te$_3$)$_{\mathrm{m}}/$MnBi$_2$Te$_4$ heterostructures, as well as the $C=2$ and $C=4$ phases in HgS and $α$-Ag$_2$Te films, respectively. This work discovers the intertwined Dirac-cone state in AFM TI thin films, which reveals a mechanism for designing the quantum anomalous Hall state with a high Chern number and also paves a way for studying highly tunable high-Chen-number flat bands of twistronics.

cond-mat.mes-hall↗

arXiv2024

IDVT: Interest-aware Denoising and View-guided Tuning for Social Recommendation

In the information age, recommendation systems are vital for efficiently filtering information and identifying user preferences. Online social platforms have enriched these systems by providing valuable auxiliary information. Socially connected users are assumed to share similar preferences, enhancing recommendation accuracy and addressing cold start issues. However, empirical findings challenge the assumption, revealing that certain social connections can actually harm system performance. Our statistical analysis indicates a significant amount of noise in the social network, where many socially connected users do not share common interests. To address this issue, we propose an innovative \underline{I}nterest-aware \underline{D}enoising and \underline{V}iew-guided \underline{T}uning (IDVT) method for the social recommendation. The first ID part effectively denoises social connections. Specifically, the denoising process considers both social network structure and user interaction interests in a global view. Moreover, in this global view, we also integrate denoised social information (social domain) into the propagation of the user-item interactions (collaborative domain) and aggregate user representations from two domains using a gating mechanism. To tackle potential user interest loss and enhance model robustness within the global view, our second VT part introduces two additional views (local view and dropout-enhanced view) for fine-tuning user representations in the global view through contrastive learning. Extensive evaluations on real-world datasets with varying noise ratios demonstrate the superiority of IDVT over state-of-the-art social recommendation methods.

cs.AI↗

arXiv2024

Joint Radar Sensing, Location, and Communication Resources Optimization in 6G Network

The possibility of jointly optimizing location sensing and communication resources, facilitated by the existence of communication and sensing spectrum sharing, is what promotes the system performance to a higher level. However, the rapid mobility of user equipment (UE) can result in inaccurate location estimation, which can severely degrade system performance. Therefore, the precise UE location sensing and resource allocation issues are investigated in a spectrum sharing sixth generation network. An approach is proposed for joint subcarrier and power optimization based on UE location sensing, aiming to minimize system energy consumption. The joint allocation process is separated into two key phases of operation. In the radar location sensing phase, the multipath interference and Doppler effects are considered simultaneously, and the issues of UE's location and channel state estimation are transformed into a convex optimization problem, which is then solved through gradient descent. In the communication phase, a subcarrier allocation method based on subcarrier weights is proposed. To further minimize system energy consumption, a joint subcarrier and power allocation method is introduced, resolved via the Lagrange multiplier method for the non-convex resource allocation problem. Simulation analysis results indicate that the location sensing algorithm exhibits a prominent improvement in accuracy compared to benchmark algorithms. Simultaneously, the proposed resource allocation scheme also demonstrates a substantial enhancement in performance relative to baseline schemes.

eess.SY↗

arXiv2024

Beyond the Edge: An Advanced Exploration of Reinforcement Learning for Mobile Edge Computing, its Applications, and Future Research Trajectories

Mobile Edge Computing (MEC) broadens the scope of computation and storage beyond the central network, incorporating edge nodes close to end devices. This expansion facilitates the implementation of large-scale "connected things" within edge networks. The advent of applications necessitating real-time, high-quality service presents several challenges, such as low latency, high data rate, reliability, efficiency, and security, all of which demand resolution. The incorporation of reinforcement learning (RL) methodologies within MEC networks promotes a deeper understanding of mobile user behaviors and network dynamics, thereby optimizing resource use in computing and communication processes. This paper offers an exhaustive survey of RL applications in MEC networks, initially presenting an overview of RL from its fundamental principles to the latest advanced frameworks. Furthermore, it outlines various RL strategies employed in offloading, caching, and communication within MEC networks. Finally, it explores open issues linked with software and hardware platforms, representation, RL robustness, safe RL, large-scale scheduling, generalization, security, and privacy. The paper proposes specific RL techniques to mitigate these issues and provides insights into their practical applications.

cs.NI↗

arXiv2024

Light-induced giant enhancement of nonreciprocal transport at KTaO3-based interfaces

Nonlinear transport is a unique functionality of noncentrosymmetric systems, which reflects profound physics, such as spin-orbit interaction, superconductivity and band geometry. However, it remains highly challenging to enhance the nonreciprocal transport for promising rectification devices. Here, we observe a light-induced giant enhancement of nonreciprocal transport at the superconducting and epitaxial CaZrO3/KTaO3 (111) interfaces. The nonreciprocal transport coefficient undergoes a giant increase with three orders of magnitude up to 105 A-1T-1. Furthermore, a strong Rashba spin-orbit coupling effective field of 14.7 T is achieved with abundant high-mobility photocarriers under ultraviolet illumination, which accounts for the giant enhancement of nonreciprocal transport coefficient. Our first-principles calculations further disclose the stronger Rashba spin-orbit coupling strength and the longer relaxation time in the photocarrier excitation process, bridging the light-property quantitative relationship. Our work provides an alternative pathway to boost nonreciprocal transport in noncentrosymmetric systems and facilitates the promising applications in opto-rectification devices and spin-orbitronic devices.

cond-mat.mes-hall↗

arXiv2024

Spin: An Efficient Secure Computation Framework with GPU Acceleration

Accuracy and efficiency remain challenges for multi-party computation (MPC) frameworks. Spin is a GPU-accelerated MPC framework that supports multiple computation parties and a dishonest majority adversarial setup. We propose optimized protocols for non-linear functions that are critical for machine learning, as well as several novel optimizations specific to attention that is the fundamental unit of Transformer models, allowing Spin to perform non-trivial CNNs training and Transformer inference without sacrificing security. At the backend level, Spin leverages GPU, CPU, and RDMA-enabled smart network cards for acceleration. Comprehensive evaluations demonstrate that Spin can be up to $2\times$ faster than the state-of-the-art for deep neural network training. For inference on a Transformer model with 18.9 million parameters, our attention-specific optimizations enable Spin to achieve better efficiency, less communication, and better accuracy.

cs.CR↗

arXiv2023

Cut-and-Paste: Subject-Driven Video Editing with Attention Control

This paper presents a novel framework termed Cut-and-Paste for real-word semantic video editing under the guidance of text prompt and additional reference image. While the text-driven video editing has demonstrated remarkable ability to generate highly diverse videos following given text prompts, the fine-grained semantic edits are hard to control by plain textual prompt only in terms of object details and edited region, and cumbersome long text descriptions are usually needed for the task. We therefore investigate subject-driven video editing for more precise control of both edited regions and background preservation, and fine-grained semantic generation. We achieve this goal by introducing an reference image as supplementary input to the text-driven video editing, which avoids racking your brain to come up with a cumbersome text prompt describing the detailed appearance of the object. To limit the editing area, we refer to a method of cross attention control in image editing and successfully extend it to video editing by fusing the attention map of adjacent frames, which strikes a balance between maintaining video background and spatio-temporal consistency. Compared with current methods, the whole process of our method is like ``cut" the source object to be edited and then ``paste" the target object provided by reference image. We demonstrate that our method performs favorably over prior arts for video editing under the guidance of text prompt and extra reference image, as measured by both quantitative and subjective evaluations.

cs.CV↗

arXiv2023

Flat bands and magnetism in $\mathrm{\mathbf{Fe_4 Ge Te_2}}$ and $\mathrm{\mathbf{Fe_5GeTe_2}}$ due to bipartite crystal lattices

$\mathrm{Fe_{n=4,5}GeTe_2}$ exhibits quasi-two-dimensional properties as a promising candidate for a near-room-temperature ferromagnet, which has attracted great interest. In this work, we notice that the crystal lattice of $\mathrm{Fe_{n=4,5}GeTe_2}$ can be approximately regarded as being stacked by three bipartite crystal lattices. By combining the model Hamiltonians of bipartite crystal lattices and first-principles calculations, we investigate the electronic structure and the magnetism of $\mathrm{Fe_{n=4,5}GeTe_2}$. We conclude that flat bands near the Fermi level originate from the bipartite crystal lattices and that these flat bands are expected to lead to the itinerant ferromagnetism in $\mathrm{Fe_{n=4,5}GeTe_2}$. Interestingly, we also find that the magnetic moment of the Fe5 atom in $\mathrm{Fe_5 Ge Te_2}$ is distinct from the other Fe atoms and is sensitive to the Coulomb interaction $U$ and external pressure. These findings may be helpful to understand the exotic magnetic behavior of $\mathrm{Fe_{n=4,5} Ge Te_2}$.

cond-mat.mtrl-sci↗

arXiv2023

Unconventional superconductivity in Sc$_2$Ir$_{4-x}$Si$_x$ by spin-orbit coupling driven flat band

The kagome lattice is very attractive as it can host many novel quantum states, such as the charge density wave, superconductivity, quantum spin liquid, etc. Meanwhile, iridates often exhibit a strong spin-orbit coupling (SOC) effect due to the large atomic mass of 5$d$ elements, which has important implications for both the energy bands and the pairing symmetry of superconductors. For the Laves phase superconductor Sc$_2$Ir$_4$ with a kagome lattice, by doping Si to the Ir sites, we observed a nonmonotonic and two-dome like doping dependence of the superconducting transition temperature $T_{\rm c}$, which is typically found in many unconventional superconducting systems. Interestingly, for some samples, especially Sc$_2$Ir$_{3.5}$Si$_{0.5}$ with the optimal $T_{\rm c}$, after the suppression of superconductivity, the normal-state resistivity exhibits a semiconducting behavior; meanwhile, the specific heat coefficient shows an upturn which follows the relation $C/T\propto{\rm ln}(T_0/T)$ at low temperatures. Around the optimal doping, the resistance measurements exhibit strong superconducting fluctuations. And the superconductivity related specific heat can be fitted by the model of a $d$-wave gap after subtracting the normal-state background. These strongly suggest unconventional superconductivity and correlation effect in the samples, which is mainly induced by a flat band near the Fermi level when considering the SOC, as supported by the first-principles calculations. Our results reveal a new unconventional superconducting system Sc$_2$Ir$_{4-x}$Si$_x$ with strong correlation effects induced by the flat band in the kagome system with strong SOC.

cond-mat.supr-con↗

arXiv2023

DRGame: Diversified Recommendation for Multi-category Video Games with Balanced Implicit Preferences

The growing popularity of subscription services in video game consumption has emphasized the importance of offering diversified recommendations. Providing users with a diverse range of games is essential for ensuring continued engagement and fostering long-term subscriptions. However, existing recommendation models face challenges in effectively handling highly imbalanced implicit feedback in gaming interactions. Additionally, they struggle to take into account the distinctive characteristics of multiple categories and the latent user interests associated with these categories. In response to these challenges, we propose a novel framework, named DRGame, to obtain diversified recommendation. It is centered on multi-category video games, consisting of two {components}: Balance-driven Implicit Preferences Learning for data pre-processing and Clustering-based Diversified Recommendation {Module} for final prediction. The first module aims to achieve a balanced representation of implicit feedback in game time, thereby discovering a comprehensive view of player interests across different categories. The second module adopts category-aware representation learning to cluster and select players and games based on balanced implicit preferences, and then employs asymmetric neighbor aggregation to achieve diversified recommendations. Experimental results on a real-world dataset demonstrate the superiority of our proposed method over existing approaches in terms of game diversity recommendations.

cs.IR↗

arXiv2023

Three-Dirac-fermion approach to unexpected universal gapless surface states of van der Waals magnetic topological insulators

Layered van der Waals (vdW) topological materials, especially the recently discovered MnBi$_2$Te$_4$-family magnetic topological insulators (TIs), have aroused great attention. However, there has been a serious debate about whether the surface states are gapped or gapless for antiferromagnetic (AFM) TI MnBi$_2$Te$_4$, which is crucial to the prospect of various magnetic topological phenomena. Here, a minimal three-Dirac-fermion approach is developed to generally describe topological surface states of nonmagnetic/magnetic vdW TIs under the modulation of the interlayer vdW gap. In particular, this approach is applied to address the controversial issues concerning the surface states of vdW AFM TIs. Remarkably, topologically protected gapless Dirac-cone surface states are found to arise due to a small expansion of the interlayer vdW gap on the surface, when the Chern number equals zero for the surface ferromagnetic layer; while the surface states remain gapped in all other cases. These results are further confirmed by our first-principles calculations on AFM TI MnBi$_2$Te$_4$. The theorectically discovered gapless Dirac-cone states provide a unique mechanism for understanding the puzzle of the experimentally observed gapless surface states in MnBi$_2$Te$_4$. This work also provides a promising way for experiments to realize the intrinsic magnetic quantum anomalous Hall effect in MnBi$_2$Te$_4$ films with a large energy gap.

cond-mat.mes-hall↗

arXiv2023

New Delay Doppler Communication Paradigm in 6G era: A Survey of Orthogonal Time Frequency Space (OTFS)

In the 6G era, space-air-Ground integrated networks (SAGIN) are anticipated to deliver global coverage, necessitating support for a diverse array of emerging applications in high-mobility, hostile environments. Under such conditions, conventional orthogonal frequency division multiplexing (OFDM) modulation, widely employed in cellular and Wi-Fi communication systems, experiences performance degradation due to significant Doppler shifts. To overcome this obstacle, a novel two-dimensional (2D) modulation approach, namely orthogonal time frequency space (OTFS), has emerged as a key enabler for future high-mobility use cases. Distinctively, OTFS modulates information within the delay-Doppler (DD) domain, as opposed to the time-frequency (TF) domain utilized by OFDM. This offers advantages such as Doppler and delay resilience, reduced signaling latency, a lower peak-to-average ratio (PAPR), and a reduced-complexity implementation. Recent studies further indicate that the direct interplay between information and the physical world in the DD domain positions OTFS as a promising waveform for achieving integrated sensing and communications (ISAC). In this article, we present an in-depth review of OTFS technology in the context of the 6G era, encompassing fundamentals, recent advancements, and future directions. Our objective is to provide a valuable resource for researchers engaged in the field of OTFS.

cs.IT↗

arXiv2023

DeepMA: End-to-end Deep Multiple Access for Wireless Image Transmission in Semantic Communication

Semantic communication is a new paradigm that exploits deep learning models to enable end-to-end communications processes, and recent studies have shown that it can achieve better noise resiliency compared with traditional communication schemes in a low signal-to-noise (SNR) regime. To achieve multiple access in semantic communication, we propose a deep learning-based multiple access (DeepMA) method by training semantic communication models with the abilities of joint source-channel coding (JSCC) and orthogonal signal modulation. DeepMA is achieved by a DeepMA network (DMANet), which is comprised of several independent encoder-decoder pairs (EDPs), and the DeepMA encoders can encode the input data as mutually orthogonal semantic symbol vectors (SSVs) such that the DeepMA decoders can detect and recover their own target data from a received mixed SSV (MSSV) superposed by multiple SSV components transmitted from different encoders. We describe frameworks of DeepMA in wireless device-to-device (D2D), downlink, and uplink channel multiplexing scenarios, along with the training algorithm. We evaluate the performance of the proposed DeepMA in wireless image transmission tasks and compare its performance with the attention module-based deep JSCC (ADJSCC) method and conventional communication schemes using better portable graphics (BPG) and Low-density parity-check code (LDPC). The results obtained show that the proposed DeepMA can achieve effective, flexible, and privacy-preserving channel multiplexing process, and demonstrate that our proposed DeepMA approach can yield comparable bandwidth efficiency compared with conventional multiple access schemes.

eess.SP↗

arXiv2023

Accelerating Wireless Federated Learning via Nesterov's Momentum and Distributed Principle Component Analysis

A wireless federated learning system is investigated by allowing a server and workers to exchange uncoded information via orthogonal wireless channels. Since the workers frequently upload local gradients to the server via bandwidth-limited channels, the uplink transmission from the workers to the server becomes a communication bottleneck. Therefore, a one-shot distributed principle component analysis (PCA) is leveraged to reduce the dimension of uploaded gradients such that the communication bottleneck is relieved. A PCA-based wireless federated learning (PCA-WFL) algorithm and its accelerated version (i.e., PCA-AWFL) are proposed based on the low-dimensional gradients and the Nesterov's momentum. For the non-convex loss functions, a finite-time analysis is performed to quantify the impacts of system hyper-parameters on the convergence of the PCA-WFL and PCA-AWFL algorithms. The PCA-AWFL algorithm is theoretically certified to converge faster than the PCA-WFL algorithm. Besides, the convergence rates of PCA-WFL and PCA-AWFL algorithms quantitatively reveal the linear speedup with respect to the number of workers over the vanilla gradient descent algorithm. Numerical results are used to demonstrate the improved convergence rates of the proposed PCA-WFL and PCA-AWFL algorithms over the benchmarks.

cs.LG↗

arXiv2023

Observation of spin-polarized surface states in a nodal line semimetal SnTaS$_{2}$

The superconductor SnTaS$_{2}$ is theoretically predicted to be an intriguing topological nodal line semimetal without consideration of spin-orbit coupling. By carrying out angle-resolved photoemission (ARPES) and spin-resolved ARPES measurements combined with band structure calculations, we have provided a complete picture of the electronic structure and spin polarization property for the prominent surface states of SnTaS$_{2}$. The low-energy electronic states are dominated by surface states; two of them are from the S-terminated surface, while four of them are from the Sn-terminated surface. These give rise to interesting Fermi surface topology of SnTaS$_{2}$: three pockets located at $\barΓ$, $\bar{M}$ and $\bar{K}$ for the S-terminated surface and two pockets surrounding $\barΓ$ and $\bar{K}$ for the Sn-terminated surface. We further reveal that two surface states that cross the Fermi level are spin-polarized. Since SnTaS$_{2}$ is also a superconductor, our observations indicate that it may provide a new platform to explore topological superconductivity and other exotic properties.

cond-mat.supr-con↗

arXiv2023

Topology hierarchy of transition metal dichalcogenides built from quantum spin Hall layers

The evolution of the physical properties of two-dimensional material from monolayer limit to the bulk reveals unique consequences from dimension confinement and provides a distinct tuning knob for applications. Monolayer 1T'-phase transition metal dichalcogenides (1T'-TMDs) with ubiquitous quantum spin Hall (QSH) states are ideal two-dimensional building blocks of various three-dimensional topological phases. However, the stacking geometry was previously limited to the bulk 1T'-WTe2 type. Here, we introduce the novel 2M-TMDs consisting of translationally stacked 1T'-monolayers as promising material platforms with tunable inverted bandgaps and interlayer coupling. By performing advanced polarization-dependent angle-resolved photoemission spectroscopy as well as first-principles calculations on the electronic structure of 2M-TMDs, we revealed a topology hierarchy: 2M-WSe2, MoS2, and MoSe2 are weak topological insulators (WTIs), whereas 2M-WS2 is a strong topological insulator (STI). Further demonstration of topological phase transitions by tunning interlayer distance indicates that band inversion amplitude and interlayer coupling jointly determine different topological states in 2M-TMDs. We propose that 2M-TMDs are parent compounds of various exotic phases including topological superconductors and promise great application potentials in quantum electronics due to their flexibility in patterning with two-dimensional materials.

cond-mat.mtrl-sci↗

arXiv2023

A Dirac-fermion approach and its application to design high Chern numbers in magnetic topological insulator multilayers

Quantum anomalous Hall (QAH) insulators host topologically protected dissipationless chiral edge states, the number of which is determined by its Chern number. Up to now, the QAH state has been realized in a few magnetic topological insulators, but usually with a low Chern number. Here, we develop a Dirac-fermion approach which is valuable to understand and design high Chern numbers in various multilayers of layered magnetic topological insulators. Based on the Dirac-fermion approach, we demonstrate how to understand and tune high Chern numbers in ferromagentic MnBi$_{2}$Te$_{4}$ films through the van der Waals (vdW) gap modulation. Further, we also employ the Dirac-fermion approach to understand the experimentally observed high Chern numbers and topological phase transition from the Chern number $C=2$ to $C=1$ in the [3QL-(Bi,Sb)$_{1.76}$Cr$_{0.24}$Te$_{3}$]/[4QL-(Bi,Sb)$_{2}$Te$_{3}$] multilayers. Our work provides a powerful tool to design the QAH states with a high Chern number in layered magnetic topological insulator multilayers.

cond-mat.mes-hall↗

arXiv2022

Towards Feature Distribution Alignment and Diversity Enhancement for Data-Free Quantization

To obtain lower inference latency and less memory footprint of deep neural networks, model quantization has been widely employed in deep model deployment, by converting the floating points to low-precision integers. However, previous methods (such as quantization aware training and post training quantization) require original data for the fine-tuning or calibration of quantized model, which makes them inapplicable to the cases that original data are not accessed due to privacy or security. This gives birth to the data-free quantization method with synthetic data generation. While current data-free quantization methods still suffer from severe performance degradation when quantizing a model into lower bit, caused by the low inter-class separability of semantic features. To this end, we propose a new and effective data-free quantization method termed ClusterQ, which utilizes the feature distribution alignment for synthetic data generation. To obtain high inter-class separability of semantic features, we cluster and align the feature distribution statistics to imitate the distribution of real data, so that the performance degradation is alleviated. Moreover, we incorporate the diversity enhancement to solve class-wise mode collapse. We also employ the exponential moving average to update the centroid of each cluster for further feature distribution improvement. Extensive experiments based on different deep models (e.g., ResNet-18 and MobileNet-V2) over the ImageNet dataset demonstrate that our proposed ClusterQ model obtains state-of-the-art performance.

cs.CV↗