ISSN 0253-2778

CN 34-1054/N

2016 Vol. 46, No. 1

Display Method:
Original Paper
Research of defected ground structure microstrip low-pass filter using stepped impedance shunt stubs structure
NI Chun, ZHANG Liang, WU Xianliang, ZHENG Juan
2016, 46(1): 1-5. doi: 10.3969/j.issn.0253-2778.2016.01.001
Defected ground structure (DGS) can change the effective permittivity of substrate materials and the equivalent circuit of the microstrip transmission line by etching the periodic or nonperiodic defected pattern on the ground plane of the microwave circuit. The traditional DGS microstrip low-pass filter has a narrow stopband, and the stopband rejection is poor. to solve this problem, the circuit structure of DGS is studied deeply . Two DGS microstrip low-pass filter circuits based on SISS structure were designed, by the adoption of rectangular defected ground structure (R-DGS), and the introduction of the rectangular stepped impedance shunt stubs (R-SISS) and semicircle stepped impedance shunt stub(S-SISS). The simulation results show that the novel DGS circuit techniques can effectively improve the RF transmission character in passband, broaden stopband bandwidth, increase stopband rejection, which are satisfactory.
A simplified non-maximum suppression with improved constraints
ZHANG Qiang, ZHANG Chenbin, CHEN Zonghai
2016, 46(1): 6-11. doi: 10.3969/j.issn.0253-2778.2016.01.002
Post processing plays an important role in object detection methods based on the sliding window method. Simplified non-maximum suppression is a typical representative of post processing methods. However, traditional simplified non-maximum suppression uses only one constraint and cannot discard repetitive detections effectively. An improved simplified non-maximum suppression with two additional constraints was proposed. Compared with the traditional simplified non-maximum suppression which only calculates the proportion of intersection area to that of candidate detection bounding box, the two additional constraints named "completely covered detection suppression" and "PASCAL VOC overlap criterion" calculate the proportions of the intersection area to that of the selected detection bounding box and to the union area, respectively. The experimental results show that the improved simplified non-maximum suppression could discard the false positives effectively and significantly improve detection performance.
Bidirectional RRT algorithm based grasping manipulation of humanoid robots
DU Shuang, SHANG Weiwei, LIU Kun, WANG Zhiling
2016, 46(1): 12-20. doi: 10.3969/j.issn.0253-2778.2016.01.003
To realize grasping manipulation effectively in practical application, whole-body motion planning should be designed for humanoid robots. Thus, degrees of freedom of all the joints in humanoid robots, and constraints of robots, environment and the physical characteristics of grasped objects should be taken into consideration. To solve the problems including multi degrees of freedom and complex constraints, a new planning method is designed by using bidirectional RRT algorithm. After receiving stable double-leg configurations and the list of grasping hand’s poses, the bidirectional RRT algorithm is adopted to realize the whole-body motion planning for humanoid robots. Some experiments are conducted to make a NAO humanoid robot to open a drawer, enables to open a drawer in the presence of obstacles, and open a drawer for taking an object, and close the drawer. The results indicate that the whole-body motion planning with bidirectional RRT algorithm is effective in achieving the grasping manipulation of humanoid robots.
A research on control-flow taint information directed symbolic execution
HUANG Hui, LU Yuliang, LIU Lintao, ZHAO Jun
2016, 46(1): 21-27. doi: 10.3969/j.issn.0253-2778.2016.01.004
Aiming at generation of test cases covering the potential vulnerable program points and combining generation base Fuzzing, static control flow analysis and static taint analysis, this paper proposes a directed dynamic symbolic execution method. By Fuzzing the test cases which could reach the function containing the vulnerable program points are generated, leading the symbolic execution fast towards the vulnerable functions along the denoted single path; By making a static control-flow analysis and a static taint analyses in the vulnerable functions, the control flow taint eachable slices are calculated directing the multi-path dynamic symbolic execution towards the desired vulnerable program points. Experiments prove effectiveness of the method in mitigating the path explosion problem common in symbolic execution applications and in generating test cases that trigger target vulnerability.
Overlapping influence of multiple spreaders in complex networks
ZHOU Mingyang, FU Zhong qian, LIAO Hao
2016, 46(1): 28-35. doi: 10.3969/j.issn.0253-2778.2016.01.005
With the development of computer technology and the Internet, network science is attracting many scientists from various fields. One field in network science is epidemic spreading, in which the key problem is the selection of source spreaders. Conventional methods select spreaders according to the importance of nodes (degree, betweenness and so on) and nodes with high importance are selected. Traditional methods perform well in characterizing the spreading ability of single nodes, but poorly in multiple nodes. An anahysis is made and the reasons poor performance of multiple spreaders is attributed to the overlapping influences that decrease the overall spreading ability of multiple nodes. Then, an improved method is proposed to suppress the overlapping influences. The validity of the proposed method is illustrated in four real-world networks in which the method could select better multiple spreaders. Further, it was found that improving the sparsity could reduce the overlapping influence of multiple spreaders, which enhances the overall spreading ability of nodes.
MCDS: Large-scale mobile communication data computation on just a PC
LIU Zhipeng
2016, 46(1): 36-46. doi: 10.3969/j.issn.0253-2778.2016.01.006
Mobile data has the characteristics of high volume, variety, velocity and value. Mobile communication data is an important part of mobile data, and it has great research value. It is of tremendous significance to efficiently store and retrieve mobile data. At present, utilizing parallel technology to perform data mining has become the main stream, but the technology is very costly in terms of hardware, and code debugging and optimization of parallel algorithms is difficult. A mobile communication data processing system operational on a single PC was proposed. MCDS is based on GraphChi, and improves GraphChi from 3 aspects: data format, sharding mechanism and memory replacement algorithm. Experimental results verify the effectiveness of MCDS, and it provides a feasible experimental environment for mobile communication data mining.
Instant traveling companion discovery based on large scale streaming ANPR data
ZHU Meiling, WANG Xiongbin, ZHANG Shouli, LIU Chen, HAN Yanbo
2016, 46(1): 47-55. doi: 10.3969/j.issn.0253-2778.2016.01.007
Traveling companions are object groups that move together in a period of time. To quickly identify traveling companions from a special kind of streaming traffic data, called automatic number plate recognition (ANPR) data, a framework and several algorithms were presented to discover companion vehicles, which can instantly detect suspicious companion vehicles with their probabilities when they pass through monitoring cameras.The framework can be used in many time-sensitive scenarios like taking surveillance on suspect trackers for specific vehicles. Experiments show that the proposed approach can process streaming ANPR data directly and discover companion vehicles in nearly real time.
Tabular-oriented data model and its query issues
HUANG Dongmei, SUN Le, SHI Shaohua, SU Cheng, ZHAO Danfeng
2016, 46(1): 56-65. doi: 10.3969/j.issn.0253-2778.2016.01.008
With the rapid development of information technologies, data storage and representation of various sources, including not only the traditional structured data such as relational databases and object-oriented databases, but also those special unstructured data like Excel, CSV documents, manifest distributed and heterogeneous characteristics. Undoubtedly, all above data features high-volume, continuously-updating, low-usability, which falls into Big Data. However, the organization and management of Excel and other forms of data by using unstructured and semi-structured methods leads to a weakly-controllable, weakly-usable data structure with poor access efficiency. To solve this problem, this paper, taking Excel data source into consideration, aims to propose a new tabular-oriented relational data model and discusses Tabular querying and optimizing issues. Firstly, the formal definition of Tabular form data is given; secondly, PartiPath tree is designed to achieve structural transformation by tabular division and its relation schema as well; then its data model is presented. After that, four basic queries and their optimization by improved DICE with user interest similarity are described. Finally, the experiment was conducted and a conclusion was drawm.
Spark/Shark-based OLAP system for smart grid applications
WANG Yaling, LIU Yue, HONG Jianguang, CUI Wei, LI Yanhu, SU Yipeng, HUANG Gaopan, ZHANG Mingming, LIU Wantao
2016, 46(1): 66-75. doi: 10.3969/j.issn.0253-2778.2016.01.009
The OLAP queries on electricity consumption information in Smart Grid have some prominent features: huge amounts of data, involving multiple tables in a joint operation, complex SQL structure, etc. Faced with this kind of applications, traditional RDBMS always leads to poor scalability, low write throughput, and unacceptable query performance, etc. A Spark/Shark-Based OLAP system for electricity consumption information in smart grid was designed. The system used distributed file system HDFS for data storage, and makes use of Shark to parse the SQL queries and Spark to execute them. However, Shark does not support fine-grained index, which hinders further improvement of query performance. To overcome this limitation, a Trie tree based fine-grained index technique TrieIndex and data re-organization scheme for better query performance was proposed. The experiment results with real electricity consumption information data and query show that the write throughput of the system is 12 times faster than that of RDBMS, and the query efficiency of the system is 10 times greater than that of original Shark.
Research on cognition-based hierarchical model for the construction and analysis of scenarios in chance discovery
CHENG Hongmei, ZHANG Zhenya
2016, 46(1): 76-81. doi: 10.3969/j.issn.0253-2778.2016.01.010
To identify and manage chance events effectively, a cognition-based hierarchical model for the construction and analysis of scenarios in chance discovery is presented according to cognitive information processing theory with the cognitive situation model, the process characteristic of cognition and the information filter mechanism of attention in cognition as references. The new model is composed of five information processes from bottom to top such as acquisition of private views, construction of private scenarios, integration of scenarios, generalization of scenarios and scenario analysis. Problems such as the acquisition of event clusters based on cognition, the representation and evolution of attention oriented to the implementation of filter mechanisms of attention, the construction of event clusters in chance diacovery scenarios, the aggregation of chance discovery scenarios, the implementation of the association phenomenon in the construction and analysis of chance discovery scenarios are discussed in detail. If cluster partition is treated as the chance discovery scenario and the chance discovery scenario is constructed as the aggregation of some private chance discovery scenarios where one private chance discovery scenario is one kind of cluster partition on dataset. Exmperimental results show that the accuracy of the cluster partition can be improved significantly.
Research on collaborative recommendation algorithms based on parallel spectral clustering
ZHENG Xiumeng, CHEN Fucai, HUANG Ruiyang
2016, 46(1): 82-86. doi: 10.3969/j.issn.0253-2778.2016.01.011
With the increase of large-scale network data, scalability has become a key factor in the recommendation system. A new collaborative recommendation algorithm is thus based on MapReduce parallel spectral clustering was proposed. First, items are clustered using the improved parallel spectral clustering method; Then, based on the user collaborative recommendation algorithm and combined with the clustered items’ ratings, an improved calculation method for similar users is proposed to establish recommendation. The test results on the dataset show that the proposed algorithm can effectively reduce time complexity, which significantly improving its accuracy and efficiency.