ISSN 
2007 Vol. 36, No. 6
2007, 36(6): 1147-1151.
Abstract:
For improving the efficiency and the potential for actively protecting against attacks, a novel security-ensured public key cryptosystem is proposed. The idea is that a large set of published parameters, which are generated with almost no more overheads, is taken into account in addition to the original public and private keys. A new set of published parameters will be chosen when a group of data to be encrypted each time. The linear time quick algorithms for deciding the solution structure and computing the solution of the quartic congruential equations are proposed for implementing the operations on the published parameters. This cryptosystem can actively protect against the chosen plaintext and the various chosen ciphertext attacks including IND-CCA2 attacks. It achieves the same security like One-Time-Pad. For the repeated encryption/decryption for a set of data the requirements for computation and memory size are deeply decreased because only a series of XOR operations are needed after first data item has been encrypted/decrypted. Due to this reason, this cryptosystem may be used in very low-end devices, such as RFID tags, sensor networks, where the computation, memory and bandwidth are very limited.
For improving the efficiency and the potential for actively protecting against attacks, a novel security-ensured public key cryptosystem is proposed. The idea is that a large set of published parameters, which are generated with almost no more overheads, is taken into account in addition to the original public and private keys. A new set of published parameters will be chosen when a group of data to be encrypted each time. The linear time quick algorithms for deciding the solution structure and computing the solution of the quartic congruential equations are proposed for implementing the operations on the published parameters. This cryptosystem can actively protect against the chosen plaintext and the various chosen ciphertext attacks including IND-CCA2 attacks. It achieves the same security like One-Time-Pad. For the repeated encryption/decryption for a set of data the requirements for computation and memory size are deeply decreased because only a series of XOR operations are needed after first data item has been encrypted/decrypted. Due to this reason, this cryptosystem may be used in very low-end devices, such as RFID tags, sensor networks, where the computation, memory and bandwidth are very limited.
2007, 36(6): 1152-1154,1159.
Abstract:
In this paper, we design an improved MFE multivariate public key encryption scheme, which can resist the HOLE attack. There are other two attacks on MPKCs, the Rank attack and XL the & Gröbner basis attack. We provide the computational complexity under the Rank attack and XL&Gröbner basis attack. Through analysis, if we choose proper values of parameters, this improved scheme can resist both the Rand attack and the XL&Gröbner basis attack. Thus, this improved scheme is a secure multivariate public key encryption scheme.
In this paper, we design an improved MFE multivariate public key encryption scheme, which can resist the HOLE attack. There are other two attacks on MPKCs, the Rank attack and XL the & Gröbner basis attack. We provide the computational complexity under the Rank attack and XL&Gröbner basis attack. Through analysis, if we choose proper values of parameters, this improved scheme can resist both the Rand attack and the XL&Gröbner basis attack. Thus, this improved scheme is a secure multivariate public key encryption scheme.
2007, 36(6): 1155-1159.
Abstract:
Ad hoc Network is a collection of wireless nodes forming a temporary computer communication network without the aid of any established infrastructure or centralized administration. The design of key management schemes and protocols, whose main objective is to provide secure and reliable communication, is one of the most important aspects and basic research filed of secure ad hoc networks. The key management in ad hoc network meets many new challenges due to its intrinsic properties such as multi-hip communication, limited resource and so on. In this paper, a method to improve the random key pre-distribution scheme based on cluster scheme is proposed.The comparison to random key pre-distribution scheme shows that the new approach can substantially improve the connectivity of network and reduce the amount of memory required at the same time.
Ad hoc Network is a collection of wireless nodes forming a temporary computer communication network without the aid of any established infrastructure or centralized administration. The design of key management schemes and protocols, whose main objective is to provide secure and reliable communication, is one of the most important aspects and basic research filed of secure ad hoc networks. The key management in ad hoc network meets many new challenges due to its intrinsic properties such as multi-hip communication, limited resource and so on. In this paper, a method to improve the random key pre-distribution scheme based on cluster scheme is proposed.The comparison to random key pre-distribution scheme shows that the new approach can substantially improve the connectivity of network and reduce the amount of memory required at the same time.
2007, 36(6): 1160-1163.
Abstract:
To cut down the dependent of key management scheme on resource of networks, this paper deeply researches polynomial-based key predistribution in distribution sensor networks for reference and presents an efficient key predistribution scheme for Ad hoc networks, a threshold-based key predistribution scheme. The analysis in this paper indicates that this scheme has a number of good properties, such as high probability to establish pairwise keys, tolerance of node captures, and low communication overhead.
To cut down the dependent of key management scheme on resource of networks, this paper deeply researches polynomial-based key predistribution in distribution sensor networks for reference and presents an efficient key predistribution scheme for Ad hoc networks, a threshold-based key predistribution scheme. The analysis in this paper indicates that this scheme has a number of good properties, such as high probability to establish pairwise keys, tolerance of node captures, and low communication overhead.
2007, 36(6): 1164-1166,1171.
Abstract:
Key management schemes based on threshold mechanism can provide high security, but lower certificate success rate and bad scalability; key management schemes based on certification chain satisfy self-organized features, but their security depends on trust degree of nodes and could not be applied high security environment. We advance hybrid key management scheme based on threshold mechanism and certification chain. Simulation shows that with less increment of communication overhead, our scheme can improve both certificate success rate and system security and better balances security and availability, and satisfies security needs of Ad hoc networks.
Key management schemes based on threshold mechanism can provide high security, but lower certificate success rate and bad scalability; key management schemes based on certification chain satisfy self-organized features, but their security depends on trust degree of nodes and could not be applied high security environment. We advance hybrid key management scheme based on threshold mechanism and certification chain. Simulation shows that with less increment of communication overhead, our scheme can improve both certificate success rate and system security and better balances security and availability, and satisfies security needs of Ad hoc networks.
2007, 36(6): 1167-1171.
Abstract:
Partially blind signatures allow signers to include some common information (the date of issue, the value of electronic coins) negotiated by signers and users. This paper presents an efficient partially blind signatures scheme with provable security in the random oracle model. Our scheme uses a public algorithm to evolve the public keys and private keys which will be used during the partially blind signatures protocol. In such a manner the scheme has the same security as the underlying blind Okamoto-Schnorr signature without bringing out additional workload.
Partially blind signatures allow signers to include some common information (the date of issue, the value of electronic coins) negotiated by signers and users. This paper presents an efficient partially blind signatures scheme with provable security in the random oracle model. Our scheme uses a public algorithm to evolve the public keys and private keys which will be used during the partially blind signatures protocol. In such a manner the scheme has the same security as the underlying blind Okamoto-Schnorr signature without bringing out additional workload.
2007, 36(6): 1172-1178,1186.
Abstract:
This article has a deep analysis on the potential danger of the Exclusive OR (XOR) and Hash chain based light weight Radio Frequency Edentification (RFID) security protocol brought by some articles. On the base of analysis, combining the traditional security schemes widely used in Internet or existing application systems and some light weight and resource-saving security mechanism, we enhance the lightweight RFID security protocol mentioning above and give a new light weight RFID security protocol. Special attention is focused on the deployment considerations for applying the protocol in real application to avoid the identified potential risk. Full consideration is given to the low cost demands for RFID tags, the power restrictions, and the non-resistant against physical tampering feature, which are common limitations in the RFID system.
This article has a deep analysis on the potential danger of the Exclusive OR (XOR) and Hash chain based light weight Radio Frequency Edentification (RFID) security protocol brought by some articles. On the base of analysis, combining the traditional security schemes widely used in Internet or existing application systems and some light weight and resource-saving security mechanism, we enhance the lightweight RFID security protocol mentioning above and give a new light weight RFID security protocol. Special attention is focused on the deployment considerations for applying the protocol in real application to avoid the identified potential risk. Full consideration is given to the low cost demands for RFID tags, the power restrictions, and the non-resistant against physical tampering feature, which are common limitations in the RFID system.
2007, 36(6): 1179-1182.
Abstract:
To improve the network environment adaptation ability of intrusion detection, a distributed intrusion detection model based on multi-level immune detector set is presented. The proposed model consists of detection hosts, detection sub-networks and central detection server. Intrusion Detection System (IDS) is deployed in detection hosts. Detection sub-networks have the same features of their superior intrusion detection network. Central detection server provides supports to detect hosts and sub-networks. Through simulating immune cell in biological immune system, immune detectors in detection host learn and evolve. Proposed model utilizes second-level detector set mechanism and cooperates with central detection server operation of vaccines reception and vaccination to decrease the size of detectors and improve the performance of detection.
To improve the network environment adaptation ability of intrusion detection, a distributed intrusion detection model based on multi-level immune detector set is presented. The proposed model consists of detection hosts, detection sub-networks and central detection server. Intrusion Detection System (IDS) is deployed in detection hosts. Detection sub-networks have the same features of their superior intrusion detection network. Central detection server provides supports to detect hosts and sub-networks. Through simulating immune cell in biological immune system, immune detectors in detection host learn and evolve. Proposed model utilizes second-level detector set mechanism and cooperates with central detection server operation of vaccines reception and vaccination to decrease the size of detectors and improve the performance of detection.
2007, 36(6): 1183-1186.
Abstract:
This paper describes the principle of Distributed Denial of Service (DDoS) attack. Several representative defense methods are analyzed to against it. A defense method against IP spoofing DDoS attack is proposed. An active IP record table is used to detect all IP packets passing through the border of autonomy system in this method. Packets of the source IP address which are not active will be discarded by the border routers or routers near the border in the autonomy system, according to the Internet Control Message Protocol (ICMP) protocol, timeout ICMP messages will be sent to the source IP hosts, and thus, IP spoofed packets will be discarded, because their source IP usually are not active. Although some legal packets will also be discarded, the retransmission will be triggered by the timeout ICMP messages immediately.
This paper describes the principle of Distributed Denial of Service (DDoS) attack. Several representative defense methods are analyzed to against it. A defense method against IP spoofing DDoS attack is proposed. An active IP record table is used to detect all IP packets passing through the border of autonomy system in this method. Packets of the source IP address which are not active will be discarded by the border routers or routers near the border in the autonomy system, according to the Internet Control Message Protocol (ICMP) protocol, timeout ICMP messages will be sent to the source IP hosts, and thus, IP spoofed packets will be discarded, because their source IP usually are not active. Although some legal packets will also be discarded, the retransmission will be triggered by the timeout ICMP messages immediately.
2007, 36(6): 1187-1191.
Abstract:
A general model of buffer overflow based attacks is described by unified modeling language. The analysis and comparison of the existing representative methods and apparatuses of defense and recovery against buffer overflow attacks are presented, including analyzing their vulnerabilities and possible means to bypass them. Highlighting the state-of-art challenging issues for facing the tradeoff of security and performance efficiency, and the continuing evolution of attach techniques, it is pointed out that security programming is the key to solve buffer overflow problems. Finally, some technical trends are given.
A general model of buffer overflow based attacks is described by unified modeling language. The analysis and comparison of the existing representative methods and apparatuses of defense and recovery against buffer overflow attacks are presented, including analyzing their vulnerabilities and possible means to bypass them. Highlighting the state-of-art challenging issues for facing the tradeoff of security and performance efficiency, and the continuing evolution of attach techniques, it is pointed out that security programming is the key to solve buffer overflow problems. Finally, some technical trends are given.
2007, 36(6): 1192-1194.
Abstract:
To improve the performance of Minimax Probability Machine (MPM) in the detection rate and the training time, Intrusion Detection Based on Kernel Fisher Discriminant Analysis and Minimax Probability Machine Classifier (KFDA-MPM) algorithm is proposed which combines the feature extraction technology and classification algorithm. In this method, the KFDA is used to extract the optimal feature set and then the MPM is adopted to classify the optimization data. Results of the experiment using the Knowledge Discovery and Data Mining Cup 1999 (KDDCUP99) datasets indicate the effectiveness of the algorithm.
To improve the performance of Minimax Probability Machine (MPM) in the detection rate and the training time, Intrusion Detection Based on Kernel Fisher Discriminant Analysis and Minimax Probability Machine Classifier (KFDA-MPM) algorithm is proposed which combines the feature extraction technology and classification algorithm. In this method, the KFDA is used to extract the optimal feature set and then the MPM is adopted to classify the optimization data. Results of the experiment using the Knowledge Discovery and Data Mining Cup 1999 (KDDCUP99) datasets indicate the effectiveness of the algorithm.
2007, 36(6): 1195-1197,1252.
Abstract:
Aiming to solve the problem of net safety evaluation index, the data extracted from NetFlow are analyzed. And the data are aggregated by several relational dimensions. Using the method of grey relation advantage analysis, the grey relation values of network safety affairs and the characters of the data are calculated. The pivot characters representing the affairs were decided. Then a set of network safety evaluation indices are proposed. The experiment results show that the indices can reflect the network situation efficiently.
Aiming to solve the problem of net safety evaluation index, the data extracted from NetFlow are analyzed. And the data are aggregated by several relational dimensions. Using the method of grey relation advantage analysis, the grey relation values of network safety affairs and the characters of the data are calculated. The pivot characters representing the affairs were decided. Then a set of network safety evaluation indices are proposed. The experiment results show that the indices can reflect the network situation efficiently.
2007, 36(6): 1198-1201.
Abstract:
To effectively evaluate the network risk of network information system, a Danger Theory based Network Risk Evaluation Model (DTREM) is proposed. With definitions of self, non-self, and immunocyte, the intrusion detection sub-model is given. DTERM is composed of memory detectors, mature detectors, and immature detectors. Furthermore, the danger theory based network risk evaluation sub-model is given. In the proposed model, the risk of each network attack, including holistic risk of the host and network, can be calculated in real time and quantificationally. Both the theory analysis and experimental results prove that DTERM provides an effective and novel approach for network risk evaluation.
To effectively evaluate the network risk of network information system, a Danger Theory based Network Risk Evaluation Model (DTREM) is proposed. With definitions of self, non-self, and immunocyte, the intrusion detection sub-model is given. DTERM is composed of memory detectors, mature detectors, and immature detectors. Furthermore, the danger theory based network risk evaluation sub-model is given. In the proposed model, the risk of each network attack, including holistic risk of the host and network, can be calculated in real time and quantificationally. Both the theory analysis and experimental results prove that DTERM provides an effective and novel approach for network risk evaluation.
2007, 36(6): 1202-1205,1218.
Abstract:
This paper presents a mobile agent system security mechanism for large scale network applications. In this mechanism, we use a bidirectional authentication technology based on mixed encryption to resolve the mobile agent system local security problem. Mobile Agent Security Management Platforms (MASMP) working on different levels is applied to deal with the key distribution and management and control the emigration and task attribution of mobile agents. In order to resolve the problem of mobile agent security transference from one network to another, a task transfer model based on the credibility between two conjoint level MASMPs is designed. All these plans together resolve the mobile agent system security problem well. Because of its hiberarchy system structure and management model, this mechanism is very suit for large scale network applications.
This paper presents a mobile agent system security mechanism for large scale network applications. In this mechanism, we use a bidirectional authentication technology based on mixed encryption to resolve the mobile agent system local security problem. Mobile Agent Security Management Platforms (MASMP) working on different levels is applied to deal with the key distribution and management and control the emigration and task attribution of mobile agents. In order to resolve the problem of mobile agent security transference from one network to another, a task transfer model based on the credibility between two conjoint level MASMPs is designed. All these plans together resolve the mobile agent system security problem well. Because of its hiberarchy system structure and management model, this mechanism is very suit for large scale network applications.
2007, 36(6): 1206-1210.
Abstract:
Peer-to-Peer (P2P) worms have become one of some major threats to peer-to-peer network security nowadays. With its strong hidden feature and destructivity, the P2P worms can attack the goal nodes accurately by manipulating the router table of infected nodes, getting information of their neighbors and constructing a target list. Through some simulation experiments, this paper analyzes the characteristic of P2P worms spreading through BitTorrent network, and tests the effect of some parameters. The performances of simulation experiments show that the spread of P2P worms is related to the state of BT network, the attack capacity of worms, the rate of initial infected node, and immune node per unit time.
Peer-to-Peer (P2P) worms have become one of some major threats to peer-to-peer network security nowadays. With its strong hidden feature and destructivity, the P2P worms can attack the goal nodes accurately by manipulating the router table of infected nodes, getting information of their neighbors and constructing a target list. Through some simulation experiments, this paper analyzes the characteristic of P2P worms spreading through BitTorrent network, and tests the effect of some parameters. The performances of simulation experiments show that the spread of P2P worms is related to the state of BT network, the attack capacity of worms, the rate of initial infected node, and immune node per unit time.
2007, 36(6): 1211-1214.
Abstract:
Session Initiation Protocol (SIP) is an important protocol adopted by the 3rd Generation Partnership Project (3GPP) for the IP Multimedia Subsystem (IMS). Facing the complex and open Internet environment, the security of SIP is poor. After the analysis of the security threats to SIP aiming at cleartext transmission of SIP message and lack of authentication methods, the scheme of data encryption and authentication are researched. And then, we discuss how using the existent techniques for improving the security of SIP and propose the ideas of the security solution to SIP.
Session Initiation Protocol (SIP) is an important protocol adopted by the 3rd Generation Partnership Project (3GPP) for the IP Multimedia Subsystem (IMS). Facing the complex and open Internet environment, the security of SIP is poor. After the analysis of the security threats to SIP aiming at cleartext transmission of SIP message and lack of authentication methods, the scheme of data encryption and authentication are researched. And then, we discuss how using the existent techniques for improving the security of SIP and propose the ideas of the security solution to SIP.
2007, 36(6): 1215-1218.
Abstract:
Internet and Web servers become the core infrastructure for companies and institutes. Simultaneously, Web servers also become a popular target for attackers. However, misuse Intrusion Detection Systems (IDSs) are only effective in detecting known attacks and it is difficult to keep up with the daily exploitation of novel and Web-related vulnerabilities; anomaly IDSs often produce a high false alarm rate. To get over the limitations of misuse and anomaly IDSs, this paper inspired by immune principles presents a novel anomaly detection approach to detect unknown Web attacks. In our proposed approach, which is referred to the immune principles Inspired Approach to Detection of Web attacks (IADW), mathematical formulas of self, non-self, antigen, library of antibody genes, immunocyte, and etc., are given, and immune-learning algorithm is described. Experiment results show that our approach can detect unknown attacks with lower false alarm rate, missing alarm rate, and higher detection rate and identification rate than the technique based on neural network and ID3. Thus, it provides an effective novel solution to detection of Web attacks.
Internet and Web servers become the core infrastructure for companies and institutes. Simultaneously, Web servers also become a popular target for attackers. However, misuse Intrusion Detection Systems (IDSs) are only effective in detecting known attacks and it is difficult to keep up with the daily exploitation of novel and Web-related vulnerabilities; anomaly IDSs often produce a high false alarm rate. To get over the limitations of misuse and anomaly IDSs, this paper inspired by immune principles presents a novel anomaly detection approach to detect unknown Web attacks. In our proposed approach, which is referred to the immune principles Inspired Approach to Detection of Web attacks (IADW), mathematical formulas of self, non-self, antigen, library of antibody genes, immunocyte, and etc., are given, and immune-learning algorithm is described. Experiment results show that our approach can detect unknown attacks with lower false alarm rate, missing alarm rate, and higher detection rate and identification rate than the technique based on neural network and ID3. Thus, it provides an effective novel solution to detection of Web attacks.
2007, 36(6): 1219-1222.
Abstract:
To detect increasing mutated script-virus, an artificial immunity-based script-virus detection model is presented. Self-tolerance and clone selection mechanism in artificial immune system are used for reference. Script-code is presented to antigen. Antibody simulates detector in the environment of virus detection. Antibody is classified according to immunity learning mechanism. The negative selection algorithm is simulated to avoid recognizing normal antigen. Self-learning of antibody is used to discover harmful mutated antigens. Mature and memory antibody are evolved to set up dynamic production and elimination mechanism of antibody to decrease false detection rate. The simulation experiment shows that proposed model is able to detect mutated script-virus and provides a new way to detect script-virus.
To detect increasing mutated script-virus, an artificial immunity-based script-virus detection model is presented. Self-tolerance and clone selection mechanism in artificial immune system are used for reference. Script-code is presented to antigen. Antibody simulates detector in the environment of virus detection. Antibody is classified according to immunity learning mechanism. The negative selection algorithm is simulated to avoid recognizing normal antigen. Self-learning of antibody is used to discover harmful mutated antigens. Mature and memory antibody are evolved to set up dynamic production and elimination mechanism of antibody to decrease false detection rate. The simulation experiment shows that proposed model is able to detect mutated script-virus and provides a new way to detect script-virus.
2007, 36(6): 1223-1227.
Abstract:
Along with the widespread concern of spam problem, at present, there are spam filtering system about the problem of semantic imperfection and spam filter low effect in the multi-send spam. This paper proposes a model of spam filtering which based on Latent Semantic Analysis (LSA) and Message-Digest algorithm 5 (MD5). By making use of the LSA marks the latent feature phrase in the spam, a semantic analysis is introduced into the spam filtering technique; the "e-mail fingerprint" of multi-send spam is born with MD5 on the LSA analytical foundation, the problem of filtering technique's low effect in the multi-send spam is resolved with this kind of method. We design a spam filtering system based on this model. This system is evaluated with an optional dataset. The results obtained are compared with Naïve Bayes algorithm filter experiment results. The experiments show the expected results, and the feasibility and advantage of the new spam filtering method is validated.
Along with the widespread concern of spam problem, at present, there are spam filtering system about the problem of semantic imperfection and spam filter low effect in the multi-send spam. This paper proposes a model of spam filtering which based on Latent Semantic Analysis (LSA) and Message-Digest algorithm 5 (MD5). By making use of the LSA marks the latent feature phrase in the spam, a semantic analysis is introduced into the spam filtering technique; the "e-mail fingerprint" of multi-send spam is born with MD5 on the LSA analytical foundation, the problem of filtering technique's low effect in the multi-send spam is resolved with this kind of method. We design a spam filtering system based on this model. This system is evaluated with an optional dataset. The results obtained are compared with Naïve Bayes algorithm filter experiment results. The experiments show the expected results, and the feasibility and advantage of the new spam filtering method is validated.
2007, 36(6): 1228-1231.
Abstract:
K-anonymity is an important privacy preserving model in the data publishing scenario. The algorithms on dataset K-anonymization are researched extensively in recent years, Median Mondrian algorithm is the only multidimensional K-anonymity partition method. However, our research shows that Median Mondrian algorithm is not well-balanced on dealing with the contradiction between data partition precision and data privacy preserving. In this paper, we propose an entropy-based multidimensional K-anonymity partition method and a new evaluation measure on K-anonymization results. The experimental results show that our new method is feasible and preserves the privacy much more efficiently than Median Mondrian algorithm.
K-anonymity is an important privacy preserving model in the data publishing scenario. The algorithms on dataset K-anonymization are researched extensively in recent years, Median Mondrian algorithm is the only multidimensional K-anonymity partition method. However, our research shows that Median Mondrian algorithm is not well-balanced on dealing with the contradiction between data partition precision and data privacy preserving. In this paper, we propose an entropy-based multidimensional K-anonymity partition method and a new evaluation measure on K-anonymization results. The experimental results show that our new method is feasible and preserves the privacy much more efficiently than Median Mondrian algorithm.
2007, 36(6): 1232-1236.
Abstract:
A rapid method for text tendency classification is proposed in this paper. By means of class space model to display the tendency of the words to the categories, the method realizes the classification based on the statistic characteristics of words. In this method, through the studies of the complexity of text tendency categorization, three statistic characteristics of word such as frequency, document frequency and the distribution of words are comprehensively taken into account, and a new method of twice feature selection is proposed:In the first characteristic selection process, using combination characteristic selection method, the words that those distributions are uniform in each category and the low-frequency words are deleted. Then in the second process, the words that those category tendencies are not obvious are deleted. The experimental results show that the algorithm is running-fast, and has high performance.
A rapid method for text tendency classification is proposed in this paper. By means of class space model to display the tendency of the words to the categories, the method realizes the classification based on the statistic characteristics of words. In this method, through the studies of the complexity of text tendency categorization, three statistic characteristics of word such as frequency, document frequency and the distribution of words are comprehensively taken into account, and a new method of twice feature selection is proposed:In the first characteristic selection process, using combination characteristic selection method, the words that those distributions are uniform in each category and the low-frequency words are deleted. Then in the second process, the words that those category tendencies are not obvious are deleted. The experimental results show that the algorithm is running-fast, and has high performance.
2007, 36(6): 1237-1240.
Abstract:
This paper puts forward an efficient algorithm for reduction of attribute in decision systems. A relatively reasonable formula measuring attribute significance is discussed and a quick method to compute partition is proposed. Then a quick algorithm for reduction of attribute is obtained. Compared with those existed algorithms, its flexibility has been increased because calculating the important value of unimportant attributes repeatedly can be avoided by removing unimportant attributes gradually from the search space. The theoretical analysis shows that this algorithm is much less time complexity than those existed algorithms. A real example and experimental results demonstrate its feasibility and effectiveness, respectively.
This paper puts forward an efficient algorithm for reduction of attribute in decision systems. A relatively reasonable formula measuring attribute significance is discussed and a quick method to compute partition is proposed. Then a quick algorithm for reduction of attribute is obtained. Compared with those existed algorithms, its flexibility has been increased because calculating the important value of unimportant attributes repeatedly can be avoided by removing unimportant attributes gradually from the search space. The theoretical analysis shows that this algorithm is much less time complexity than those existed algorithms. A real example and experimental results demonstrate its feasibility and effectiveness, respectively.
2007, 36(6): 1241-1244.
Abstract:
An adaptive Generalized Hebbian Algorithm (GHA) is presented which can be used to approach the intrinsic dimension of an input data set. The classifier design based on adaptive GHA networks is given in detail and the determination method of the classifier parameters is also described. The classifier can be trained by using supervised manner. We applied this approach to the domain of intrusion detection. Some simulations are carried out for anomaly detection by using labeled normal type network connections, and the misuse detection are performed on specified type attacks of denial-of-service intrusions. All the training and testing datasets are based on the KDD CUP 1999 intrusion evaluation data set. Performance comparisons are also made with other recent published methods.
An adaptive Generalized Hebbian Algorithm (GHA) is presented which can be used to approach the intrinsic dimension of an input data set. The classifier design based on adaptive GHA networks is given in detail and the determination method of the classifier parameters is also described. The classifier can be trained by using supervised manner. We applied this approach to the domain of intrusion detection. Some simulations are carried out for anomaly detection by using labeled normal type network connections, and the misuse detection are performed on specified type attacks of denial-of-service intrusions. All the training and testing datasets are based on the KDD CUP 1999 intrusion evaluation data set. Performance comparisons are also made with other recent published methods.
2007, 36(6): 1245-1248.
Abstract:
Neural networks are used more frequently in lossy data coding domains such as audio, image, etc than in general lossless data coding, because standard neural networks must be trained off-line and they are too slow to be practical. In this paper, an adaptive arithmetic coding algorithm based on maximum entropy and neural networks are proposed for data compression. This adaptive algorithm with simply structure can do on-line learning and does not need to be trained off-line. The experiments show that this algorithm surpasses those traditional coding method, such as Limper-Ziv compressors (zip, gzip), in compressing rate and is competitive in speed and time with those traditional coding method such as PPM and Burrows-Wheeler algorithms. The compressor is a bit-level predictive arithmetic which using a 2 layer network with muti-input and one output. The arithmetic, according with the context constriction, improves the precision of prediction and reduces the coding time.
Neural networks are used more frequently in lossy data coding domains such as audio, image, etc than in general lossless data coding, because standard neural networks must be trained off-line and they are too slow to be practical. In this paper, an adaptive arithmetic coding algorithm based on maximum entropy and neural networks are proposed for data compression. This adaptive algorithm with simply structure can do on-line learning and does not need to be trained off-line. The experiments show that this algorithm surpasses those traditional coding method, such as Limper-Ziv compressors (zip, gzip), in compressing rate and is competitive in speed and time with those traditional coding method such as PPM and Burrows-Wheeler algorithms. The compressor is a bit-level predictive arithmetic which using a 2 layer network with muti-input and one output. The arithmetic, according with the context constriction, improves the precision of prediction and reduces the coding time.
2007, 36(6): 1249-1252.
Abstract:
In this paper, we investigate an online forum and construct a social network with the replying relations between comments mapped to the comment authors' relations. We analyze the network's feature and apply the statistical physics methods to this complex network research. Then we validate its characteristics of small world networks and successfully find out the opinion leaders in on-line community based on small-world network.
In this paper, we investigate an online forum and construct a social network with the replying relations between comments mapped to the comment authors' relations. We analyze the network's feature and apply the statistical physics methods to this complex network research. Then we validate its characteristics of small world networks and successfully find out the opinion leaders in on-line community based on small-world network.
2007, 36(6): 1253-1256.
Abstract:
This paper studies application grid clustering method to distinguish the target observation data which is received by the same radar. The fusion of the same goal's observation data received by different radars is realized through the integration of different observation data, so as to realize real-time tracking of multiple targets. The basic thought of grid observation data clustering, and the formal description algorithm are studied. The Kalman filter equations for maneuvering target tracking are described. Parameter matrix theoretical basis for the simplified calculation and the corresponding initial matrix are given for air traffic control system. The simulation result indicates that the grid clustering is able to discriminate different targets well and to carry on the track fusion more accurate.
This paper studies application grid clustering method to distinguish the target observation data which is received by the same radar. The fusion of the same goal's observation data received by different radars is realized through the integration of different observation data, so as to realize real-time tracking of multiple targets. The basic thought of grid observation data clustering, and the formal description algorithm are studied. The Kalman filter equations for maneuvering target tracking are described. Parameter matrix theoretical basis for the simplified calculation and the corresponding initial matrix are given for air traffic control system. The simulation result indicates that the grid clustering is able to discriminate different targets well and to carry on the track fusion more accurate.
2007, 36(6): 1257-1259,1272.
Abstract:
A method of fuzzy clustering based on genetic algorithms is proposed in this paper. This method applies the improved genetic arithmetic to optimization of the Fuzzy C-Mean (FCM) arithmetic. FCM arithmetic has the limitation of converging to the local infinitesimal point, in our method, some interrelated key technique problems, such as encoding method, genetic operators, restrict condition, fitness function for the traditional genetic algorithm, are further reformed. Experiment results show that the method can search global optimum partly so that the clustering results are better than those of only using the FCM.
A method of fuzzy clustering based on genetic algorithms is proposed in this paper. This method applies the improved genetic arithmetic to optimization of the Fuzzy C-Mean (FCM) arithmetic. FCM arithmetic has the limitation of converging to the local infinitesimal point, in our method, some interrelated key technique problems, such as encoding method, genetic operators, restrict condition, fitness function for the traditional genetic algorithm, are further reformed. Experiment results show that the method can search global optimum partly so that the clustering results are better than those of only using the FCM.
2007, 36(6): 1260-1263.
Abstract:
Traditional methods have poor efficiency and effect to deal with the scientific data series forecasting. In this paper, a forecasting algorithm based on grey theory and self-organized map neural networks is proposed. Firstly, the scientific data time series cluster in self-organized mannar. Then the forecast model is established with grey theory. In clustering, a distance criterion is proposed to scale the difference between series. In grey theory, the whiten parameter is optimized. The experiments show that this algorithm surpasses those traditional forecasting methods in precision and time efficiency.
Traditional methods have poor efficiency and effect to deal with the scientific data series forecasting. In this paper, a forecasting algorithm based on grey theory and self-organized map neural networks is proposed. Firstly, the scientific data time series cluster in self-organized mannar. Then the forecast model is established with grey theory. In clustering, a distance criterion is proposed to scale the difference between series. In grey theory, the whiten parameter is optimized. The experiments show that this algorithm surpasses those traditional forecasting methods in precision and time efficiency.
2007, 36(6): 1264-1267.
Abstract:
This paper proposes a clustering algorithm based on Particle Swarm Optimization Algorithm with Immunity-Vaccination (IV-PSO-KMEANS). It combines Particle Swarm Optimization (PSO) algorithm and K-means for clustering. Synchronously, Immunity-vaccination and immunity-selection mechanisms of immune system are introduced into the iterative procedure. mmunity-vaccination is used to direct the procedure of particle swarm and immunity-selection is applied to select from the results of vaccination. In result, the swarm is made to move towards a better direction. The experiments show that the IV-PSO-KMEANS algorithm overcomes the problem of K-means algorithm that the results are related to the initial clustering centers, and the results of clustering are steadier and better than algorithms based on PSO.
This paper proposes a clustering algorithm based on Particle Swarm Optimization Algorithm with Immunity-Vaccination (IV-PSO-KMEANS). It combines Particle Swarm Optimization (PSO) algorithm and K-means for clustering. Synchronously, Immunity-vaccination and immunity-selection mechanisms of immune system are introduced into the iterative procedure. mmunity-vaccination is used to direct the procedure of particle swarm and immunity-selection is applied to select from the results of vaccination. In result, the swarm is made to move towards a better direction. The experiments show that the IV-PSO-KMEANS algorithm overcomes the problem of K-means algorithm that the results are related to the initial clustering centers, and the results of clustering are steadier and better than algorithms based on PSO.
2007, 36(6): 1268-1272.
Abstract:
Enlightened by biological immune system, this paper applies the idea of vaccine extraction and vaccination, this paper proposes quantum-behaved particle swarm optimization with immunity algorithm (IQPSO). In this algorithm, vaccination can guide the particles to evolve towards much better direction. Experiments show that attribute reduction based on IQPSO algorithm achieve much better result both in convergence speed and optimization capabilities in comparison with other algorithms, such as Hu algorithm, particle swarm optimization, and quantum-behaved particle swarm optimization.
Enlightened by biological immune system, this paper applies the idea of vaccine extraction and vaccination, this paper proposes quantum-behaved particle swarm optimization with immunity algorithm (IQPSO). In this algorithm, vaccination can guide the particles to evolve towards much better direction. Experiments show that attribute reduction based on IQPSO algorithm achieve much better result both in convergence speed and optimization capabilities in comparison with other algorithms, such as Hu algorithm, particle swarm optimization, and quantum-behaved particle swarm optimization.
2007, 36(6): 1273-1277.
Abstract:
Detecting and eliminating approximately duplicated records is one of the main problems needed to be solved for data mining and data quality improvement. An algorithm for detecting approximately duplicated database records is presented based on rank group. Firstly, each property of the data is endowed with certain weight according rank-based weights method. Secondly, in term of group thought, large data sets are divided into many non-intersect small data sets. Finally, approximately duplicated records are detected and eliminated in each small data set. To avoid missing, the above steps can be repeated. The theory analysis and experiment show that this algorithm has a good detecting precision better efficiency of time, and therefore is an effective approach to solve approximately duplicate records of massive data.
Detecting and eliminating approximately duplicated records is one of the main problems needed to be solved for data mining and data quality improvement. An algorithm for detecting approximately duplicated database records is presented based on rank group. Firstly, each property of the data is endowed with certain weight according rank-based weights method. Secondly, in term of group thought, large data sets are divided into many non-intersect small data sets. Finally, approximately duplicated records are detected and eliminated in each small data set. To avoid missing, the above steps can be repeated. The theory analysis and experiment show that this algorithm has a good detecting precision better efficiency of time, and therefore is an effective approach to solve approximately duplicate records of massive data.
2007, 36(6): 1278-1281.
Abstract:
A new model of electricity price forecasting based on cluster analysis is proposed. The complex forecasting problem is divided into simpler problems in the presented model. The whole input space is partitioned into several disjointed regions. Then, support vector machine is used for modeling and forecasting for each region. In the process of cluster analysis, K-means algorithm is used for further optimizing after the number of partitioned regions and initial cluster centers are automatically obtained by using subtractive clustering method. The simulation research using the historical data from PJM market shows that the proposed model can improve the precision of electricity price forecasting effectively and stably.
A new model of electricity price forecasting based on cluster analysis is proposed. The complex forecasting problem is divided into simpler problems in the presented model. The whole input space is partitioned into several disjointed regions. Then, support vector machine is used for modeling and forecasting for each region. In the process of cluster analysis, K-means algorithm is used for further optimizing after the number of partitioned regions and initial cluster centers are automatically obtained by using subtractive clustering method. The simulation research using the historical data from PJM market shows that the proposed model can improve the precision of electricity price forecasting effectively and stably.
2007, 36(6): 1282-1285.
Abstract:
Data mining can help business enterprise get valuable information from continual accumulated and updated data sources. This paper uses seller's commodity sale database to investigate the correlations among commodities. Especially, aiming to the shortage of association rule algorithm in mining the correlation among commodities, this paper proposes a new algorithm. Based on daily sale data record of commodities, we obtain their sale data time series according to the change of commodities' sales, then compare these time series, measure their distance, and finally get correlations of commodities. Some experiments on real data sets validate the effectiveness of our proposed method. And we obtain some valuable results, which can guide the business application.
Data mining can help business enterprise get valuable information from continual accumulated and updated data sources. This paper uses seller's commodity sale database to investigate the correlations among commodities. Especially, aiming to the shortage of association rule algorithm in mining the correlation among commodities, this paper proposes a new algorithm. Based on daily sale data record of commodities, we obtain their sale data time series according to the change of commodities' sales, then compare these time series, measure their distance, and finally get correlations of commodities. Some experiments on real data sets validate the effectiveness of our proposed method. And we obtain some valuable results, which can guide the business application.
2007, 36(6): 1286-1288,1295.
Abstract:
In order to make the outlier detection more automatic and decrease the users' difficulty for the selection of parameters, an outlier detection method with a new definition of average density is proposed. In this method, the outlier's density is considered smaller than the average density of data set and the none-outlier's density shouldn't decrease with its closed interval compression. An experiment is used to identify the outline of the animal's body. The experimental results show that the method identifies the face's outline effectively.
In order to make the outlier detection more automatic and decrease the users' difficulty for the selection of parameters, an outlier detection method with a new definition of average density is proposed. In this method, the outlier's density is considered smaller than the average density of data set and the none-outlier's density shouldn't decrease with its closed interval compression. An experiment is used to identify the outline of the animal's body. The experimental results show that the method identifies the face's outline effectively.
2007, 36(6): 1289-1291,1314.
Abstract:
Clustering analysis is a very useful tool in the domain of data mining for searching distributing mode from a great deal of data. Its main algorithms are partition-based algorithm, hierarchy-based algorithm, density-based algorithm, grid-based algorithm, and model-based algorithm. The paper mainly discusses a clustering algorithm based on density and grid in data mining, which has high clustering efficiency and low time complexity. It is efficient and effective for multi-density and uniformity density data sets with noise and suitable for batch update. After that an incremental clustering technique is presented. This technique not only makes best use of the former clustering results and improves the efficiency of clustering analysis, but also brings to the reduction of enormous expenditure on knowledge base maintenance. At last an application of the algorithm in Customer Relationship Management (CRM) is gien.
Clustering analysis is a very useful tool in the domain of data mining for searching distributing mode from a great deal of data. Its main algorithms are partition-based algorithm, hierarchy-based algorithm, density-based algorithm, grid-based algorithm, and model-based algorithm. The paper mainly discusses a clustering algorithm based on density and grid in data mining, which has high clustering efficiency and low time complexity. It is efficient and effective for multi-density and uniformity density data sets with noise and suitable for batch update. After that an incremental clustering technique is presented. This technique not only makes best use of the former clustering results and improves the efficiency of clustering analysis, but also brings to the reduction of enormous expenditure on knowledge base maintenance. At last an application of the algorithm in Customer Relationship Management (CRM) is gien.
2007, 36(6): 1292-1295.
Abstract:
An efficient and quick method based on 2-D Principal Component Analysis (PCA) and coherence measure is introduced. The coordinates are achieved by projecting the high dimensional data to the 2-D space after the principle component space is built and feature extraction is finished at one time. Every principle component is the linear combination of the original variables and is irrelevant to each other. A novel coherence measure is introduced and designed for effectively measuring the coherence of a new specimen of unknown type with the training samples. The spectrum can be classified quickly and exactly by the classifier.
An efficient and quick method based on 2-D Principal Component Analysis (PCA) and coherence measure is introduced. The coordinates are achieved by projecting the high dimensional data to the 2-D space after the principle component space is built and feature extraction is finished at one time. Every principle component is the linear combination of the original variables and is irrelevant to each other. A novel coherence measure is introduced and designed for effectively measuring the coherence of a new specimen of unknown type with the training samples. The spectrum can be classified quickly and exactly by the classifier.
2007, 36(6): 1296-1299.
Abstract:
This paper analyzes the international standard of Near Field Communication (NFC). Authors survey the developing history of the technology; describe the current status, as well as point out the tendency of them. The model in commercial environment and the problems of application development are analyzed. Security has become an important factor in application. So authors also discuss the security problems from link level and application level. This paper analyses the existing problems and the goal of further developments.
This paper analyzes the international standard of Near Field Communication (NFC). Authors survey the developing history of the technology; describe the current status, as well as point out the tendency of them. The model in commercial environment and the problems of application development are analyzed. Security has become an important factor in application. So authors also discuss the security problems from link level and application level. This paper analyses the existing problems and the goal of further developments.
2007, 36(6): 1300-1303.
Abstract:
Most of the existed Ad hoc routing protocols have the disadvantages of high delay and can not satisfy the transmission requirements of real-time applications. Based on the Ad hoc On-demand Distance Vector Routing (AODV) protocol, a Real-time aware Ad hoc On-demand Distance Vector routing (RAODV) protocol is proposed in this paper. The performances of RAODV and AODV under different network conditions are compared by simulation experiments. The experiment results show that RAODV can obtain as good packet delivery fraction as AODV, but has lower average end-to-end delay of data packets and lower routing load than AODV under normal network load condition. Under high network load condition, the performances of the two protocols both declined. However, the performance of RAODV declined much smaller than that of AODV:the former has higher packet delivery fraction, lower average end-to-end delay of data packets, and routing load than the later.
Most of the existed Ad hoc routing protocols have the disadvantages of high delay and can not satisfy the transmission requirements of real-time applications. Based on the Ad hoc On-demand Distance Vector Routing (AODV) protocol, a Real-time aware Ad hoc On-demand Distance Vector routing (RAODV) protocol is proposed in this paper. The performances of RAODV and AODV under different network conditions are compared by simulation experiments. The experiment results show that RAODV can obtain as good packet delivery fraction as AODV, but has lower average end-to-end delay of data packets and lower routing load than AODV under normal network load condition. Under high network load condition, the performances of the two protocols both declined. However, the performance of RAODV declined much smaller than that of AODV:the former has higher packet delivery fraction, lower average end-to-end delay of data packets, and routing load than the later.
2007, 36(6): 1304-1307.
Abstract:
Low-power design is one of the most important issues in Wireless Sensor Networks (WSNs), where nodes are likely to be largely scattered with constrained battery energy.Transmitting power control is a simple method to make the power consumption down, while network connectivity and excessive interferences should be considered. In this paper,an optimal Transmitting power is derived by considering average interferences and network connectivity. To make the network energy efficient, we present a novel protocol via Adaptive Rate Control with Optimal Power (ARCOP). Compared with MCP(Maximum Common Power,e.g. 802.11DCF with fixed maximum power) and Basic protocol, Simulation results show that the new protocol can achieve high average throughput while minimizing Energy Consumption Per Useful Bit (ECPUB).
Low-power design is one of the most important issues in Wireless Sensor Networks (WSNs), where nodes are likely to be largely scattered with constrained battery energy.Transmitting power control is a simple method to make the power consumption down, while network connectivity and excessive interferences should be considered. In this paper,an optimal Transmitting power is derived by considering average interferences and network connectivity. To make the network energy efficient, we present a novel protocol via Adaptive Rate Control with Optimal Power (ARCOP). Compared with MCP(Maximum Common Power,e.g. 802.11DCF with fixed maximum power) and Basic protocol, Simulation results show that the new protocol can achieve high average throughput while minimizing Energy Consumption Per Useful Bit (ECPUB).
2007, 36(6): 1308-1310.
Abstract:
The iterative threshold choosing algorithm is used to changes the gray image into binary image in the low half of a known face image, the area of mouth and nose can be obtained without manual work. This algorithm improves the traditional method limitation, such as vague mouth contour and poor connection quality. The erosion and dilation arithmetic is used to delete the minor noise area; the minimum enclosing rectangle is used to describe the shape of every area; and the mouth area is determined according to the prior knowledge of the mouth shape and size. The mouth center is represented by the centroid of the mouth connection area. Harris corner detection arithmetic is used to detect the two corners of the mouth in the mouth area of the original gray image. Experiment shows that the algorithm presented in this paper is quicker and more accurate than traditional methods.
The iterative threshold choosing algorithm is used to changes the gray image into binary image in the low half of a known face image, the area of mouth and nose can be obtained without manual work. This algorithm improves the traditional method limitation, such as vague mouth contour and poor connection quality. The erosion and dilation arithmetic is used to delete the minor noise area; the minimum enclosing rectangle is used to describe the shape of every area; and the mouth area is determined according to the prior knowledge of the mouth shape and size. The mouth center is represented by the centroid of the mouth connection area. Harris corner detection arithmetic is used to detect the two corners of the mouth in the mouth area of the original gray image. Experiment shows that the algorithm presented in this paper is quicker and more accurate than traditional methods.
2007, 36(6): 1311-1314.
Abstract:
In this paper, a new content-based image retrieval method is proposed. By using the information entropy theory, the method describes the color, shape, and texture features of images, respectively. An integrated similarity match algorithm is presented. Firstly, an image is segmented into several sub-areas, and the angle information of the shape is computed after shape extracted from the image. Secondly, mutual information of color, shape, and texture among images is computed using information entropy. Compared with the other algorithms, experiments indicate that this approach has the advantages of good effect.
In this paper, a new content-based image retrieval method is proposed. By using the information entropy theory, the method describes the color, shape, and texture features of images, respectively. An integrated similarity match algorithm is presented. Firstly, an image is segmented into several sub-areas, and the angle information of the shape is computed after shape extracted from the image. Secondly, mutual information of color, shape, and texture among images is computed using information entropy. Compared with the other algorithms, experiments indicate that this approach has the advantages of good effect.
2007, 36(6): 1315-1318.
Abstract:
The modes and applications of identity authentication in common used E-Procurement System (EPS) are analyzed in this paper firstly. Then, the dynamic identity authentication policy model is acquired according to Policy, Protection, Detection, and Response (P2DR) model and architecture. Finally, the identity authentication policy model based on P2DR is proposed for the security requirement of EPS. Considering all kinds of factors such as policy, cost, working scope and so on, this policy model can adjust the modes of EPS dynamically and realize the homeostasis among "security, cost and efficiency" for different security requirements and users in each stage of e-procurement.
The modes and applications of identity authentication in common used E-Procurement System (EPS) are analyzed in this paper firstly. Then, the dynamic identity authentication policy model is acquired according to Policy, Protection, Detection, and Response (P2DR) model and architecture. Finally, the identity authentication policy model based on P2DR is proposed for the security requirement of EPS. Considering all kinds of factors such as policy, cost, working scope and so on, this policy model can adjust the modes of EPS dynamically and realize the homeostasis among "security, cost and efficiency" for different security requirements and users in each stage of e-procurement.
2007, 36(6): 1319-1321.
Abstract:
In data mining reseach, the classification algorithms generally pursue more highly accuracy. It is based on the assumption that all misclassifications have the same cost. However, the assumption is not correct in the real world, so that the normal classification algorithms do not perform well. By improving the encode/decode methods and taking different misclassification cost into account, this paper concerns a new cost-sensitive algorithm called CSC-GEP based on Gene Expression Programming (GEP). The experimental results show that the new algorithm is effective.
In data mining reseach, the classification algorithms generally pursue more highly accuracy. It is based on the assumption that all misclassifications have the same cost. However, the assumption is not correct in the real world, so that the normal classification algorithms do not perform well. By improving the encode/decode methods and taking different misclassification cost into account, this paper concerns a new cost-sensitive algorithm called CSC-GEP based on Gene Expression Programming (GEP). The experimental results show that the new algorithm is effective.
2007, 36(6): 1322-1324,1353.
Abstract:
In this paper, the activity based costing is introduced to overcome the limitation of traditional cost estimating method and complex software process. By confirming and measuring the cost of an Activity, the proposed method can provide useful information to improve those "value-added" operations. Three different software processes' costs are analyzed. The result show that the activity based costing can describe the software process cost accurately.
In this paper, the activity based costing is introduced to overcome the limitation of traditional cost estimating method and complex software process. By confirming and measuring the cost of an Activity, the proposed method can provide useful information to improve those "value-added" operations. Three different software processes' costs are analyzed. The result show that the activity based costing can describe the software process cost accurately.
2007, 36(6): 1325-1328.
Abstract:
According to the multiple accessing technique of Code Division Multiple Access (CDMA) system, multiple watermarks are encoded with convolution and interleaved in block. Orthogonal Gold sequences are used to spread spectrum of the copyright messages. CDMA encoded copyright messages are embedded into the wavelet sub-bands except HH1 sub-band. The embedded amplitude is decided by Watson's perceptual model of wavelet transform domain, and the embedded position in the selected wavelet sub-bands is decided randomly by a Pseudorandom Noise (PN) sequence. Watermarks are extracted without the need of original image. The capacity of proposed algorithm is also discussed. Experimental results show that proposed algorithm improves the detection Bits Error Rate (BER) and that multiple watermarks have preferable robustness and invisibility.
According to the multiple accessing technique of Code Division Multiple Access (CDMA) system, multiple watermarks are encoded with convolution and interleaved in block. Orthogonal Gold sequences are used to spread spectrum of the copyright messages. CDMA encoded copyright messages are embedded into the wavelet sub-bands except HH1 sub-band. The embedded amplitude is decided by Watson's perceptual model of wavelet transform domain, and the embedded position in the selected wavelet sub-bands is decided randomly by a Pseudorandom Noise (PN) sequence. Watermarks are extracted without the need of original image. The capacity of proposed algorithm is also discussed. Experimental results show that proposed algorithm improves the detection Bits Error Rate (BER) and that multiple watermarks have preferable robustness and invisibility.
2007, 36(6): 1329-1332,1349.
Abstract:
Radio Frequency Identification (RFID) is one of the most promising technology of information collection. This paper brings forward a model based on service-oriented architecture and web services, which is used for integrating the information systems of enterprises. On the basis of the model, the information platform makes full use of the data mining and computational intelligence to build the intelligent decision support system for the enterprises and to support the daily operation and long-term layout of enterprises. The solution and model will be an upstanding reference for the application of RFID in the logistics industry and the construction of IT infrastructure of logistics enterprises.
Radio Frequency Identification (RFID) is one of the most promising technology of information collection. This paper brings forward a model based on service-oriented architecture and web services, which is used for integrating the information systems of enterprises. On the basis of the model, the information platform makes full use of the data mining and computational intelligence to build the intelligent decision support system for the enterprises and to support the daily operation and long-term layout of enterprises. The solution and model will be an upstanding reference for the application of RFID in the logistics industry and the construction of IT infrastructure of logistics enterprises.
2007, 36(6): 1333-1337.
Abstract:
The rapid development of Peer-to-Peer (P2P) network has enriched the performance of Internet. However, with the violent increment of data flow and bandwidth using with no restrict, the P2P applications have brought huge impact to Internet base establishment and the advance service of Internet Service Provider (ISP) and Application Service Provider (ASP). Therefore, to utilize the Internet base establishment and P2P technology correctly and to process P2P applications effectively, the technology of P2P network traffic identification and traffic filtering should be researched while restraining the illegal content being transfered in P2P network. This paper focuses on the deep research and comparison of available algorithms of P2P network traffic identification, and gives a technology reference and standard for the P2P network traffic management.
The rapid development of Peer-to-Peer (P2P) network has enriched the performance of Internet. However, with the violent increment of data flow and bandwidth using with no restrict, the P2P applications have brought huge impact to Internet base establishment and the advance service of Internet Service Provider (ISP) and Application Service Provider (ASP). Therefore, to utilize the Internet base establishment and P2P technology correctly and to process P2P applications effectively, the technology of P2P network traffic identification and traffic filtering should be researched while restraining the illegal content being transfered in P2P network. This paper focuses on the deep research and comparison of available algorithms of P2P network traffic identification, and gives a technology reference and standard for the P2P network traffic management.
2007, 36(6): 1338-1341.
Abstract:
The requirement and current status of deep packet inspection are analyzed with applications in Peer-to-Peer (P2P). The difference of traditional network and Next Generation Network (NGN) is presented, and the operator's measures of managing P2P are also summarized. The possibility and necessity of operating P2P are also discussed. After the analysis of current Deep Packet Inspection (DPI) technology, the fact that DPI can not satisfy the telecommunication level requirement is pointed out. A concrete mechanism of telecommunication operator dominating P2P service operation is proposed. This mechanism is different from traditional method in utilizing the superiority of P2P technology properly and synthesizing every participant's advantage.
The requirement and current status of deep packet inspection are analyzed with applications in Peer-to-Peer (P2P). The difference of traditional network and Next Generation Network (NGN) is presented, and the operator's measures of managing P2P are also summarized. The possibility and necessity of operating P2P are also discussed. After the analysis of current Deep Packet Inspection (DPI) technology, the fact that DPI can not satisfy the telecommunication level requirement is pointed out. A concrete mechanism of telecommunication operator dominating P2P service operation is proposed. This mechanism is different from traditional method in utilizing the superiority of P2P technology properly and synthesizing every participant's advantage.
2007, 36(6): 1342-1344.
Abstract:
Based on Peer-to-Peer (P2P), combining the characteristics of mobile computing environment, a structure of mobile P2P network which is adapted to mobile computing environment is put forward. The core idea of this structure is that the mobile embedded devices are indirectly connected to the P2P network with Agent, and then share their resources each other through the P2P network. Then, based on the network structure, a structure of mobile P2P middleware is designed. It is composed of Agent and mobile client. Finally, based on the P2P of JXTA and J2ME mobile computing platform, a file-sharing experiment between two mobile embedded devices validates the effectiveness and practicability of the middleware structure.
Based on Peer-to-Peer (P2P), combining the characteristics of mobile computing environment, a structure of mobile P2P network which is adapted to mobile computing environment is put forward. The core idea of this structure is that the mobile embedded devices are indirectly connected to the P2P network with Agent, and then share their resources each other through the P2P network. Then, based on the network structure, a structure of mobile P2P middleware is designed. It is composed of Agent and mobile client. Finally, based on the P2P of JXTA and J2ME mobile computing platform, a file-sharing experiment between two mobile embedded devices validates the effectiveness and practicability of the middleware structure.
2007, 36(6): 1345-1349.
Abstract:
In order to increase the convergent speed and to improve the overall searching ability of the algorithm,a Particle Swarm Optimization (PSO) method is proposed with adaptive inertia weight by the change of the number of iterations based on the analysis of inertia weight global best fitness of the PSO. The improved PSO increases the ability to avoid local optimum. Then a speaker recognition method using this improved algorithm to train Support Vector Machine (SVM) is presented. The experimental results show that the presented SVM method optimized by PSO for speaker recognition can achieve higher recognition accuracy and higher recognition speed.
In order to increase the convergent speed and to improve the overall searching ability of the algorithm,a Particle Swarm Optimization (PSO) method is proposed with adaptive inertia weight by the change of the number of iterations based on the analysis of inertia weight global best fitness of the PSO. The improved PSO increases the ability to avoid local optimum. Then a speaker recognition method using this improved algorithm to train Support Vector Machine (SVM) is presented. The experimental results show that the presented SVM method optimized by PSO for speaker recognition can achieve higher recognition accuracy and higher recognition speed.
2007, 36(6): 1350-1353.
Abstract:
Replication strategies directly effect on use of grid resources in data grid. As the replication strategies are very important to the data grid, this paper builds up a simulation environment with OptorSim simulators, and analyzes different replication strategies according to different accessing models. Reasearch of data structure of grid, replication strategies, policy of job scheduler, mode of access, and effect on data grid. It is verified by the simulation results that the replication strategies adopted in this paper can cut down the system cost of single operation and make the most use of computing unit.
Replication strategies directly effect on use of grid resources in data grid. As the replication strategies are very important to the data grid, this paper builds up a simulation environment with OptorSim simulators, and analyzes different replication strategies according to different accessing models. Reasearch of data structure of grid, replication strategies, policy of job scheduler, mode of access, and effect on data grid. It is verified by the simulation results that the replication strategies adopted in this paper can cut down the system cost of single operation and make the most use of computing unit.
