Low Space Complexity Research Articles

Data-driven based automatic maritime routing from massive AIS trajectories in the face of disparity

Motivated by new technologies, such as Internet of Things, Big Data, Cloud Computing, data-driven based smart maritime related research is attracting more and more attention. With the establishment of China Automatic Identification System (AIS) Asia-Pacific Data Center and its network with IALA-Net, worldwide ship trajectory data is becoming increasingly available, thus providing plenty of materials for data-driven based smart maritime researches. This paper presents a novel approach to automatic maritime routing algorithm, given a set of ship trajectories, infer a routable road network by combining data-driven based algorithms, specially by focusing on: (i) After appropriate pre-processing, simplify AIS trajectory data using Douglas-Peucker algorithm, which can simplify AIS trajectory data by extracting characteristic points, in the meanwhile, compressing redundant information and improving subsequent processing efficiency. (ii) By setting similarity metric of characteristic points, cluster them using DBSCAN algorithm, consequently, deduce general behavior and spatial pattern of trajectories following the similar route. Also, the turning nodes of routes can be obtained in this step. Unlike past published works that focus on partitioned trajectory and require dense-sampling, trajectory point based method proposed in this paper has the merits of robustness to noise and disparity, low computational complexity and space complexity. (iii) Determine the connectivity of turning nodes according to ship trajectories. In this step, isolated turning nodes are linked together to form a coherent chain modeled as a directed graph. (iv) Given a starting location, infer the optimal route to the destination using Ant Colony Algorithm. By comparing the main route generated automatically and the macroscopic traffic flow situation, the result indicate that the method proposed in this paper is effective.

Concurrent use of write-once memory

We consider the problem of implementing general shared-memory objects on top of write-once bits, which can be changed from 0 to 1 but not back again. In a sequential setting, write-once memory (WOM) codes have been developed that allow simulating memory that support multiple writes, even of large values, setting an average of 1+o(1) write-once bits per write. We show that similar space efficiencies can be obtained in a concurrent setting, though at the cost of high time complexity and fixed bound on the number of write operations. As an alternative, we give an implementation that permits unboundedly many writes and has much better amortized time complexity, but at the cost of unbounded space complexity. Whether one can obtain both low time complexity and low space complexity in the same implementation remains open.

Scalable and Unified Digit-Serial Processor Array Architecture for Multiplication and Inversion Over GF( $2^{m}$ )

This paper proposes a scalable and unified digit-serial structure, with low space complexity to perform multiplication and inversion operations in $GF(2^{m})$ , based on the bit serial multiplication algorithm and the previously modified extended Euclidean inversion algorithm. In this structure, the multiplier and inverter shares the data-path and thus saves more area resources and power than the case of using separate data-path for each operation. Also, this structure is suitable for fixed size processor that only reuse the core and does not require to modulate the core size when the field size $m$ is modified. This structure is extracted by applying a nonlinear methodology that gives the designer more flexibility to control the processing element workload. Implementation results for of the proposed scalable and unified digit-serial design and previously reported efficient designs show that the proposed scalable structure achieves a significant reduction in area ranging from 64.3% to 85.5% and also achieves a significant saving in energy ranging from 21.9% to 92.5% over them, but it has lower throughput compared with them. This makes the proposed design more suitable for constrained implementations of cryptographic primitives in ultra-low power devices, such as wireless sensor nodes and radio frequency identification devices.

A heuristic task scheduling algorithm based on server power efficiency model in cloud environments

As the ever-growing energy consumption rises as a global concern, energy conservation in cloud data centers has become a topic of interest. Based on extensive research and experimental analysis of the energy consumption characteristics and performance records from SPEC data set, this paper presents a power efficiency model for cloud servers. With the model, we use server power efficiency to guide task scheduling and propose a heuristic task scheduling algorithm (ECOTS) for optimizing energy consumption in cloud environments. ECOTS takes into account multiple key factors such as task resource requirements, server power efficiency model and performance degradation in order to reduce system energy consumption at a minimal cost of performance. ECOTS algorithm has low time and space complexity with a better global searching ability to approach the optimal scheduling plan. We simulated a heterogeneous cluster environment and conducted experiments to evaluate the effectiveness of ECOTS. Simulation results show that ECOTS algorithm is the most energy-efficient without violating all the cloud tasks’ resources requirement. Moreover, ECOTS algorithm effectively reduces total energy consumption by more than 20% compared with the Improvement of Min-Min algorithm.

Spatial Pyramid Covariance-Based Compact Video Code for Robust Face Retrieval in TV-Series.

We address the problem of face video retrieval in TV-series, which searches video clips based on the presence of specific character, given one face track of his/her. This is tremendously challenging because on one hand, faces in TV-series are captured in largely uncontrolled conditions with complex appearance variations, and on the other hand, retrieval task typically needs efficient representation with low time and space complexity. To handle this problem, we propose a compact and discriminative representation for the huge body of video data, named compact video code (CVC). Our method first models the face track by its sample (i.e., frame) covariance matrix to capture the video data variations in a statistical manner. To incorporate discriminative information and obtain more compact video signature suitable for retrieval, the high-dimensional covariance representation is further encoded as a much lower dimensional binary vector, which finally yields the proposed CVC. Specifically, each bit of the code, i.e., each dimension of the binary vector, is produced via supervised learning in a max margin framework, which aims to make a balance between the discriminability and stability of the code. Besides, we further extend the descriptive granularity of covariance matrix from traditional pixel-level to more general patch-level, and proceed to propose a novel hierarchical video representation named spatial pyramid covariance along with a fast calculation method. Face retrieval experiments on two challenging TV-series video databases, i.e., the Big Bang Theory and Prison Break, demonstrate the competitiveness of the proposed CVC over the state-of-the-art retrieval methods. In addition, as a general video matching algorithm, CVC is also evaluated in traditional video face recognition task on a standard Internet database, i.e., YouTube Celebrities, showing its quite promising performance by using an extremely compact code with only 128 bits.

Fusion based Multimodal Biometric Security for Social Networks Communication

modern means of communication increase in their potential and receptiveness, they instigate additional demands in terms of security. Therefore, communication between users via social network becomes complicated, that increases possibility of threats with respect to user authentication. With the objective of ensuring security in social networks, different user authentication and cryptographic mechanism were designed. But, with different heuristic and computational algorithm, user authentication through single modal biometric is easily broken and vulnerabilities of multimodal approaches in social network still remain unexplored. This paper proposes a novel Fusion based Multimodal Biometric Security (FMBS) method utilizing Face and Fingerprint features of human individuals on social networks. Feature extraction for FMBS is performed utilizing combination of Binomial Feature Distribution Algorithm and Neighborhood Dominant Attribute Identification for both face and fingerprint features. Then, dominant attributes are stored in spatial vector for both the modalities to form biometric fusion template. Finally, Structural Biometric Fusion Template Matching algorithm designed to compute matching accuracy of test data to available training data. Experimental evaluation with Biometric Research Repositories is conducted. Performance evaluation show that the method significantly improve matching accuracy of human biometric samples, compared to conventional biometrics user authentication that only make use of single modal biometrics. The result shows that the method has low social network authentication time and network space complexity suited for deployment in real time social network sites.

Incremental semi-supervised kernel construction with self-organizing incremental neural network and application in intrusion detection

The semi-supervised learning (SSL) problems are often solved by graph based algorithms, semi-definite program- mings etc. These methods always require high space complexities, and thus are not efficient for network intrusion detection systems. Apart from the space complexity challenge, a network intrusion detection system should be able to handle the distribution drifting of data flow as well. A common solution for this concept drift problem is by SSL. In this paper, an incre- mental SSL training framework is proposed to combine the low space complexity advantage of topology learning and SSL for network intrusion detection. First, the unsupervised self-organizing incremental neural network is extended to process labeled and unlabeled information incrementally. Second, a kernel function is constructed from the training results of the previous step. Finally, a kernel machine is trained with the constructed kernel function. The proposed method reduces the space complexity of SSL to the magnitude similar to supervised learning. The experiments are carried out on the NSL-KDD datasets, and the results show that the proposed method outperforms the mainstream methods such as Transductive Support Vector Machine and Label Propagation.

HDIdx: High-dimensional indexing for efficient approximate nearest neighbor search

Fast Nearest Neighbor (NN) search is a fundamental challenge in large-scale data processing and analytics, particularly for analyzing multimedia contents which are often of high dimensionality. Instead of using exact NN search, extensive research efforts have been focusing on approximate NN search algorithms. In this work, we present “HDIdx”, an efficient high-dimensional indexing library for fast approximate NN search, which is open-source and written in Python. It offers a family of state-of-the-art algorithms that convert input high-dimensional vectors into compact binary codes, making them very efficient and scalable for NN search with very low space complexity.

A Study on Data Loss Compensation of WiFi-Based Wireless Sensor Networks for Structural Health Monitoring

In wireless data transmissions processes, data loss is an important factor that reduces the robustness of wireless sensor networks (WSNs). In many practical engineering applications, data loss compensation algorithms are then required to maintain the robustness, and such algorithms are typically based on compressive sensing with a large memory of microcontrollers needed. This paper presents an improved algorithm, based on random demodulator, to overcome the difficulty of microcontroller-dependence in the traditional data loss algorithms. The newly developed algorithm demonstrates the following advantages: 1) low space complexity; 2) low floating-point calculations; and 3) low time complexity. Therefore, it is more suitable to be embedded into ordinary nodes, comparing with the traditional algorithms. In this paper, a WSN based on WiFi is also developed for verifying the effectiveness and feasibility of the proposed algorithm. Field experiment on Xinghai Bay Bridge is done. Experimental results show that the WSN works properly and steady. Moreover, the data loss can be compensated effectively and efficiently through the use of the present algorithm.

Decoding schedule generating method for successive‐cancellation decoder of polar codes

Successive cancellation (SC) is the first and widely known decoder of polar codes, which has received a lot of attentions. However, its decoding schedule generating methods are still primitive. Based on a newly found factor, this study proposes an online method to generate the decoding schedule of SC decoder by a non-recursive way. The decoding schedule of SC decoder includes two parts. One is to determine which of likelihood ratios (LRs) can be shared and which of LRs need to be calculated. The other is to select a calculation formula for a LR to be computed. As shown by the comparisons among the proposed method and existing methods, the proposed method solves the first part of the decoding schedule with constant critical path delay and lower space complexity, and solves the second part of the decoding schedule with less calculations. Besides, experimental results show that the proposed method does not affect the error performance of polar codes.

CDNASA: clustering data with noise and arbitrary shape

In many data domains, especially for spatial data, clusters of data are of arbitrary shape, size and density. Traditional clustering methods often fail to identify clusters efficiently or accurately in those situations. But the need for scalable spatial clustering algorithms has emerged with the rapid growth of spatial data in recent years. In this paper we propose a spatial clustering method, named CDNASA, based on the idea that each data object belongs to a certain space and if the two spaces have overlapping sections, they can be merged into one cluster. The data points which cannot be merged by any cluster are noise points. The effectiveness and efficiency of the proposed algorithm are tested on both synthetic and real data sets. Experimental results show that the quality of clusters discovered by CDNASA is much better than those by existing algorithms, especially for arbitrary shaped clusters. CDNASA also has the characteristics of noise-tolerance as well as low time and space complexity.

Robust K-Median and K-Means Clustering Algorithms for Incomplete Data

Incomplete data with missing feature values are prevalent in clustering problems. Traditional clustering methods first estimate the missing values by imputation and then apply the classical clustering algorithms for complete data, such as K-median and K-means. However, in practice, it is often hard to obtain accurate estimation of the missing values, which deteriorates the performance of clustering. To enhance the robustness of clustering algorithms, this paper represents the missing values by interval data and introduces the concept of robust cluster objective function. A minimax robust optimization (RO) formulation is presented to provide clustering results, which are insensitive to estimation errors. To solve the proposed RO problem, we propose robust K-median and K-means clustering algorithms with low time and space complexity. Comparisons and analysis of experimental results on both artificially generated and real-world incomplete data sets validate the robustness and effectiveness of the proposed algorithms.

Efficient Scalable Digit-Serial Inverter Over GF( $2^{m}$ ) for Ultra-Low Power Devices

This paper proposes a novel scalable digit-serial inverter structure with low space complexity to perform inversion operation in GF(2 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m</sup> ) based on a previously modified extended Euclidean algorithm. This structure is suitable for fixed size processor that only reuse the core and does not require to modulate the core size when m modified. This structure is extracted by applying a nonlinear methodology that gives the designer more flexibility to control the processing element workload and also reduces the overhead of communication between processing elements. Implementation results of the proposed scalable design and previously reported efficient designs show that the proposed scalable structure achieves a significant reduction in the area ranging from 83.0% to 88.3% and also achieves a significant saving in energy ranging from 75.0% to 85.0% over them, but it has lower throughput compared to them. This makes the proposed design more suitable for constrained implementations of cryptographic primitives in ultra-low power devices such as wireless sensor nodes and radio frequency identification (RFID) devices.

Low space-complexity and low power semi-systolic multiplier architectures over GF(2m) based on irreducible trinomial

This paper proposes a three bit-serial and digit-serial semi-systolic GF(2m) multipliers using Progressive Product Reduction (PPR) technique. These architectures are obtained by converting the GF(2m) multiplication algorithm into an iterative algorithm using systematic techniques for scheduling the computational tasks and mapping them to Processing Elements (PE). Three different semi systolic arrays were obtained. ASIC implementation of the proposed designs and previously published schemes were used to verify the performance of the proposed designs. One proposed design has at least 29% lower area compared to previously published bit/digit serial multipliers. This design has also at least 70% lower power compared to previously published bit/digit serial multipliers. Another proposed design has at least 12% lower power-delay product (PDP) compared to previously published bit/digit serial multipliers. This makes the proposed designs more suited to resource-constrained embedded applications.

Look-back-upon tree recurrence method for Geiger-mode avalanche photodiode performance prediction.

This Letter, for the first time, proposes the look-back-upon tree recurrence (LTR) method based on the Poisson statistics for discrete time to predict the performance of Geiger-mode avalanche photodiodes with a short dead time. The results of the proposed method correspond to the results of previous methods for diverse input fluxes. The LTR method possesses very low time and space complexity to allow for the real-time analysis of the Geiger-mode avalanche photodiodes' performance over the entire ranges of dead time and input diversity.

Loda: Lightweight on-line detector of anomalies

In supervised learning it has been shown that a collection of weak classifiers can result in a strong classifier with error rates similar to those of more sophisticated methods. In unsupervised learning, namely in anomaly detection such a paradigm has not yet been demonstrated despite the fact that many methods have been devised as counterparts to supervised binary classifiers. This work partially fills the gap by showing that an ensemble of very weak detectors can lead to a strong anomaly detector with a performance equal to or better than state of the art methods. The simplicity of the proposed ensemble system (to be called Loda) is particularly useful in domains where a large number of samples need to be processed in real-time or in domains where the data stream is subject to concept drift and the detector needs to be updated on-line. Besides being fast and accurate, Loda is also able to operate and update itself on data with missing variables. Loda is thus practical in domains with sensor outages. Moreover, Loda can identify features in which the scrutinized sample deviates from the majority. This capability is useful when the goal is to find out what has caused the anomaly. It should be noted that none of these favorable properties increase Loda's low time and space complexity. We compare Loda to several state of the art anomaly detectors in two settings: batch training and on-line training on data streams. The results on 36 datasets from UCI repository illustrate the strengths of the proposed system, but also provide more insight into the more general questions regarding batch-vs-on-line anomaly detection.

Learning multiple linear mappings for efficient single image super-resolution.

Example learning-based superresolution (SR) algorithms show promise for restoring a high-resolution (HR) image from a single low-resolution (LR) input. The most popular approaches, however, are either time- or space-intensive, which limits their practical applications in many resource-limited settings. In this paper, we propose a novel computationally efficient single image SR method that learns multiple linear mappings (MLM) to directly transform LR feature subspaces into HR subspaces. In particular, we first partition the large nonlinear feature space of LR images into a cluster of linear subspaces. Multiple LR subdictionaries are then learned, followed by inferring the corresponding HR subdictionaries based on the assumption that the LR-HR features share the same representation coefficients. We establish MLM from the input LR features to the desired HR outputs in order to achieve fast yet stable SR recovery. Furthermore, in order to suppress displeasing artifacts generated by the MLM-based method, we apply a fast nonlocal means algorithm to construct a simple yet effective similarity-based regularization term for SR enhancement. Experimental results indicate that our approach is both quantitatively and qualitatively superior to other application-oriented SR methods, while maintaining relatively low time and space complexity.

Low-Complexity Digit-Serial and Scalable SPB/GPB Multipliers Over Large Binary Extension Fields Using (b,2)-Way Karatsuba Decomposition

Shifted polynomial basis (SPB) and generalized polynomial basis (GPB) are two variations of polynomial basis representation. SPB/GPB have potential for efficient bit-level and digit-level implementations of multiplication over binary extension fields. This paper presents a (b,2)-way KA decomposition for digit-serial multiplication with low-space complexity. Based on the proposed parallel (b,2)-way KA scheme, we derive a novel scalable SPB/GPB multiplier. Analytical results show that the proposed multiplier could achieve the desired trade-off between space and time complexities. Our proposed multiplier is modular, regular, and suitable for very-large-scale integration (VLSI) implementations. It involves significantly less area complexity, less computation time and less energy consumption compared to the existing digit-serial and scalable multipliers.

Type-2 fuzzy multi-intersection traffic signal control with differential evolution optimization

This paper proposes a multi-agent type-2 fuzzy logic control (FLC) method optimized by differential evolution (DE) for multi-intersection traffic signal control. Type-2 fuzzy sets can deal with models’ uncertainties efficiently because of its three-dimensional membership functions, but selecting suitable parameters of membership functions and rule base is not easy. DE is adopted to decide the parameters in the type-2 fuzzy system, as it is easy to understand, simple to implement and possesses low space complexity. In order to avoid the computational complexity, the expert rule base and the parameters of membership functions (MF) are optimized by turns. An eleven-intersection traffic network is studied in which each intersection is governed by the proposed controller. A secondary layer controller is set in every intersection to select the proper phase sequence. Furthermore, the communication among the adjacent intersections is implemented using multi-agent system. Simulation experiments are designed to compare communicative type-2 FLC optimized by DE with type-1 FLC, fixed-time signal control, etc. Experimental results indicate that our proposed method can enhance the vehicular throughput rate and reduce delay, queue length and parking rate efficiently.

Fast Clustering-Based Anonymization Algorithm for Data Streams

PDF HTML阅读 XML下载 导出引用 引用提醒 基于聚类的快速数据流匿名方法 DOI: 10.3724/SP.J.1001.2013.04330 作者: 作者单位: 作者简介: 通讯作者: 中图分类号: 基金项目: 国家自然科学基金(70871024); 福建省自然科学基金(2010J01358); 福州大学科技发展基金(201-xy-16) Fast Clustering-Based Anonymization Algorithm for Data Streams Author: Affiliation: Fund Project: 摘要 | 图/表 | 访问统计 | 参考文献 | 相似文献 | 引证文献 | 资源附件 | 文章评论 摘要:为了防止敏感信息的泄漏,保护用户隐私,常采用概化和抑制等技术在共享数据前对其准标识符进行匿名化.与静态数据集不同,数据流具有潜在无限、高度动态等特性,使得数据流匿名需要解决更加复杂的问题,不能直接应用静态数据集的匿名方法.在分析现有数据流匿名方法的基础上,提出一种采用聚类思想进行数据流匿名的方法,通过单遍扫描数据识别和重用满足匿名条件的簇,以实现数据流的快速匿名.真实数据集上的实验结果表明,该方法在满足匿名要求的同时能够降低概化和抑制处理带来的信息损失,并且具有较低的时间和空间复杂度. Abstract:In order to prevent the disclosure of sensitive information and protect users’ privacy, the generalization and suppression of technology is often used to anonymize the quasi-identifiers of the data before its sharing. Data streams are inherently infinite and highly dynamic which are very different from static datasets, so that the anonymization of data streams needs to be capable of solving more complicated problems. The methods for anonymizing static datasets cannot be applied to data streams directly. In this paper, an anonymization approach for data streams is proposed with the analysis of the published anonymization methods for data streams. This approach scans the data only once to recognize and reuse the clusters that satisfy the anonymization requirements for speeding up the anonymization process. Experimental results on the real dataset show that the proposed method can reduce the information loss that is caused by generalization and suppression and also satisfies the anonymization requirements and has low time and space complexity. 参考文献 相似文献 引证文献