Xueqing Peng, Lingfei Qian, Yan Wang, Ruoyu Xiang, Yueru He, Yang Ren, Mingyang Jiang, Vincent Jim Zhang, Yuqing Guo, Jeff Zhao, Huan He, Yi Han, Yun Feng, Yuechen Jiang, Yupeng Cao, Haohang Li, Yangyang Yu, Xiaoyu Wang, Penglei Gao, Shengyuan Lin, Keyi Wang, Shanshan Yang, Yilun Zhao, Zhiwei Liu, Peng Lu, Jerry Huang, Suyuchen Wang, Triantafillos Papadopoulos, Polydoros Giannouris, Efstathia Soufleri, Nuo Chen, Zhiyang Deng, Heming Fu, Yijia Zhao, Mingquan Lin, Meikang Qiu, Kaleb E Smith, Arman Cohan, Xiao-Yang Liu, Jimin Huang, Guojun Xiong, Alejandro Lopez-Lira, Xi Chen, Junichi Tsujii, Jian-Yun Nie, Sophia Ananiadou, Qianqian Xie. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Going beyond simple text processing, financial auditing requires detecting semantic, structural, and numerical inconsistencies across large-scale disclosures. As financial reports are filed in XBRL, a structured XML format governed by accounting standards, auditing becomes a structured information extraction and reasoning problem involving concept alignment, taxonomy-defined relations, and cross-document consistency. Although large language models (LLMs) show promise on isolated financial tasks, their capability in professional-grade auditing remains unclear. We introduce FinAuditing, a taxonomy-aligned, structure-aware benchmark built from real XBRL filings. It contains 1,102 annotated instances averaging over 33k tokens and defines three tasks: Financial Semantic Matching (FinSM), Financial Relationship Extraction (FinRE), and Financial Mathematical Reasoning (FinMR). Evaluations of 13 state-of-the-art LLMs reveal substantial gaps in concept retrieval, taxonomy-aware relation modeling, and consistent cross-document reasoning. These findings highlight the need for realistic, structure-aware benchmarks. We release the evaluation code1 and dataset2 publicly, and the task currently serves as the official benchmark of an ongoing public evaluation contest3.

Network traffic forecasting is crucial for dynamic resource allocation and network management. However, real-time network traffic forecasting in real-world scenarios is challenging due to the limitation of incomplete data. In this paper, we propose a masked generative model (MGM) for real-time network traffic forecasting from incomplete data. Firstly, we formulate the forecasting task as a low-tubal-rank tensor completion problem and verify that generative models can produce low-tubal-rank tensors through low-dimensional latent variables. Secondly, we propose MGM, which adapts masked autoencoders to robustly learn latent variables from incomplete traffic data. The variables are then mapped to complete low-tubal-rank traffic tensors through pretrained generative models for real-time forecasting. We also establish a performance guarantee that quantifies the error bound of the proposed approach. Finally, experiments on real-world datasets demonstrate that our approach achieves accurate network traffic forecasting within 100 ms, with a normalized root mean squared error (NRMSE) below 0.1.

Network latency estimation is critical for network performance monitoring and management. However, with the escalating demand for real-time performance monitoring and rapid network adjustments in contemporary networks, existing latency estimation methodologies fall short of meeting the need for instantaneous estimation. In this article, we propose a pretrained generative model-based scheme (PGM) for real-time network latency estimation. PGM operates in two stages. First, we employ a pretrained generative model to relax the low-rank constraint typically associated with latency matrix completion (MC). The pretrained generative model well learns the low-rank characteristics of latency matrices in the pretraining stage and can map a condensed latent representation to the matrix space. Second, instead of directly optimizing the matrix, we turn to optimizing the latent representation. Leveraging the low-rank structure achieved by the pretrained generative model simplifies our optimization process, enabling real-time estimation. We also provide a theoretical recovery guarantee to reveal the error bound of PGM. Experimental results on real-world datasets show that the proposed scheme can achieve accurate latency estimation within 50 ms while maintaining the relative squared error (RSE) of estimation at no more than 0.11 (as evidenced using the PlanetLab dataset).

In this article, we propose a novel spectral tensor layer for communication-free distributed deep learning. The overall framework is as follows: first, we represent the data in tensor form (instead of vector form) and replace the matrix product in conventional neural networks with the tensor product, which in effect imposes certain transformed-induced structure on the original weight matrices, e.g., a block-circulant structure; then, we apply a linear transform along a certain dimension to split the original dataset into multiple spectral subdatasets; as a result, the proposed spectral tensor network consists of parallel branches where each branch is a conventional neural network trained on a spectral subdataset with ZERO communication cost. The parallel branches are directly ensembled (i.e., the weighted sum of their outputs) to generate an overall network with substantially stronger generalization capability than that of each branch. Moreover, the proposed method enjoys a byproduct of decentralization gain in terms of memory and computation, compared with traditional networks. It is a natural yet elegant solution for heterogeneous data in federated learning (FL), where data at different nodes have different resolutions. Finally, we evaluate the proposed spectral tensor networks on the MNIST, CIFAR-10, ImageNet-1K, and ImageNet-21K datasets, respectively, to verify that they simultaneously achieve communication-free distributed learning, distributed storage reduction, parallel computation speedup, and learning with multiresolution data.

Large language models (LLMs) are computationally intensive. The computation workload and the memory footprint grow quadratically with the dimension (layer width). Most of LLMs' parameters come from the linear layers of the transformer structure and are highly redundant. These linear layers contribute more than 80% of the computation workload and 99% of the model size. To pretrain and finetune LLMs efficiently, there are three major challenges to address: 1) reducing redundancy of the linear layers; 2) reducing GPU memory footprint; 3) improving GPU utilization when using distributed training. Prior methods, such as LoRA and QLoRA, utilized low-rank structure and quantization to reduce the number of trainable parameters and model size, respectively. However, the resulting model still consumes a large amount of GPU memory. In this paper, we present high-performance GPU-based methods for both pretraining and finetuning quantized LLMs with low-rank structures. We replace a single linear layer in the transformer structure with two narrower linear layers, significantly reducing the number of parameters by several orders of magnitude. By quantizing the pretrained parameters into low precision (8-bit and 4-bit), the memory consumption of the resulting model is further reduced. Compared with existing LLMs, our methods achieve a speedup of 1.3x and a model compression ratio of 2.64 x for pretraining without accuracy drop. For finetuning, our methods achieve an average accuracy score increase of 6.3 and 24.0 in general tasks and financial tasks, respectively, and GPU memory consumption is reduced by 6.3x. The sizes of our models are smaller than 0.59 GB, allowing inference on a smartphone.

Despite Spanish's pivotal role in the global finance industry, a pronounced gap exists in Spanish financial natural language processing (NLP) and application studies compared to English, especially in the era of large language models (LLMs).To bridge this gap, we unveil Toisón de Oro, the first bilingual framework that establishes instruction datasets, finetuned LLMs, and evaluation benchmark for financial LLMs in Spanish joint with English.We construct a rigorously curated bilingual instruction dataset including over 144K Spanish and English samples from 15 datasets covering 7 tasks.Harnessing this, we introduce FinMA-ES, an LLM designed for bilingual financial applications.We evaluate our model and existing LLMs using FLARE-ES, the first comprehensive bilingual evaluation benchmark with 21 datasets covering 9 tasks.The FLARE-ES benchmark results

LLMs have transformed NLP and shown promise in various fields, yet their potential in finance is underexplored due to a lack of comprehensive benchmarks, the rapid development of LLMs, and the complexity of financial tasks. In this paper, we introduce FinBen, the first extensive open-source evaluation benchmark, including 42 datasets spanning 24 financial tasks, covering eight critical aspects: information extraction (IE), textual analysis, question answering (QA), text generation, risk management, forecasting, decision-making, and bilingual (English and Spanish). FinBen offers several key innovations: a broader range of tasks and datasets, the first evaluation of stock trading, novel agent and Retrieval-Augmented Generation (RAG) evaluation, and two novel datasets for regulations and stock trading. Our evaluation of 21 representative LLMs, including GPT-4, ChatGPT, and the latest Gemini, reveals several key findings: While LLMs excel in IE and textual analysis, they struggle with advanced reasoning and complex tasks like text generation and forecasting. GPT-4 excels in IE and stock trading, while Gemini is better at text generation and forecasting. Instruction-tuned LLMs improve textual analysis but offer limited benefits for complex tasks such as QA. FinBen has been used to host the first financial LLMs shared task at the FinNLP-AgentScen workshop during IJCAI-2024, attracting 12 teams. Their novel solutions outperformed GPT-4, showcasing FinBen's potential to drive innovations in financial LLMs. All datasets and code are publicly available for the research community, with results shared and updated regularly on the Open Financial LLM Leaderboard.

Snapshot compressive imaging (SCI) cameras compress high-speed videos or hyperspectral images into measurement frames. However, decoding the data frames from measurement frames is compute-intensive. Existing state-of-the-art decoding algorithms suffer from low decoding quality or heavy running time or both, which are not practical for real-time applications. In this article, we exploit the powerful learning ability of deep neural networks (DNN) and propose a novel tensor fast iterative shrinkage-thresholding algorithm net (Tensor FISTA-Net) as a real-time decoder for SCI cameras. Since SCI cameras have an accurate physical model, we can trade training time for the decoding time by generating abundant synthetic data and training a decoder on the cloud. Tensor FISTA-Net not only learns a sparse representation of the frames through convolution layers but also reduces the decoding time and memory consumption significantly through tensor operations, which makes Tensor FISTA-Net an appropriate approach for a real-time decoder. Our proposed Tensor FISTA-Net obtains an average PSNR improvement of 0.79-2.84 dB (video images) and 2.61-4.43 dB (hyperspectral images) over the state-of-the-art algorithms, along with more clear and detailed visual results on real SCI datasets, Hammer and Wheel, respectively. Our Tensor FISTA-Net reaches 45 frames per second in video datasets and 70 frames per second in hyperspectral datasets, meeting the real-time requirement. Besides, the trained model occupies only a 12 -MB memory footprint, making it applicable to real-time Internet of Things (IoT) applications.

Google's quantum supremacy announcement has received broad questions from academia and industry due to the debatable estimate of 10,000 years' running time for the classical simulation task on the Summit supercomputer. Has quantum supremacy already come? Or will it come in one or two decades later? To avoid hasty advertisements of quantum supremacy by tech giants or quantum startups and eliminate the cost of dedicating a team to the classical simulation task, we advocate an open-source approach to maintain a trustable benchmark performance. In this paper, we take a reinforcement learning approach for the classical simulation of quantum circuits and demonstrate its great potential by reporting an estimated simulation time of less than 4 days, a speedup of 5.40x over the state-of-the-art method. Specifically, we formulate the classical simulation task as a tensor network contraction ordering problem using the K-spin Ising model and employ a novel Hamiltonina-based reinforcement learning algorithm. Then, we establish standard criteria to evaluate the performance of classical simulation of quantum circuits. We develop a dozen of massively parallel environments to simulate quantum circuits. We open-source our parallel gym environments and benchmarks. We hope the AI/ML community and quantum physics community will collaborate to maintain reference curves for validating an unequivocal first demonstration of empirical quantum supremacy.

Extracting information from large-scale high-dimensional data is a fundamentally important task in high performance computing, where the hierarchical Tucker (HT) tensor learning approach (learning a tensor-tree structure) has been widely used in many applications. However, HT tensor learning algorithms are compute-intensive due to the “curse of dimensionality,” i.e., the time complexity grows exponentially with the order of the data tensor. The computation of HT tensor learning algorithms boils down to tensor primitives, which are amenable to computing on GPU tensor cores. Existing work does not support HT tensor learning using GPU tensor cores. There are three main challenges to address: 1) to accelerate tensor learning primitives using GPU tensor cores; 2) to implement the tensor learning algorithms using GPU tensor cores and multiple GPUs; 3) to support large-scale data tensors exceeding the GPU memory capacity. In this paper, we present efficient HT tensor learning primitives using GPU tensor cores and demonstrate three applications. First, we utilize GPU tensor cores to optimize HT tensor learning primitives, including tensor contractions, tensor matricizations and tensor singular value decomposition (SVD). We employ the optimized primitives to optimize HT tensor decomposition algorithms for Big Data analysis. Second, we propose a novel HT tensor layer for deep neural networks, whose training process only involves a forward pass without back propagation. The forward pass consists of tensor operations, thus further exploiting the computing power of GPU tensor cores. Third, we apply the optimized primitives to develop a tensor-tree structured quantum machine learning algorithmtree-tensor network (TTN). Compared with TensorLy and TensorNetwork on NVIDIA A100 GPUs, our third-order HT tensor decomposition algorithm achieves up to$8.92 \times$and$6.42 \times$speedups, respectively, and our high-order case achieves up to$32.67 \times$and$23.97 \times$speedups, respectively. Our HT tensor layer for a fully connected neural network achieves$49.2 \times$compression at the cost of 0.5% drops in accuracy and$1.42 \times$speedup compared with the implementation on CUDA cores; for the AlexNet, our HT tensor layer achieves$9.45 \times$compression at the cost of 0.8% drops in accuracy and$1.87 \times$speedup compared with the implementation on CUDA cores. Our TTN algorithm achieves up to$11.17\times$speedup compared with TensorNetwork, indicating the potential of optimized tensor learning primitives for the classical simulation of quantum machine learning algorithms.

Tensor learning is a powerful tool for big data analytics and machine learning, e.g., gene analysis and deep learning. However, tensor learning algorithms are compute-intensive since their time and space complexities grow exponentially with the order of tensors, which hinders their application. In this paper, we exploit the parallelism of tensor learning primitives using GPU tensor cores and develop high-performance tensor learning algorithms. First, we propose novel hardware-oriented optimization strategies for tensor learning primitives on GPU tensor cores. Second, for big data analytics, we employ the optimized tensor learning primitives to accelerate the CP tensor decomposition and then apply it for gene analysis. Third, we optimize the Tucker tensor decomposition and propose a novel Tucker tensor layer to compress deep neural networks. We employ natural gradients to train the neural networks, which only involve a forward pass without backpropagation and thus are suitable for GPU computations. Compared with TensorLab and TensorLy libraries on an A100 GPU, our third-order CP tensor decomposition achieves up to$16.32\times$and$32.25\times$speedups; and$6.09\times$and$6.72\times$speedups for our third-order Tucker tensor decomposition. The proposed fourth-order CP and Tucker tensor decompositions achieve up to$30.65\times$and$5.41\times$speedups over the TensorLab. Our CP tensor decomposition for gene analysis achieves up to$5.88\times$speedup over TensorLy. Compared with a conventional fully connected neural network, our Tucker tensor layer neural network achieves an accuracy of$97.9\%$, a speedup of$4.47\times$, and a compression ratio of$2.92$at the cost of$0.4\%$drop in accuracy.

It is important to estimate the global network traffic data from partial traffic measurements for many network management tasks, including status monitoring and fault detection. However, existing estimation approaches cannot well handle the topological correlations hidden in network traffic and suffer from limited imputation performance. This paper proposes a deep learning approach for network traffic imputation, which well exploits the topological structure of network traffic. We first model the network traffic as a novel graph-tensor and derive a theoretical recovery guarantee. Then we develop an iterative graph-tensor completion algorithm and propose a graph neural network for network traffic imputation by unfolding the iterative algorithm. The proposed graph neural network well captures the topological correlations of network traffic and achieves accurate imputation. Extensive experiments on real-world datasets show that the proposed graph neural network achieves about one-half lower relative square error while at least ten times faster imputation speed than the existing methods.

In recommendation systems, global positioning, system identification and mobile social networks, it is a fundamental routine that a server completes a low-rank matrix from an observed subset of its entries. However, sending data to a cloud server raises up the data privacy concern due to eavesdropping attacks and the single-point failure problem, e.g., the Netflix prize contest was canceled after a privacy lawsuit. In this paper, we propose a homomorphic matrix completion algorithm for privacy-preserving data completion. First, we formulate a \textit{homomorphic matrix completion} problem where a server performs matrix completion on cyphertexts, and propose an encryption scheme that is fast and easy to implement. Secondly, we prove that the proposed scheme satisfies the \textit{homomorphism property} that decrypting the recovered matrix on cyphertexts will obtain the target complete matrix in plaintext. Thirdly, we prove that the proposed scheme satisfies an $(\epsilon, \delta)$-differential privacy property. While with similar level of privacy guarantee, we reduce the best-known error bound $O(\sqrt[10]{n_1^3n_2})$ to EXACT recovery at a price of more samples. Finally, on numerical data and real-world data, we show that both homomorphic nuclear-norm minimization and alternating minimization algorithms achieve accurate recoveries on cyphertexts, verifying the homomorphism property.

Finance is a particularly challenging playground for deep reinforcement learning. However, establishing high-quality market environments and benchmarks for financial reinforcement learning is challenging due to three major factors, namely, low signal-to-noise ratio of financial data, survivorship bias of historical data, and backtesting overfitting. In this paper, we present an openly accessible FinRL-Meta library that has been actively maintained by the AI4Finance community. First, following a DataOps paradigm, we will provide hundreds of market environments through an automatic data curation pipeline that processes dynamic datasets from real-world markets into gym-style market environments. Second, we reproduce popular papers as stepping stones for users to design new trading strategies. We also deploy the library on cloud platforms so that users can visualize their own results and assess the relative performance via community-wise competitions. Third, FinRL-Meta provides tens of Jupyter/Python demos organized into a curriculum and a documentation website to serve the rapidly growing community. FinRL-Meta is available at: \url{https://github.com/AI4Finance-Foundation/FinRL-Meta}

Recommendations based on prediction of user preferences from partial information are widely used in various applications. However, recommendations using smart devices have some challenges related to limited data device resources, data sparsity, and data privacy. Since there are many multidimensional data in smart devices, recommendations may collect a large amount of user private data. In this article, we study privacy-preserving recommendations with high-dimensional tensor data in smart devices. First, we propose a federated tensor completion scheme to infer the user’s preferences and we prove that this scheme satisfies the differential privacy guarantee. Our scheme consists of a global update and a local update, which reduce information exposure and guarantee local data privacy. Second, we mathematically analyze the privacy and utility of the proposed algorithm. Third, we provide empirical evaluations on synthetic data sets and real-world data sets. Results show that our scheme has a low recovery error and provides strong privacy protection.

In intelligent transportation systems (ITS), incomplete traffic data due to sensor malfunctions and communication faults, seriously restricts the related applications of ITS. Recovering missing data from incomplete traffic data becomes an important issue for ITS. Existing works on traffic data imputation cannot achieve satisfactory accuracy due to inefficiently exploiting the underlying topological structure of the traffic data. In this paper, we model the topology of the road network as a graph and introduce graph Fourier transform (GFT) to process the traffic data. Then we utilize an algebraic framework termed as graph-tensor singular value decompositions (GT-SVD) to extract the hidden spatial information of traffic data. Furthermore, we propose a novel graph spectral regularized tensor completion algorithm based on GT-SVD and construct temporal regularized constraints to improve the recovery accuracy. The extensive experimental results on real traffic datasets demonstrate that the proposed algorithm outperforms the state-of-the-art methods under different missing patterns.

Given a target binary function, the binary code search retrieves top-K similar functions in the repository, and similar functions represent that they are compiled from the same source codes. Searching binary code is particularly challenging due to large variations of compiler tool-chains and options and CPU architectures, as well as thousands of binary codes. Furthermore, there are some pivotal issues in current binary code search schemes, including inaccurate text-based or token-based analysis, slow graph matching, or complex deep learning processes. In this paper, we present an unsupervised tensor embedding scheme, Codee, to carry out code search efficiently and accurately at the binary function level. First, we use an NLP-based neural network to generate the semantic-aware token embedding. Second, we propose an efficient basic block embedding generation algorithm based on the network representation learning model. We learn both the semantic information of instructions and the control flow structural information to generate the basic block embedding. Then we use all basic block embeddings in a function to obtain a variable-length function feature vector. Third, we build a tensor to generate function embeddings based on the tensor singular value decomposition, which compresses the variable-length vectors into short fixed-length vectors to facilitate efficient search afterward. We further propose a dynamic tensor compression algorithm to incrementally update the function embedding database. Finally, we use the local sensitive hash method to find the top-$K$similar matching functions in the repository. Compared with state-of-the-art cross-optimization-level code search schemes, such as Asm2Vec and DeepBinDiff, our scheme achieves higher average search accuracy, shorter feature vectors, and faster feature generation performance using four datasets, OpenSSL, Coreutils, libgmp and libcurl. Compared with other cross-platform and cross-optimization-level code search schemes, such as Gemini, Safe, the average recall of our method also outperforms others.

This paper addresses the task of segmenting class-agnostic objects in semi-supervised setting. Although previous detection based methods achieve relatively good performance, these approaches extract the best proposal by a greedy strategy, which may lose the local patch details outside the chosen candidate. In this paper, we propose a novel spatiotemporal graph neural network (STG-Net) to reconstruct more accurate masks for video object segmentation, which captures the local contexts by utilizing all proposals. In the spatial graph, we treat object proposals of a frame as nodes and represent their correlations with an edge weight strategy for mask context aggregation. To capture temporal information from previous frames, we use a memory network to refine the mask of current frame by retrieving historic masks in a temporal graph. The joint use of both local patch details and temporal relationships allow us to better address the challenges such as object occlusions and missing. Without online learning and fine-tuning, our STG-Net achieves state-of-the-art performance on four large benchmarks, demonstrating the effectiveness of the proposed approach.

Recurrent Neural Networks (RNNs) have been widely used in sequence analysis and modeling. However, when processing high-dimensional data, RNNs typically require very large model sizes, thereby bringing a series of deployment challenges. Although various prior works have been proposed to reduce the RNN model sizes, executing RNN models in the resource-restricted environments is still a very challenging problem. In this paper, we propose to develop extremely compact RNN models with fully decomposed hierarchical Tucker (FDHT) structure. The HT decomposition does not only provide much higher storage cost reduction than the other tensor decomposition approaches, but also brings better accuracy performance improvement for the compact RNN models. Meanwhile, unlike the existing tensor decomposition-based methods that can only decompose the input-to-hidden layer of RNNs, our proposed fully decomposition approach enables the comprehensive compression for the entire RNN models with maintaining very high accuracy. Our experimental results on several popular video recognition datasets show that, our proposed fully decomposed hierarchical tucker-based LSTM (FDHT-LSTM) is extremely compact and highly efficient. To the best of our knowledge, FDHT-LSTM, for the first time, consistently achieves very high accuracy with only few thousand parameters (3,132 to 8,808) on different datasets. Compared with the state-of-the-art compressed RNN models, such as TT-LSTM, TR-LSTM and BT-LSTM, our FDHT-LSTM simultaneously enjoys both order-of-magnitude (3,985× to 10,711×) fewer parameters and significant accuracy improvement (0.6% to 12.7%).

Cloud removal is a ubiquitous and important task in remote sensing image processing, which aims at restoring the ground regions shadowed by clouds. It is challenging to remove the clouds for a single satellite image due to the difficulty of distinguishing clouds from white objects on the ground and filling the irregular missing regions with visual consistency. In this article, we propose a novel two-stage cloud removal method. The first stage is cloud segmentation, i.e., extracting the clouds and removing the thin clouds directly using U-Net. The second stage is image restoration, i.e., removing the thick cloud and recovering the corresponding irregular missing regions using generative adversarial network (GAN). We evaluate the proposed scheme on both synthetic images and real satellite images (over$20\,000\, \times \,20\,000$pixels). On synthetic images for cloud coverage less than 40%, the proposed scheme achieves improvements of 0.049–0.078 in Structural SIMilarity (SSIM) and 3.8–6.2 dB in peak signal-to-noise ratio (PSNR), while the$\ell _{1}$-norm error reduces by 49%–78%, compared with a state-of-the-art deep learning method Pix2Pix. On real satellite images, we demonstrate the consistent visual results of the proposed scheme.

Recently, the residual learning strategy has been integrated into the convolutional neural network (CNN) for single image super-resolution (SISR), where the CNN is trained to estimate the residual images. Recognizing that a residual image usually consists of high-frequency details and exhibits cartoon-like characteristics, in this paper, we propose a deep shearlet residual learning network (DSRLN) to estimate the residual images based on the shearlet transform. The proposed network is trained in the shearlet transform-domain which provides an optimal sparse approximation of the cartoon-like image. Specifically, to address the large statistical variation among the shearlet coefficients, a dual-path training strategy and a data weighting technique are proposed. Extensive evaluations on general natural image datasets as well as remote sensing image datasets show that the proposed DSRLN scheme achieves close results in PSNR to the state-of-the-art deep learning methods, using much less network parameters.

High-accuracy location awareness in indoor environments is fundamentally important for mobile computing and mobile social networks. However, accurate radio frequency (RF) fingerprint-based localization is challenging due to real-time response requirements, limited RF fingerprint samples, and limited device storage. In this article, we propose a tensor generative adversarial net (Tensor-GAN) scheme for real-time indoor localization, which achieves improvements in terms of localization accuracy and storage consumption. First, with verification on real-world fingerprint data set, we model RF fingerprints as a 3-D low-tubal-rank tensor to effectively capture the multidimensional latent structures. Second, we propose a novel Tensor-GAN that is a three-player game among a regressor, a generator, and a discriminator. We design a tensor completion algorithm for the tubal-sampling pattern as the generator that produces new RF fingerprints as training samples, and the regressor estimates locations for RF fingerprints. Finally, on real-world fingerprint data set, we show that the proposed Tensor-GAN scheme improves localization accuracy from 0.42 m (state-of-the-art methods kNN, DeepFi, and AutoEncoder) to 0.19 m for 80% of 1639 random testing points. Moreover, we implement a prototype Tensor-GAN that is downloaded as an Android smartphone App, which has a relatively small memory footprint, i.e., 57 KB.

Snapshot compressive imaging (SCI) cameras capture high-speed videos by compressing multiple video frames into a measurement frame. However, reconstructing video frames from the compressed measurement frame is challenging. The existing state-of-the-art reconstruction algorithms suffer from low reconstruction quality or heavy time consumption, making them not suitable for real-time applications. In this paper, exploiting the powerful learning ability of deep neural networks (DNN), we propose a novel Tensor Fast Iterative Shrinkage-Thresholding Algorithm Net (Tensor FISTA-Net) as a decoder for SCI video cameras. Tensor FISTA-Net not only learns the sparsest representation of the video frames through convolution layers, but also reduces the reconstruction time significantly through tensor calculations. Experimental results on synthetic datasets show that the proposed Tensor FISTA-Net achieves average PSNR improvement of 1.63∼3.89dB over the state-of-the-art algorithms. Moreover, Tensor FISTA-Net takes less than 2 seconds running time and 12MB memory footprint, making it practical for real-time IoT applications.

Audio adversarial examples, imperceptible to humans, have been constructed to attack automatic speech recognition (ASR) systems. However, the adversarial examples generated by existing approaches usually incorporate noticeable noises, especially during the periods of silences and pauses. Moreover, the added noises often break temporal dependency property of the original audio, which can be easily detected by state-of-the-art defense mechanisms. In this paper, we propose a new Iterative Proportional Clipping (IPC) algorithm that preserves temporal dependency in audios for generating more robust adversarial examples. We are motivated by an observation that the temporal dependency in audios imposes a significant effect on human perception. Following our observation, we leverage a proportional clipping strategy to reduce noise during the low-intensity periods. Experimental results and user study both suggest that the generated adversarial examples can significantly reduce human-perceptible noises and resist the defenses based on the temporal structure.

Query-based moment localization is a new task that localizes the best matched segment in an untrimmed video according to a given sentence query. In this localization task, one should pay more attention to thoroughly mine visual and linguistic information. To this end, we propose a novel Cross- and Self-Modal Graph Attention Network (CSMGAN) that recasts this task as a process of iterative messages passing over a joint graph. Specifically, the joint graph consists of Cross-Modal interaction Graph (CMG) and Self-Modal relation Graph (SMG), where frames and words are represented as nodes, and the relations between cross- and self-modal node pairs are described by an attention mechanism. Through parametric message passing, CMG highlights relevant instances across video and sentence, and then SMG models the pairwise relation inside each modality for frame (word) correlating. With multiple layers of such a joint graph, our CSMGAN is able to effectively capture high-order interactions between two modalities, thus enabling a further precise localization. Besides, to better comprehend the contextual details in the query, we develop a hierarchical sentence encoder to enhance the query understanding. Extensive experiments on four public datasets demonstrate the effectiveness of our proposed model, and GCSMAN significantly outperforms the state-of-the-arts.

Video frame synthesis is an important task in computer vision and has drawn great interests in wide applications. However, existing neural network methods do not explicitly impose tensor low-rankness of videos to capture the spatiotemporal correlations in a high-dimensional space, while existing iterative algorithms require hand-crafted parameters and take relatively long running time. In this paper, we propose a novel multi-phase deep neural network Transform-Based Tensor-Net that exploits the low-rank structure of video data in a learned transform domain, which unfolds an Iterative Shrinkage-Thresholding Algorithm (ISTA) for tensor signal recovery. Our design is based on two observations: (i) both linear and nonlinear transforms can be implemented by a neural network layer, and (ii) the soft-thresholding operator corresponds to an activation function. Further, such an unfolding design is able to achieve nearly real-time at the cost of training time and enjoys an interpretable nature as a byproduct. Experimental results on the KTH and UCF-101 datasets show that compared with the state-of-the-art methods, i.e., DVF and Super SloMo, the proposed scheme improves Peak Signal-to-Noise Ratio (PSNR) of video interpolation and prediction by 4.13 dB and 4.26 dB, respectively.

Purchase-related micro-behaviors, e.g., favorite, add to cart, read reviews, etc., provide implicit feedback of users' decision-making process. Such informative feedback can lead to fine-grained post-click conversion rate (CVR) modeling of the buying process. However, most existing works on CVR estimation either neglect these informative feedback, or model them as a sequential pattern with Recurrent Neural Networks. We argue such modeling could be inappropriate since different orders of micro-behaviors may represent similar user buying intention, and micro-behaviors often correlate with each other.

Post-click conversion rate (CVR) estimation is a critical task in e-commerce recommender systems. This task is deemed quite challenging under industrial setting with two major issues: 1) selection bias caused by user self-selection, and 2) data sparsity due to the rare click events. A successful conversion typically has the following sequential events: ”exposure → click → conversion”. Conventional CVR estimators are trained in the click space, but inference is done in the entire exposure space. They fail to account for the causes of the missing data and treat them as missing at random. Hence, their estimations are highly likely to deviate from the real values by large. In addition, the data sparsity issue can also handicap many industrial CVR estimators which usually have large parameter spaces.

Today, the knowledge base question answering (KB-QA) system is promising to achieve a large-scale high-quality reply in the e-commerce industry. However, there exist two major challenges to efficiently support large-scale KB-QA systems. On the one hand, it is difficult to serve tens of thousands of online stores (i.e., constrained by the tuning and deployment time), and it would perform poorly if the systems start without a sufficient number of chat records. On the other hand, current KB-QA systems cannot be updated in an efficient way due to the high cost of knowledge base (KB) updating. In this article, we propose an automatic learning scheme for KB-QA systems, called ALKB-QA , using a vector modeling method to address the preceding two main challenges. The ALKB-QA system provides online stores with basic KB templates that are suitable for many common occasions, and this feature enables the ability to deploy chatbots for a large number of online stores in a short time. Then, the KBs are further updated automatically to adapt to their own businesses (meet different specific needs), leading to increased reply accuracy. Our work has three main contributions. First, the proposed ALKB-QA system has a good business model in the e-commerce industry (serving tens of thousands of online stores with low cost), breaking the scalability limitations of existing KB-QA systems. Second, we assess the reply accuracy of the proposed ALKB-QA system using human evaluations, and the results show that it outperforms human annotation-base approaches. Third, we launched our ALKB-QA system as a real-world business application, and it supports tens of thousands of online stores.

With the rapid development and proliferation of multimedia systems and applications, there is a growing need to handle multimedia heterogeneous data safely and efficiently on the cloud. Tensor models are effective in representing multimedia multidimensional data, and the tensor decomposition is one of the basic building blocks of data analysis and learning models. In this article, we propose a secure tensor singular value decomposition (${S}$-tSVD), in which the time-domain operation is converted into a scheme featuring frequency-domain multilinear circular unfolding–folding. First, we represent various multimedia data as cipher subtensors, using fully homomorphic encryption. We then take the fast Fourier transform (FFT) approach to launch a new multiplication operation along the tubal fibers of a unified high-order tensor. Second, relying on the homomorphism of addition and multiplication theory, we prove the fully homomorphic consistency of the proposed${S}$-tSVD algorithm. Third, we provide an elegant solution to tackle the typical dimensionality inconsistency problem while working with multiple subtensors. Finally, we carry out theoretical analyses with respect to dimensionality reduction, reconstruction error of${S}$-tSVD, running time, and data security. We use real unstructured video data and semistructured XML documents, integrating them within a unified tensor model for decomposition. We demonstrate that the error ratio of the${S}$-tSVD is lower than the same compression ratio compared to the SVD decomposition and tSVD-slice approaches. Moreover, the specific${S}$-tSVD decomposition not only enables effective data mining and dimensionality reduction but also ensures the accuracy of the decomposition result and data privacy protection.

In a vertical industry alliance, Internet of Things (IoT) deployed in different smart factories are similar. For example, most automobile manufacturers have the similar assembly lines and IoT surveillance systems. It is common to observe the industrial knowledge using deep learning and data mining methods based on the IoT data. However, some knowledge is not easy to be mined from only one factory's data because the samples are still few. If multiple factories within an alliance can gather their data together, more knowledge could be mined. However, the key concern of these factories is the data security. Existing matrix-based methods can guarantee the data security inside a factory but do not allow the data sharing among factories, and thus their mining performance is poor due to lack of correlation. To address this concern, in this article we propose the novel federated tensor mining (FTM) framework to federate multisource data together for tensor-based mining while guaranteeing the security. The key contribution of FTM is that every factory only needs to share its ciphertext data for security issue, and these ciphertexts are adequate for tensor-based knowledge mining due to its homomorphic attribution. Real-data-driven simulations demonstrate that FTM not only mines the same knowledge compared with the plaintext mining, but also is enabled to defend the attacks from distributed eavesdroppers and centralized hackers. In our typical experiment, compared with the matrix-based privacy-preserving compressive sensing (PPCS), FTM increases up to 24% on mining accuracy.

Snapshot compressive imaging (SCI) is a promising approach to capture high-dimensional data with low dimensional sensors. With modest modifications to off-the-shelf cameras, SCI cameras encode multiple frames into a single measurement frame. These correlated frames can then be retrieved by reconstruction algorithms. Existing reconstruction algorithms suffer from low speed or low fidelity. In this paper, we propose a novel reconstruction algorithm, namely, Shearlet enhanced Snapshot Compressive Imaging (SeSCI), which exploits the sparsity of the image representation in both frequency domain and shearlet domain. Towards this end, we first derive our SeSCI algorithm under the alternating direction method of multipliers (ADMM) framework. We then propose an efficient solution of SeSCI algorithm. Moreover, we prove that the improved SeSCI algorithm converges to a fixed point. Experimental results on both synthetic data and real data captured by SCI cameras demonstrate the significant advantages of SeSCI, which outperforms the conventional algorithms by more than 2dB in PSNR. At the same time, the SeSCI achieves a speed-up more than 100× over the state-of-the-art algorithm.

The low-tubal-rank tensor model has been recently proposed for real-world multidimensional data. In this paper, we study the low-tubal-rank tensor completion problem, i.e., to recover a third-order tensor by observing a subset of its elements selected uniformly at random. We propose a fast iterative algorithm, called Tubal-AltMin, that is inspired by a similar approach for low-rank matrix completion. The unknown low-tubal-rank tensor is represented as the product of two much smaller tensors with the low-tubal-rank property being automatically incorporated, and Tubal-AltMin alternates between estimating those two tensors using tensor least squares minimization. First, we note that tensor least squares minimization is different from its matrix counterpart and nontrivial as the circular convolution operator of the low-tubal-rank tensor model is intertwined with the sub-sampling operator. Secondly, the theoretical performance guarantee is challenging since Tubal-AltMin is iterative and nonconvex. We prove that 1) Tubal-AltMin generates a best rank-r approximate up to any predefined accuracy ε at an exponential rate, and 2) for an n × n × k tensor M with tubal-rank r ≪ n, the required sampling complexity is O((nr2kIIMIIF2log3n)/σ2rk), where σ̅rk is the rk-th singular value of the block diagonal matrix representation of M in the frequency domain, and the computational complexity is O(n2r2k3logn log(n/ε)). Finally, on both synthetic data and real-world video data, evaluation results show that compared with tensor-nuclear norm minimization using alternating direction method of multipliers (TNN-ADMM), Tubal-AltMin-Simple (a simplified implementation of Tubal-AltMin) improves the recovery error by several orders of magnitude. In experiments, Tubal-AltMin-Simple is faster than TNN-ADMM by a factor of 5 for a 200 × 200 × 20 tensor.

In recent years, end-to-end network latency estimation has attracted much attention because of its significance for network performance evaluation. Given the widespread use of personal devices, latency estimation from partially observed samples becomes more complicated due to unstable communication conditions, while measuring the latencies between all nodes in a large-scale network is infeasible and costly. Hence, reducing the measurement cost becomes critical for the latency estimation of personal device network. In this paper, we propose an adaptive sampling scheme based on leverage scores to reduce the measurement cost while achieving high estimation accuracy. Furthermore, we provide theoretical analysis to characterize the performance bounds of the proposed scheme in terms of sampling complexity and estimation error. Finally, we demonstrate the efficiency of the proposed scheme by conducting extensive simulations on both synthetic and real datasets. The results show that the proposed scheme is able to not only improve the estimation accuracy of network latency but also reduce the sample budget compared to the state-of-the-art approaches.

Data completion is a problem of filling missing or unobserved elements of partially observed datasets. Data completion algorithms have received wide attention and achievements in diverse domains including data mining, signal processing, and computer vision. We observe a ubiquitous tubal-sampling pattern in big data and Internet of Things (IoT) applications, which is introduced by many reasons such as high data acquisition cost, downsampling for data compression, sensor node failures, and packet losses in low-power wireless transmissions. To meet the time and accuracy requirements of applications, data completion methods are expected to be accurate as well as fast. However, the existing methods for data completion with the tubal-sampling pattern are either accurate or fast, but not both. In this article, we propose high-performance graphics processing unit (GPU) tensor completion for data completion with the tubal-sampling pattern. First, by exploiting the convolution theorem, we split a tensor least-squares minimization problem into multiple least-squares sub-problems in the frequency domain. In this way, massive parallelisms are exposed for many-core GPU architectures while still preserving high recovery accuracy. Second, we propose computing slice-level and tube-level tasks in batches to improve GPU utilization. Third, we reduce the data transfer cost by eliminating the accesses to the CPU memory inside algorithm loop structures. The experimental results show that the proposed tensor completion is both fast and accurate. Using synthetic data of varying sizes, the proposed GPU tensor completion achieves maximum 248.18×, 7, 403.27×, and 33.27× speedups over the CPU MATLAB implementation, GPU element-sampling tensor completion in the cuTensor-tubal library, and GPU high-performance matrix completion, respectively. With a 50 percent sampling rate, the proposed GPU tensor completion achieves a recovery error of 1.40e-5, which is comparable with that of the GPU element-sampling tensor completion and three orders of magnitude better than that of the GPU high-performance matrix completion. To utilize multiple GPUs in servers, we design a multi-GPU scheme for tubal-sampling tensor completion. The multi-GPU tensor completion achieves maximum 1.89× speedup on two GPUs versus on a single GPU for medium or big tensors. We further evaluate the performance of the proposed GPU tensor completion in three real applications, namely, video transmission in wireless camera networks, RF fingerprint-based indoor localization, and seismic data completion, and it achieves maximum speedups of 448.68×, 24.63×, and 311.54×, respectively. We integrate this high-performance GPU tensor completion implementation into the cuTensor-tubal library to support various applications.

Tensors are the cornerstone data structures in high-performance computing, big data analysis and machine learning. However, tensor computations are compute-intensive and the running time increases rapidly with the tensor size. Therefore, designing high-performance primitives on parallel architectures such as GPUs is critical for the efficiency of ever growing data processing demands. Existing GPU basic linear algebra subroutines (BLAS) libraries (e.g., NVIDIA cuBLAS) do not provide tensor primitives. Researchers have to implement and optimize their own tensor algorithms in a case-by-case manner, which is inefficient and error-prone. In this paper, we develop the cuTensor-tubal library of seven key primitives for the tubal-rank tensor model on GPUs: t-FFT, inverse t-FFT, t-product, t-SVD, t-QR, t-inverse, and t-normalization. cuTensor-tubal adopts a frequency domain computation scheme to expose the separability in the frequency domain, then maps the tube-wise and slice-wise parallelisms onto the single instruction multiple thread (SIMT) GPU architecture. To achieve good performance, we optimize the data transfer, memory accesses, and design the batched and streamed parallelization schemes for tensor operations with data-independent and data-dependent computation patterns, respectively. In the evaluations oft-product, t-SVD, t-QR, t-inverse and t-normalization, cuTensor-tubal achieves maximum 16.91x, 27.03x, 38.97x, 22.36x,15.43x speedups respectively over the CPU implementations running on dual 10-core Xeon CPUs. Two applications, namely, t-SVD-based video compression and low-tubal-rank tensor completion, are tested using our library and achieve maximum 9.80x and 269.26x speedups over multi-core CPU implementations.

AI technologies are helping more and more companies leverage their resources to expand business, reach higher financial performance and become more valuable for investors. However, it is difficult to capture and predict the impacts of AI technologies on companies' stock prices through traditional financial factors. Moreover, common information sources such as company's earnings calls and news are not enough to quantity and predict the actual AI premium for a certain company. In this paper, we utilize scholar data as alternative data for trading strategy development and propose a practical machine learning approach to quantity the AI premium of a company and capture the scholar data driven alpha in the AI industry. First, we collect the scholar data from the Microsoft Academic Graph database, and conduct feature engineering based on AI publication and patent data, such as conference/journal publication counts, patent counts, fields of studies and paper citations. Second, we apply machine learning algorithms to weight and re-balance stocks using the scholar data and traditional financial factors every month, and construct portfolios using the “buy-and-hold-long only” strategy. Finally, we evaluate our factor and portfolio in terms of factor performance and portfolio's cumulative return. The proposed scholar data driven approach achieves a cumulative return of 1029.1% during our backtesting period, which significantly outperforms the Nasdaq 100 index's 529.5% and S&P 500's 222.6%. The traditional financial factors approach only leads to 776.7%, which indicates that our scholar data driven approach is better at capturing investment alpha in AI industry than traditional financial factors.

In this paper, we propose a novel method for generating high-granularity three-dimensional (3D) seismic data from low-granularity data based on tensor sparse coding, which jointly trains a high-granularity dictionary and a low-granularity dictionary. First, considering the high-dimensional properties of seismic data, we introduce tensor sparse coding to seismic data interpolation. Second, we propose that the dictionary pairs trained by low-granularity seismic data and high-granularity seismic data have the same sparse representation, which are used to recover high-granularity data with the high-granularity dictionary. Finally, experiments on the seismic data of an actual field show that the proposed method effectively perform seismic trace interpolation and can improve the resolution of seismic data imaging.

In this paper, we optimize the computations of third-order low-tubal-rank tensor operations on many-core GPUs. Tensor operations are compute-intensive and existing studies optimize such operations in a case-by-case manner, which can be inefficient and error-prone. We develop and optimize a BLAS-like library for the low-tubal-rank tensor model called cuTensor-tubal, which includes efficient GPU primitives for tensor operations and key processes. We compute tensor operations in the frequency domain and fully exploit tube-wise and slice-wise parallelisms. We design, implement, and optimize four key tensor operations namely t-FFT, inverse t-FFT, t-product, and t-SVD. For t-product and t-SVD, cuTensor-tubal demonstrates significant speedups: maximum 29.16 ×, 6.72× speedups over the non-optimized GPU counterparts, and maximum 16.91× and 27.03× speedups over the CPU implementations running on dual 10-core Xeon CPUs.

Tensor provides a brief and natural representation for large-scale multidimensional data by way of appropriate low-rank approximations, thus we can discover significant latent structures of complex data and generalize data representation. To date, tensor has gained tremendous success in various science and technology fields, especially in machine learning and big data applications. However, tensor computation, especially tensor decomposition, is usually expensive due to the inherent large-size characteristic of tensors, and hence would potentially hinder their future wide deployment. In this paper, we develop a hardware architecture to accelerate tensor singular value decomposition (t-SVD), which is a new tensor decomposition technique that has been successfully applied to high-dimensional data classification and video recovery. Specifically, design consideration of each key computing unit is analyzed and discussed. Then, the proposed t-SVD hardware architecture is implemented and synthesized using CMOS 28nm technology. Comparison with real-world CPU-based implementations shows that the proposed hardware accelerator is expected to provide average 14× speedup on various t-SVD workloads.

Snapshot compressive imaging (SCI) systems have been developed to capture high-dimensional (≥ 3) signals using low-dimensional off-the-shelf sensors, i.e., mapping multiple video frames into a single measurement frame. One key module of a SCI system is an accurate decoder that recovers the original video frames. However, existing model-based decoding algorithms require exhaustive parameter tuning with prior knowledge and cannot support practical applications due to the extremely long running time. In this paper, we propose a deep tensor ADMM-Net for video SCI systems that provides high-quality decoding in seconds. Firstly, we start with a standard tensor ADMM algorithm, unfold its inference iterations into a layer-wise structure, and design a deep neural network based on tensor operations. Secondly, instead of relying on a pre-specified sparse representation domain, the network learns the domain of low-rank tensor through stochastic gradient descent. It is worth noting that the proposed deep tensor ADMM-Net has potentially mathematical interpretations. On public video data, the simulation results show the proposed method achieves average 0.8 ~ 2.5 dB improvement in PSNR and 0.07 ~ 0.1 in SSIM, and 1500× ~ 3600× speedups over the state-of-the-art methods. On real data captured by SCI cameras, the experimental results show comparable visual results with the state-of-the-art methods but in much shorter running time.

In this paper, we develop a new library TenDeC++ for tensor decompositions in C++. TenDeC++ supports popular tensor decomposition functions including Canonical Polyadic, Tucker, tensor-train, and t-SVD, assisting C++ programmers to shorten the development cycle of deep learning applications. Compared with the resource-intensive Python and MATLAB, C++ has the nature advantages on fast running time and high compatibility. To further explore potentials of C++, we propose a novel underlying technology PointerDefomer leveraging the unique pointer. Since the transformation between tensor and size-specific matrix is indispensable in tensor decompositions, PointerDefomer can virtually achieve such a transformation by controlling the movement of pointer in memory address. As a result, the conventional transformation steps can be skipped to accelerate the decomposition process and there is no memory needed for saving the intermediate results of tensor transformation. In our experiment, TenDeC++ reduces decomposition time and support larger size of tensor compared with the classic Tensorly in Python and TensorLab in MATLAB, respectively.

In recent years Wi-Fi fingerprinting has attracted much attention in indoor localization because of the availability of high-quality signal and pervasive deployment of wireless LANs. For fingerprint-based localization, however, offline site survey is usually time-consuming and labor-intensive. Therefore, reducing the burden of offline site survey becomes an important issue for fingerprint-based indoor localization. In this paper, using a low-tubal-rank tensor to model Wi-Fi fingerprints of all reference points (RPs), we propose an adaptive sampling scheme via approximate volume sampling to improve reconstruction accuracy of radio map with reduced expenditure. We propose a rank-increasing strategy to effectively estimate the rank of the underlying fingerprint tensor to alleviate the computation burden for tensor completion. We provide a theoretical foundation to analyze the proposed scheme and derive the performance bounds in terms of sample complexity and reconstruction error. We prove that the proposed scheme can achieve a relative error guarantee. Finally, we validate the effectiveness of the proposed scheme through extensive simulations using both synthetic and real datasets. The simulation results demonstrate that the proposed scheme is able to not only reduce reconstruction error and improve localization accuracy but also reduce running time compared to the state-of-the-art schemes.

Envisioning the potentials of energy harvesting technology and the improved spectrum reuse by joint utilization of overlay and underlay modes, this paper studies the throughput performance of a novel cognitive radio network (CRN) scenario with a mobile energy-harvesting secondary transmitter (ST). The hybrid overlay-underlay scheme allows the secondary users to access the spectrum even when the primary signal is detected. We are the first to partition the unit area into three parts for secondary users: overlay mode area, underlay mode area, and harvesting zone. Then, we propose a metric to classify the CRN into the spectrum-limited state and the energy-limited state, and accordingly maximize the throughput through the monotonicity analysis of throughput and collision probability. The secondary throughput is maximized under the energy constraint and collision constraint. Moreover, we quantitatively discuss the impacts of underlay mode transmission on the classification of network states and the corresponding optimal spectrum sensing, respectively. We find that with a relatively small detection threshold, ST transmits the considerable amount of packets in underlay mode, while it transmits few packets in overlay mode. Theoretical results are validated by simulations, and our findings shed light on the design and operation of mobile energy-harvesting CRNs.

Today search engines are tightly coupled with social networks, and present users with a double-edged sword: they are able to acquire information interesting to users but are also capable of spreading viruses introduced by hackers. It is challenging to characterize how a search engine spreads viruses, since the search engine serves as a virtual virus pool and creates propagation paths over the underlying network structure. In this paper, we quantitatively analyze virus propagation effects and the stability of the virus propagation process in the presence of a search engine. First, although social networks have a community structure that impedes virus propagation, we find that a search engine generates a propagation wormhole. Second, we propose an epidemic feedback model and quantitatively analyze propagation effects based on a model employing four metrics: infection density, the propagation wormhole effect, the epidemic threshold, and the basic reproduction number. Third, we verify our analyses on four real-world data sets and two simulated data sets. Moreover, we prove that the proposed model has the property of partial stability. Evaluation results show that, compared the cases without a search engine, virus propagation with the search engine has a higher infection density, shorter network diameter, greater propagation velocity, lower epidemic threshold, and larger basic reproduction number.

As efficient traffic-management platforms, public vehicle (PV) systems are envisioned to be a promising approach to solving traffic congestion and pollution for future smart cities. PV systems provide online/dynamic peer-to-peer ride-sharing services with the goal of serving a sufficient number of customers with a minimum number of vehicles and the lowest possible cost. A key component of the PV system is the online ride-sharing scheduling strategy. In this paper, an efficient path-planning strategy based on a greedy algorithm is proposed, which focuses on a limited potential search area for each vehicle by filtering out the requests that violate the passenger service quality level, so that the global search is reduced to a local search. Moreover, the proposed heuristic can be easily used in the future globally optimal algorithm (if it will exist) to speed the computation time. The performance of the proposed solution, such as reduction ratio of computational complexity, is analyzed. Simulations based on the Manhattan taxi data set show that the computing time is reduced by 22% compared with the exhaustive search method under the same service quality performance.

Energy-efficient and secure wireless communications have recently earned tremendous interests due to economic, environmental, and military concerns. This paper investigates the tradeoff between the secrecy throughput and the energy efficiency in cognitive radio networks (CRNs), where primary and secondary users with different priorities of spectrum access can either interfere or cooperate with each other. To gain an understanding of the intricate effects that system parameters have on underlay network's performance, we exclusively focus on characterizing several key aspects that may have potential impacts on secure underlay CRNs, including the transmission power, the number of interfering users, and the designed interference resistance coefficient. Based on the obtained analytical results, we further propose a cooperative spectrum sharing paradigm to improve both the secrecy throughput and the energy efficiency of primary users. The main idea is that primary users allow secondary users to simultaneously access the licensed spectrum and in return, the secondary transmitter acts as both a relay for primary transmissions and a friendly jammer against eavesdropping, in case the primary transmission fails. Both theoretical and numerical results reveal that: (i) When the interference from secondary transmitters is small, there is an optimal transmission power that maximizes the secrecy throughput for primary users compared to CRNs without the security issue. (ii) When the interference from secondary transmitters is large, the secrecy throughput increases with the transmission power for primary users. (iii) The transmission power that maximizes the energy efficiency is smaller than that maximizes the secrecy throughput for primary users. (iv) The number of interfering users has a slight impact on the secrecy throughput and the energy efficiency of primary users due to the secondary power control. (v) The proposed cooperative paradigm is an efficient approach to boost both the secrecy throughput and the energy efficiency of primary users compared with the traditional non-cooperative spectrum sharing, and provides an alternative method to compensate for the interference caused by secondary users.

This paper characterizes the throughput and delay performance of Cognitive Radio Networks (CRNs), where both primary and secondary networks coexist in a unit torus. Specifically, the primary network consists of static primary nodes (PNs) of density n, which have a higher priority to access the spectrum. In contrast, the secondary network consists of mobile secondary nodes (SNs) of density m = nβwithβ≥ 1, which move according to a hybrid random walk mobility model and have opportunistic access to the spectrum without affecting primary packet transmissions. Motivated by the fact that cooperation between primary and secondary nodes leads to possible improvement on the performance of CRNs, as well as the fact that the heterogeneous moving regions of secondary nodes will bring about further improvement, we propose a novel hierarchical cooperative scheduling mechanism, where secondary nodes serve as relays for primary packet transmissions by exploiting their mobility heterogeneity and geographic information. Our findings include: (i) For the primary network, stronger mobility heterogeneity of secondary nodes leads to better delay performance of the primary network, and meanwhile the delay scaling can be significantly reduced to Θ (n√(β/(4 log n)) log3/2n) when a near-optimal per-node throughput of Θ(1/log n) is obtained. (ii) For the secondary network, we also adopt a similar hierarchical cooperative scheduling mechanism, and obtain a near-optimal per-node throughput of Θ (1/log m) with the delay scaling of Θ(m1-(1√logm)). (iii) The delay of secondary source-destination pairs is determined by the moving region of destinations and has no relation with sources. Our work provides deeper understandings of the cooperation, heterogeneous mobility, and geographic information on the performance of CRNs, and sheds light on designing more efficient CRNs.

Due to the random and dynamic energy-harvesting process, it is challenging to conduct optimal rate control in Energy-Harvesting Communication Systems (EHCSs). Existing works mainly focus on two cases: (1) the traffic load is infinite (as long as there is energy, there is data to transmit), in which the objective is to optimize the rate control policy subject to the dynamic energy arrivals, thus maximizing the average system throughput; and (2) the traffic load is finite, in which the objective is to optimize the rate control policy, thus minimizing the time by which all packets are delivered. In this work, we focus on the optimal rate control of EHCSs from another important and practical perspective, where the data and energy arrivals are both random. Given any deadline of T , our goal is to maximize the total throughput in [0, T ]. Specifically, two scenarios are considered: (1) energy is ready before the transmission; and (2) energy arrives randomly during the transmission. In both scenarios, we assume that the data arrive randomly during the transmission. For the first scenario, we develop a novel Stepwise Searching Algorithm (SSA) based on the cumulative curve methodology, which is shown to achieve the optimal solution and the complexity grows only linearly with the problem size. In addition, the SSA can provide a simple and appealing graphical visualization of approximating the optimal solution. For the second scenario, we provide a simplified case study that can be solved by the SSA with low computation overhead and demonstrate the difficulties in solving the general setting, which initiates a first step toward the full understanding of the scenario when energy arrives randomly during the transmission.

In this paper, we propose two novel multi-dimensional tensor sparse coding (MDTSC) schemes using the t-linear combination. Based on the t-linear combination, the shifted versions of the bases are used for the data approximation, but without need to store them. Therefore, the dictionaries of the proposed schemes are more concise and the coefficients have richer physical explanations. Moreover, we propose an efficient alternating minimization algorithm, including the tensor coefficient learning and the tensor dictionary learning, to solve the proposed problems. For the tensor coefficient learning, we design a tensor-based fast iterative shrinkage algorithm. For the tensor dictionary learning, we first divide the problem into several nearly-independent subproblems in the frequency domain, and then utilize the Lagrange dual to further reduce the number of optimization variables. Experimental results on multi-dimensional signals denoising and reconstruction (3DTSC, 4DTSC, 5DTSC) show that the proposed algorithms are more efficient and outperform the state-of-the-art tensor-based sparse coding models.

In this paper, we propose a novel low-rank tensor completion (LRTC) model under the circulant algebra for color image inpainting, which simultaneously preserves the low-rank structures of images, and also explore the local smooth and piecewise priors of the images in the spatial domain. First, color images are naturally represented by 3-order tensors which preserve the intrinsic structures of color images. Second, we preserve the low-rank structures of these tensors with tensor nuclear norm, which can simultaneously exploit the correlations among the spatial and channel domains. Third, we integrate an anisotropic total variation into our low-rank tensor completion model, which preserve the local smooth and piecewise priors of color images. Then, an efficient alternating direction method of multipliers (ADMM) is proposed to solve the resulting optimization problem. Experimental results on eight widely used color images demonstrate the effectiveness and superiority of the proposed algorithm.

The problem of testing whether an incomplete tensor lies in a given tensor subspace, called tensor matched subspace detection, is significant when it is unavoidable to have missing entries. Compared with the matrix case, the tensor matched subspace detection problem is much more challenging due to the curse of dimensionality and the intertwinement between the sampling operator and the tensor product operation. In this paper, we investigate the subspace detection problem for the transform-based tensor models. Under this framework, tensor subspaces and the orthogonal projection onto a given subspace are defined, and the energies of a tensor outside the given subspace (also called residual energy in statistics) with tubal-sampling and elementwise-sampling are derived. We have proved that the residual energy of sampling signals is bounded with high probability. Based on the residual energy, the reliable detection is feasible.

Localization technology is important for the development of indoor location-based services (LBS). Global Positioning System (GPS) becomes invalid in indoor environments due to the non-line-of-sight issue, so it is urgent to develop a real-time high-accuracy localization approach for smartphones. However, accurate localization is challenging due to issues such as real-time response requirements, limited fingerprint samples and mobile device storage. To address these problems, we propose a novel deep learning architecture: Tensor-Generative Adversarial Network (TGAN). We first introduce a transform-based 3D tensor to model fingerprint samples. Instead of those passive methods that construct a fingerprint database as a prior, our model applies artificial neural network with deep learning to train network classifiers and then gives out estimations. Then we propose a novel tensorbased super-resolution scheme using the generative adversarial network (GAN) that adopts sparse coding as the generator network and a residual learning network as the discriminator. Further, we analyze the performance of TGAN and implement a trace-based localization experiment, which achieves better performance. Compared to existing methods for smartphones indoor positioning, that are energy- consuming and high demands on devices, TGAN can give out an improved solution in localization accuracy, response time and implementation complexity.

Radio-frequency (RF) tomographic imaging is a promising technique for inferring multi-dimensional physical space by processing RF signals traversed across a region of interest. However, conventional RF tomography schemes are generally based on vector compressed sensing, which ignores the geometric structures of the target spaces and leads to low recovery precision. The recently proposed transform-based tensor model is more appropriate for sensory data processing, as it helps exploit the geometric structures of the three-dimensional target and improve the recovery precision. In this paper, we propose a novel tensor sensing approach that achieves highly accurate estimation for real-world three-dimensional spaces. First, we use the transform-based tensor model to formulate a tensor sensing problem, and propose a fast alternating minimization algorithm called Alt-Min. Secondly, we drive an algorithm which is optimized to reduce memory and computation requirements. Finally, we present evaluation of our Alt-Min approach using IKEA 3D data and demonstrate significant improvement in recovery error and convergence speed compared to prior tensor-based compressed sensing.

The emerging Internet of Things (IoT) systems are fueling an exponential explosion of sensory data. The major challenge to effective implementation of IoT systems is the presence of massive missing data entries, measurement noise, and anomaly readings, which motivates us to investigate the robust recovery of sensory big data. In this paper, we propose an LS-decomposition approach that decomposes a sensory reading matrix as the superposition of a Low-rank matrix and a Sparse anomaly matrix. First, based on data sets from three representative real-world IoT projects, i.e., the IntelLab project (indoor environment), the GreenOrbs project (mountain environment), and the NBDC-CTD project (ocean environment), we observe that anomaly readings are ubiquitous and cannot be ignored. Second, we prove that the convex surrogate of the LS-decomposition problem guarantees bounded recovery error under proper conditions. Third, we propose an accelerated proximal gradient algorithm that converges to the optimal solution at a rate that is inversely proportional to the square of the number of iterations. Evaluations on the above three data sets show that the proposed scheme achieves (relative) recovery error ≤ 0.05 for missing data rate ≤ 50 percent and almost exact recovery for missing data rate ≤ 40 percent, while previous methods have (relative) recovery error 0.04 ~0.15 even at only 10 percent missing data rate.

Public vehicle (PV) systems are promising transportation systems for future smart cities which provide dynamic ride-sharing services according to passengers' requests. PVs are driverless/self-driving electric vehicles which require frequent recharging from smart grids. For such systems, the challenge lies in both the efficient scheduling scheme to satisfy transportation demands with service guarantee and the cost-effective charging strategy under the real-time electricity pricing. In this paper, we study the joint transportation and charging scheduling for PV systems to balance the transportation and charging demands, ensuring the long-term operation. We adopt a cake cutting game model to capture the interactions among PV groups, the cloud and smart grids. The cloud announces strategies to coordinate the allocation of transportation and energy resources among PV groups. All the PV groups try to maximize their joint transportation and charging utilities. We propose an algorithm to obtain the unique normalized Nash equilibrium point for this problem. Simulations are performed to confirm the effects of our scheme under the real taxi and power grid data sets of New York City. Our results show that our scheme achieves almost the same transportation performance compared with a heuristic scheme, namely, transportation with greedy charging; however, the average energy price of the proposed scheme is 10.86% lower than the latter one.

Many realistic wireless social networks are evolving over time. While network evolution has its important influence on network performances, it is nevertheless overlooked in most existing studies on information diffusion. Motivated by this, in this paper, we investigate the delivery accuracy of interest-aware information diffusion in evolving social networks. In doing so, we adopt a model, named affiliation networks, to characterize network evolution from three aspects, i.e., the arrival of new users, the generation of new interests, and the creation of new links between them. Based on that, we consider a publishing based information diffusion mechanism that widely exists in wireless networking services such as Facebook, Twitter, and Sina Weibo, where a user receives data items from his friends and then republishes the ones he is interested in to all his friends. Under the above network model, we study how the performance metric such as delivery accuracy is affected by the network evolution. The publishing based information diffusion mechanism is a blind targeting one that may suffer a low delivery accuracy. However, our analytical results demonstrate a contrary finding that the delivery accuracy is improved over time, and even more surprisingly, we disclose that with a sufficiently long evolving time, the delivery accuracy can achieve a perfect state where those who receive the data are exactly the ones that are interested in it. In addition, our theoretical findings are verified by experimental measurements through a social network dataset from Facebook.

The broad implementation of cloud computing have led to a dramatically growing in exchanging and using data throughout multiple parties. The main problem restricting the implementation of cloud computing is that users lack controls in cloud systems, from which security and privacy concerns become a major issue for cloud users. Logically, an applicable Fully Homomorphic Encryption (FHE) scheme is an effective solution to protecting data throughout the data usage lifecycle in the cloud system, due to the full control on users' own. However, there is no efficacious FHE scheme developed yet for meeting practical demands by the reason of either unqualified accuracy rate or intolerable latency time. Focus on this issue, we propose an advanced FHE scheme designed for operating real numbers, which is named as Full Homomorphic Encryption over Real Numbers (FHE-RN). Our approach has superb performances in both accuracy and efficiency, which has been proved by our experimental evaluations.

Sparse coding (SC) is a unsupervised learning scheme that has received an increasing amount of interests in recent years. However, conventional SC vectorizes the input images, which destructs the intrinsic spatial structures of the images. In this paper, we propose a novel graph regularized tensor sparse coding (GTSC) for image representation. GTSC preserves the local proximity of elementary structures in the image by adopting the newly proposed tubal-tensor representation. Simultaneously, it considers the intrinsic geometric properties by imposing graph regularization that has been successfully applied to uncover the geometric distribution for the image data. Moreover, the learned sparse representations by GTSC have better physical explanations as the key operation (i.e., circular convolution) in the tubal-tensor model preserves the shifting invariance property. Experimental results on image clustering demonstrate the effectiveness of the proposed scheme.

In many realistic networks, entities of different types usually form an evolving network with hybrid interactions. However, how to mathematically model such networks remains unexplored. Motivated by this, we develop a novel evolving model, which, as validated by our empirical results, can well capture some basic features such as power-law distribution, densification and shrinking diameter. Particularly, in our proposed model, named Evolving K-Graph, the hybrid interactions among entities are classified into inter-type and intra-type connections that are respectively characterized by two joint graphs evolving over time. By empirical validation, we disclose two new network properties: a positive correlation of any two layers of the network, and an earlier occurrence of network connectivity resulted by our model.

In this paper, we consider tensor completion under adaptive sampling of tensor (a multidimensional array) fibers. Tensor fibers or tubes are vectors obtained by fixing all but one index of the array. This sampling is in contrast to the cases considered so far where one performs an adaptive element-wise sampling. In this context we exploit a recently proposed algebraic framework to model tensor data [1] and model the underlying data as a tensor with low tensor tubal-rank. Under this model we then present an algorithm for adaptive sampling and recovery, which is shown to be nearly optimal in terms of sampling complexity. We apply this algorithm for robust estimation of RF fingerprints for accurate indoor localization. We show the performance on real and synthetic data sets. Compared to existing methods, that are primarily based on non-adaptive matrix completion methods, adaptive tensor completion achieves significantly better performance.

A well-designed energy-efficient routing protocol is an indispensable part for prolonging the lifetime of wireless sensor networks (WSNs) because a sensor node usually has limited energy. Many research efforts are contributed on routing design in WSNs. With the development of green technology, the energy harvesting technique is being applied to real WSNs. Therefore, existing routing protocols are not suitable for such new WSNs with energy harvesting. In this paper, we concentrate on designing a novel routing protocol, named energy harvesting routing (EHR), which takes energy harvesting as one major factor into routing design to improve the energy efficiency. First, we introduce a hybrid routing metric combining the effect of residual energy and energy harvesting rate. Then we propose an updating mechanism allowing every node to maintain dynamic energy information of its neighbors. Based on the hybrid metric and the neighbor information, EHR is able to locally select the optimal next hop. Extensive simulations are conducted to evaluate the performance of EHR. Results demonstrate that EHR outperforms existing routing protocols in energy harvesting WSNs in term of the energy efficiency.

In the wireless communication, there are many cases where the transmission path is obstructed by unknown objects. With the rapid development of the drone technology in recent years, the drones are advocated to serve as mobile relays to forward data streams. However, the challenges are that data transmission may suffer severe signal attenuation due to the existence of the obstructions and it is challenging to find the best location for mobile relays due to the dynamic environment and unpredictable interference. To address the problem, this paper proposes an approach that a drone can automatically find the location with the optimal link quality. We design a novel algorithm, named Path-sampling Online Tensor Update (POTU), to estimate the link quality in the space and find the optimal location. Furthermore, the algorithm is practical to the real applications due to the simplicity of implementation. In the experiment, we construct a realistic scene and compare the performance of our algorithm with the classic and the state-of-the-art algorithms. As a result, POTU outperforms existing methods in achieving the trade-off between time cost and estimation accuracy.

Public vehicle (PV) systems will be efficient traffic-management platforms in future smart cities, where PVs provide ridesharing trips with balanced QoS (quality of service). PV systems differ from traditional ridesharing due to that the paths and scheduling tasks are calculated by a server according to passengers' requests, and all PVs corporate with each other to achieve higher transportation efficiency. Path planning is the primary problem. The current path planning strategies become inefficient especially for traffic big data in cities of large population and urban area. To ensure real-time scheduling, we propose one efficient path planning strategy with balanced QoS (e.g., waiting time, detour) by restricting search area for each PV, so that a large number of computation is saved. Simulation results based on the Shanghai (China) urban road network show that, the computation can be reduced by 34% compared with the exhaustive search method since many requests violating QoS are excluded.

This paper advocates a new paradigm of transportation systems for future smart cities, namely, public vehicles (PVs), that provides dynamic ridesharing trips at requests. Passengers will enjoy more convenient and flexible transportation services with much less expense. In the PV system, both the number of vehicles and required parking spaces will be significantly reduced. There will be less traffic congestion, less energy consumption, and less pollution. In this paper, the concept, method, and algorithm for the PV system are described. The key issue of effectively implementing the PV system is to design efficient planning and scheduling algorithms. The PV-path problem is formulated, which is NP-complete. Then, a practical approach is proposed, which can serve people anywhere and anytime. The simulation results show that, to achieve the same performance (e.g., total time, waiting time, and travel time), the number of vehicles in the PV system can be reduced by around 90% and 57% compared with the conventional vehicle system and Uber Pool, respectively, and the total traveling distance can be reduced by 34% and 14%.

Indoor localization is a supporting technology for a broadening range of pervasive wireless applications. One promising approach is to locate users with radio frequency fingerprints. However, its wide adoption in real-world systems is challenged by the time- and manpower-consuming site survey process, which builds a fingerprint databasea priorifor localization. To address this problem, we visualize the 3-D RF fingerprint data as a function of locations (x-y) and indices of access points (fingerprint), as atensorand use tensor algebraic methods for anadaptivetubal-sampling of this fingerprint space. In particular, using a recently proposed tensor algebraic framework in[1], we capture the complexity of the fingerprint space as a low-dimensional tensor-column space. In this formulation, the proposed scheme exploits adaptivity to identify reference points which are highly informative for learning this low-dimensional space. Further, under certain incoherency conditions, we prove that the proposed scheme achieves bounded recovery error and near-optimal sampling complexity. In contrast to several existing work that rely on random sampling, this paper shows that adaptivity in sampling can lead to significant improvements in localization accuracy. The approach is validated on both data generated by the ray-tracing indoor model which accounts for the floor plan and the impact of walls and the real world data. Simulation results show that, while maintaining the same localization accuracy of existing approaches, the amount of samples can be cut down by$71$percent for the high SNR case and$55$percent for the low SNR case.

The propagation methods of viruses are diverse and studying the virus propagation is a hot topic. There appears a new way that the search engine quickly spreads network viruses, and many researches overlook its impact of propagation and few researches set a model to quantifiabely analyze how the search engine spread viruses. Based on community networks, this paper designs a specific model how the search engine spreads viruses. Moreover, this paper quantifiabely calculate the virus propagation velocity and the propagation effect. First, by analyzing the propagation process of viruses under the search engine condition, we design a positive feedback model to analyze how the search engine and the community network influence the propagation process of viruses. Second, we define relationship functions of propagation factors and calculate the rate of infected nodes while establishing mathematical propagation formulas for two situations. One situation is with the search engine, and another is without the search engine. Third, we design the experiment to verify the model analysis. Compared with two situations, we show that viruses have a much quicker propagation velocity, and the growth rate of infected rate is larger under the search engine condition. When the immune vaccine replaces the virus, this paper is also applicable.

Substantial technology advances have been made in areas of autonomous and connected vehicles, which opens a wide landscape for future transportation systems. We propose a new type of transportation system, Public Vehicle (PV) system, to provide effective, comfortable, and convenient service. The PV system is to improve the efficiency of current transportation systems, \eg, taxi system. Meanwhile, the design of such a system targets on significant reduction in energy consumption, traffic congestion, and provides solutions with affordable cost. The key issue of implementing an effective PV system is to design efficient scheduling algorithms. We formulate it as the PV Path (PVP) problem, and prove it is NP-Complete. Then we introduce a real time approach, which is based on solutions of the Traveling Salesman Problem (TSP) and it can serve people efficiently with lower costs. Our results show that to achieve the same performance (e.g., the total time: waiting and travel time), the number of vehicles can be reduced by 47%-69%, compared with taxis. The number of vehicles on roads is reduced, thus traffic congestion is relieved.

Location based services have experienced an explosive growth and evolved from utilizing a single location to the whole trajectory. Due to the hardware and energy constraints, there are usually many missing data within a trajectory. In order to accurately recover the complete trajectory, crowdsensing provides a promising method. This method resorts to the correlation among multiple users' trajectories and the advanced compressive sensing technique, which significantly outperforms conventional interpolation methods on accuracy. However, as trajectories exposes users' daily activities, the privacy issue is a major concern in crowdsensing. While existing solutions independently tackle the accurate trajectory recovery and privacy issues, yet no single design is able to address these two challenges simultaneously. Therefore in this paper, we propose a novel Privacy Preserving Compressive Sensing (PPCS) scheme, which encrypts a trajectory with several other trajectories while maintaining the homomorphic obfuscation property for compressive sensing. Under PPCS, adversaries can only capture the encrypted data, so the user privacy is preserved. Furthermore, the homomorphic obfuscation property guarantees that the recovery accuracy of PPCS is comparable to the state-of-the-art compressive sensing design. Based on two publicly available traces with numerous users and long durations, we conduct extensive simulations to evaluate PPCS. The results demonstrate that PPCS achieves a high accuracy of9,000 m even when up to 50% original data are missing.

Environment reconstruction is to rebuild the physical environment in the cyberspace using the sensory data collected by sensor networks, which is a fundamental method for human to understand the physical world in depth. A lot of basic scientific work such as nature discovery and organic evolution heavily relies on the environment reconstruction. However, gathering large amount of environmental data costs huge energy and storage space. The shortage of energy and storage resources has become a major problem in sensor networks for environment reconstruction applications. Motivated by exploiting the inherent feature of environmental data, in this paper, we design a novel data gathering protocol based on compressive sensing theory and time series analysis to further improve the resource efficiency. This protocol adapts the duty cycle and sensing probability of every sensor node according to the dynamic environment, which cannot only guarantee the reconstruction accuracy, but also save energy and storage resources. We implement the proposed protocol on a 51-node testbed and conduct the simulations based on three real datasets from Intel Indoor, GreenOrbs and Ocean Sense projects. Both the experiment and simulation performances demonstrate that our method significantly outperforms the conventional methods in terms of resource efficiency and reconstruction accuracy.

Data collection is a crucial operation in wireless sensor networks. The design of data collection schemes is challenging due to the limited energy supply and the hot spot problem. Leveraging empirical observations that sensory data possess strong spatiotemporal compressibility, this paper proposes a novel compressive data collection scheme for wireless sensor networks. We adopt a power-law decaying data model verified by real data sets and then propose a random projection-based estimation algorithm for this data model. Our scheme requires fewer compressed measurements, thus greatly reduces the energy consumption. It allows simple routing strategy without much computation and control overheads, which leads to strong robustness in practical applications. Analytically, we prove that it achieves the optimal estimation error bound. Evaluations on real data sets (from the GreenOrbs, IntelLab and NBDC-CTD projects) show that compared with existing approaches, this new scheme prolongs the network lifetime by 1.5X to 2X for estimation error 5-20 percent.

The data gathering is a fundamental operation in wireless sensor networks. Among approaches of the data gathering, the compressive data gathering (CDG) is an effective solution, which exploits the spatiotemporal correlation of raw sensory data. However, in the multi-attribute scenario, the performance of CDG decreases in every attribute's capacity because more measurements are on demand. In this paper, under the general framework of CDG, we propose a multi-attribute compressive data gathering protocol, taking into account the observed interattribute correlation in the multi-attribute scenario. Firstly, we find that 1) the rapid growth of the demand on measurements may decline the network capacity, 2) according to the compressive sensing theory, correlations among attributes can be utilized to reduce the demand on measurements without the loss of accuracy, and 3) such correlations can be found on real data sets. Secondly, motivated by these observations, we propose our approach to decline measurements. Finally, the real-trace simulation shows that our approach outperforms the original CDG under multiattribute scenario. Compared to the CDG, our approach can save 16% demand on measurements.

Electric vehicle (EV) is a promising transportation with plenty of advantages, e.g., low carbon emission, high energy efficiency. However, it requires frequent and long time charging. In public charging stations, EVs spend long time on queuing especially during peak hours. Hence, it requires an efficient method to reduce the total charging time for EVs. We study the Electric Vehicle Charging-Scheduling (EVCS) problem in this paper. First we prove that EVCS is NP-Complete, which can be reduced from one Parallel Machine Scheduling (PMS) problem. Then two heuristic algorithms are proposed: the Earliest Start Time (EST) algorithm, and the Earliest Finish Time (EFT) algorithm. EST tries to advance the start charging time to get customers in service as early as possible, while EFT focuses on the possible finish charging time to get customers served as soon as possible. Finally simulations show that, the proposed algorithms outperform the classic greedy nearest scheduling algorithm: assign each EV to its nearest charging station, then choose the outlet where the fewest EVs are queuing. Typically, under our simulation settings, the average finish time and maximum finish time can be reduced by about one hour, and six hours respectively.

Reconstructing the environment by sensory data is a fundamental operation for understanding the physical world in depth. A lot of basic scientific work (e.g., nature discovery, organic evolution) heavily relies on the accuracy of environment reconstruction. However, data loss in wireless sensor networks is common and has its special patterns due to noise, collision, unreliable link, and unexpected damage, which greatly reduces the reconstruction accuracy. Existing interpolation methods do not consider these patterns and thus fail to provide a satisfactory accuracy when the missing data rate becomes large. To address this problem, this paper proposes a novel approach based on compressive sensing to reconstruct the massive missing data. Firstly, we analyze the real sensory data from Intel Indoor, GreenOrbs, and Ocean Sense projects. They all exhibit the features of low-rank structure, spatial similarity, temporal stability and multi-attribute correlation. Motivated by these observations, we then develop an environmental space time improved compressive sensing (ESTI-CS) algorithm with a multi-attribute assistant (MAA) component for data reconstruction. Finally, extensive simulation results on real sensory datasets show that the proposed approach significantly outperforms existing solutions in terms of reconstruction accuracy.

Coverage is a fundamental problem in wireless sensor networks (WSNs). Conventional studies on this topic focus on 2D ideal plane coverage and 3D full space coverage. The 3D surface of a field of interest (FoI) is complex in many real-world applications. However, existing coverage studies do not produce practical results. In this paper, we propose a new coverage model called surface coverage. In surface coverage, the field of interest is a complex surface in 3D space and sensors can be deployed only on the surface. We show that existing 2D plane coverage is merely a special case of surface coverage. Simulations point out that existing sensor deployment schemes for a 2D plane cannot be directly applied to surface coverage cases. Thus, we target two problems assuming cases of surface coverage to be true. One, under stochastic deployment, what is the expected coverage ratio when a number of sensors are adopted? Two, if sensor deployment can be planned, what is the optimal deployment strategy with guaranteed full coverage with the least number of sensors? We show that the latter problem is NP-complete and propose three approximation algorithms. We further prove that these algorithms have a provable approximation ratio. We also conduct extensive simulations to evaluate the performance of the proposed algorithms.

In wireless sensor networks (WSNs), since many basic scientific works heavily rely on the complete sensory data, data recovery is an indispensable operation against the data loss. Several works have studied the missing value problem. However, existing solutions cannot achieve satisfactory accuracy due to special loss patterns and high loss rates in WSNs. In this work, we propose a multiple attributes-based recovery algorithm which can provide high accuracy. Firstly, based on two real datasets, the Intel Indoor project and the GreenOrbs project, we reveal that such correlations are strong, e.g., the change of temperature and light illumination usually has strong correlation. Secondly, motivated by this observation, we develop a Multi-Attribute-assistant Compressive-Sensing-based (MACS) algorithm to optimize the recovery accuracy. Finally, real trace-driven simulation is performed. The results show that MACS outperforms the existing solutions. Typically, MACS can recover all data with less than 5% error when the loss rate is less than 60%. Even when losing 85% data, all missing data can be estimated by MACS with less than 10% error.

Reconstructing the environment in cyber space by sensory data is a fundamental operation for understanding the physical world in depth. A lot of basic scientific work (e.g., nature discovery, organic evolution) heavily relies on the accuracy of environment reconstruction. However, data loss in wireless sensor networks is common and has its special patterns due to noise, collision, unreliable link, and unexpected damage, which greatly reduces the accuracy of reconstruction. Existing interpolation methods do not consider these patterns and thus fail to provide a satisfactory accuracy when missing data become large. To address this problem, this paper proposes a novel approach based on compressive sensing to reconstruct the massive missing data. Firstly, we analyze the real sensory data from Intel Indoor, GreenOrbs, and Ocean Sense projects. They all exhibit the features of spatial correlation, temporal stability and low-rank structure. Motivated by these observations, we then develop an environmental space time improved compressive sensing (ESTICS) algorithm to optimize the missing data estimation. Finally, the extensive experiments with real-world sensory data shows that the proposed approach significantly outperforms existing solutions in terms of reconstruction accuracy. Typically, ESTICS can successfully reconstruct the environment with less than 20% error in face of 90% missing data.

An urban vehicular network is a typical type of Delay Tolerant Network (DTN). Based on the routing analysis in a DTN, we first put forward a Minimum Delay and Hop Algorithm (MDHA), which requires both historical and future information on all the vehicles in the network. Since MDHA is not practical, we then design a Traffic Aware Routing Algorithm (TARA), which uses the historical and the real-time vehicle information to make routing decisions on the road structure level. A simulation using real GPS data in Shanghai shows that TARA significantly reduces the transmission delay and the hop count compared to the traditional GEO routing and GPSR.

Data loss is ubiquitous in wireless sensor networks (WSNs) mainly due to the unreliable wireless transmission, which results in incomplete sensory data sets. However, the completeness of a data set directly determines its availability and usefulness. Thus, sensory data recovery is an indispensable operation against the data loss problem. However, existing solutions cannot achieve satisfactory accuracy due to special loss patterns and high loss rates in WSNs. In this work, we propose a novel sensory data recovery algorithm which exploits the spatial and temporal joint-sparse feature. Firstly, by mining two real datasets, namely the Intel Indoor project and the GreenOrbs project, we find that: (1) for one attribute, sensory readings at nearby nodes exhibit inter-node correlation; (2) for two attributes, sensory readings at the same node exhibit inter-attribute correlation; (3) these inter-node and inter-attribute correlations can be modeled as the spatial and temporal joint-sparse features, respectively. Secondly, motivated by these observations, we propose two Joint-Sparse Sensory Data Recovery (JSSDR) algorithms to promote the recovery accuracy. Finally, real data-based simulations show that JSSDR outperforms existing solutions. Typically, when the loss rate is less than 65%, JSSDR can estimate missing values with less than 10% error. And when the loss rate reaches as high as 80%, the missing values can be estimated by JSSDR with less than 20% error.