Email alert
2016 Vol. 38, No. 9
Display Method:
2016, 38(9): 2130-2140.
doi: 10.11999/JEIT160517
Abstract:
Many enterprises try to promote their products in online social network since information propagation in this network have several advantages such as fast transmission speed, low marketing costs, and large influence area. However, it is a challenging task for enterprises to select suitable seed nodes to publish marketing information so that marketing information can influence or cover most users under a given cost and realize performance maximization. By means of literature search and review, this paper systematically summarizes information propagation models in social marketing, introduces algorithms for social marketing performance maximization problem with respect to network topology, user historical data, compete and non-compete condition. Finally, this paper concludes an exploration of future directions of this research filed.
Many enterprises try to promote their products in online social network since information propagation in this network have several advantages such as fast transmission speed, low marketing costs, and large influence area. However, it is a challenging task for enterprises to select suitable seed nodes to publish marketing information so that marketing information can influence or cover most users under a given cost and realize performance maximization. By means of literature search and review, this paper systematically summarizes information propagation models in social marketing, introduces algorithms for social marketing performance maximization problem with respect to network topology, user historical data, compete and non-compete condition. Finally, this paper concludes an exploration of future directions of this research filed.
2016, 38(9): 2141-2149.
doi: 10.11999/JEIT160515
Abstract:
Devoted to enriching the research system of online group, and laying the foundation for exploring deep scientific problem in the future, this paper discusses the definition of online group, online topic, common classification methods, and primarily introduces a new qualitative method of online topic classification based on observing the trend of a social system response function. Through this method, online topic discussed by online group can be divided into exogenous critical topic, exogenous subcritical topic, endogenous critical topic, and endogenous subcritical topic. The standardized steps to this method are put forward, and it is figured out that the problems may occur when applying it to the practice. Whats more, this method is tried to estimate the distribution of four types of topics in the two representatives of online social network platform Sina microblog and Baidu Tieba.
Devoted to enriching the research system of online group, and laying the foundation for exploring deep scientific problem in the future, this paper discusses the definition of online group, online topic, common classification methods, and primarily introduces a new qualitative method of online topic classification based on observing the trend of a social system response function. Through this method, online topic discussed by online group can be divided into exogenous critical topic, exogenous subcritical topic, endogenous critical topic, and endogenous subcritical topic. The standardized steps to this method are put forward, and it is figured out that the problems may occur when applying it to the practice. Whats more, this method is tried to estimate the distribution of four types of topics in the two representatives of online social network platform Sina microblog and Baidu Tieba.
2016, 38(9): 2150-2157.
doi: 10.11999/JEIT160544
Abstract:
Group recommendation confronts two major problems, i.e., unambiguous definition and identification of groups and efficient recommendation to users in groups. To tackle the two problems, a group recommendation framework based on social network community is proposed. The framework takes into account social network structural information to identify overlapping groups, which is well interpreted; and fulfills the task of recommending to groups by performing aggregation and allocation strategies using the membership of users related to groups, which considers how much users contribute to groups and benefit from groups. Experimental results on publicly open datasets demonstrate its efficiency and accuracy on the task of group recommendation.
Group recommendation confronts two major problems, i.e., unambiguous definition and identification of groups and efficient recommendation to users in groups. To tackle the two problems, a group recommendation framework based on social network community is proposed. The framework takes into account social network structural information to identify overlapping groups, which is well interpreted; and fulfills the task of recommending to groups by performing aggregation and allocation strategies using the membership of users related to groups, which considers how much users contribute to groups and benefit from groups. Experimental results on publicly open datasets demonstrate its efficiency and accuracy on the task of group recommendation.
2016, 38(9): 2158-2164.
doi: 10.11999/JEIT151136
Abstract:
The trajectory K-anonymous is the mainstream of the current trajectory privacy protection, but the method has some defects such as privacy leakage. In this paper, a method of trajectory privacy preserving is proposed Based on Agent Forwarding Mechanism (BAFM) in mobile social networks, which uses secure multi-party computation and inner product secure computation to find the best matching user by the trusted server as the agent. The agent forwards the users request to the server to query, which hides the correlation between users real trajectory and the server in order to achieve users trajectory privacy. Security analysis shows that the propose method can effectively protect the user's trajectory privacy. Experiments show that the proposed method is more effective, it reduces the overhead of servers query and communication.
The trajectory K-anonymous is the mainstream of the current trajectory privacy protection, but the method has some defects such as privacy leakage. In this paper, a method of trajectory privacy preserving is proposed Based on Agent Forwarding Mechanism (BAFM) in mobile social networks, which uses secure multi-party computation and inner product secure computation to find the best matching user by the trusted server as the agent. The agent forwards the users request to the server to query, which hides the correlation between users real trajectory and the server in order to achieve users trajectory privacy. Security analysis shows that the propose method can effectively protect the user's trajectory privacy. Experiments show that the proposed method is more effective, it reduces the overhead of servers query and communication.
2016, 38(9): 2165-2172.
doi: 10.11999/JEIT151479
Abstract:
In mobile social networks, people can quickly find potential friends with the same attributes by sharing personal attribute profile. These attribute profiles, however, usually contain sensitive information, if this information gets intercepted by malicious attackers it may result in unpredictable consequences. In this paper, a dual handshake privacy-preserving scheme is proposed based on user pseudo identity anonymous and hash value authentication, which is combined with identity authentication, one-way hash function and key agreement to ensure that malicious attackers can not get the real content of personal profile by identity fraud, attribute forgery, hacking security attributes and eavesdrop secure channel, thus the personal privacy can be protected. At the same time, the scheme relies on the powerful computing and anti-attack ability to trusted third party to reduce the computation cost of the intelligent terminal and security risks. Security and performance analysis demonstrates that this scheme is of high privacy, non-repudiation and verifiability and is more effective than existing solutions.
In mobile social networks, people can quickly find potential friends with the same attributes by sharing personal attribute profile. These attribute profiles, however, usually contain sensitive information, if this information gets intercepted by malicious attackers it may result in unpredictable consequences. In this paper, a dual handshake privacy-preserving scheme is proposed based on user pseudo identity anonymous and hash value authentication, which is combined with identity authentication, one-way hash function and key agreement to ensure that malicious attackers can not get the real content of personal profile by identity fraud, attribute forgery, hacking security attributes and eavesdrop secure channel, thus the personal privacy can be protected. At the same time, the scheme relies on the powerful computing and anti-attack ability to trusted third party to reduce the computation cost of the intelligent terminal and security risks. Security and performance analysis demonstrates that this scheme is of high privacy, non-repudiation and verifiability and is more effective than existing solutions.
2016, 38(9): 2173-2179.
doi: 10.11999/JEIT160350
Abstract:
The model based on fully-trusted third party is a major model for location privacy protection in location-based services, but the model has some risk of exposing privacy. In this paper, a location privacy protection method based on Grid Identifier Matching (GIM) is proposed. In this method the user first divides the query area into grid and combines the order-preserving symmetric encryption and K-anonymity mechanism. Then, the K-anonymity paradigm is formed in anonymizer. Finally, the query results are returned to users by utilizing GIM. In the query process, the anonymizer dose not have any knowlegdge about a users real location, which can enhance the users location privacy. Meanwhile, the anonymizer only does simple comparison and matching operations, which relieves effectively is performance bottleneck of the anonymizer. Security analysis shows that the proposed approach can effectively protect the users location privacy. Experimental evaluations show that the proposed approach can decrease processing time overhead of the anonymizer.
The model based on fully-trusted third party is a major model for location privacy protection in location-based services, but the model has some risk of exposing privacy. In this paper, a location privacy protection method based on Grid Identifier Matching (GIM) is proposed. In this method the user first divides the query area into grid and combines the order-preserving symmetric encryption and K-anonymity mechanism. Then, the K-anonymity paradigm is formed in anonymizer. Finally, the query results are returned to users by utilizing GIM. In the query process, the anonymizer dose not have any knowlegdge about a users real location, which can enhance the users location privacy. Meanwhile, the anonymizer only does simple comparison and matching operations, which relieves effectively is performance bottleneck of the anonymizer. Security analysis shows that the proposed approach can effectively protect the users location privacy. Experimental evaluations show that the proposed approach can decrease processing time overhead of the anonymizer.
2016, 38(9): 2180-2187.
doi: 10.11999/JEIT151338
Abstract:
To study the effects of information coupling of local topology on the complex network evolution, a new weighted method is proposed based on local topology information, which can measure the closeness of connection and the coupling degree of topology information between nodes. In this paper, to demonstrate the efficiency of the information coupling of local topology, an empirical research is made on characteristic statistics of evolving model and real network data testing of link prediction respectively. Firstly, the weighted method is applied to BA model; TwBA and the local world model TwLW are proposed based on the topology weighted method. Simulation experiments show that the degree distribution of TwBA can be rapidly changed from exponential distribution to power law distribution with the increasing of the connection numbers for new added nodes, which confirmes that the phenomenon of accelerating growth appears widely in the evolution of many real scale-free networks. Then, based on TwBA model, an accelerating growth model A-TwBA is proposed, and the A-TwBA model presents power law distribution for different accelerating growth rates. The degree distribution of TwLW is changed from stretched exponential distribution to power law distribution for different sizes of local world. Finally, the proposed weighted method is applied to link prediction methods (including CN, Salton and RA index), and three weighted indices are proposed. Empirical study shows that the weighted proposed method can significantly improve the prediction accuracy of these basic indices, and some of them are higher than those of the global indices.
To study the effects of information coupling of local topology on the complex network evolution, a new weighted method is proposed based on local topology information, which can measure the closeness of connection and the coupling degree of topology information between nodes. In this paper, to demonstrate the efficiency of the information coupling of local topology, an empirical research is made on characteristic statistics of evolving model and real network data testing of link prediction respectively. Firstly, the weighted method is applied to BA model; TwBA and the local world model TwLW are proposed based on the topology weighted method. Simulation experiments show that the degree distribution of TwBA can be rapidly changed from exponential distribution to power law distribution with the increasing of the connection numbers for new added nodes, which confirmes that the phenomenon of accelerating growth appears widely in the evolution of many real scale-free networks. Then, based on TwBA model, an accelerating growth model A-TwBA is proposed, and the A-TwBA model presents power law distribution for different accelerating growth rates. The degree distribution of TwLW is changed from stretched exponential distribution to power law distribution for different sizes of local world. Finally, the proposed weighted method is applied to link prediction methods (including CN, Salton and RA index), and three weighted indices are proposed. Empirical study shows that the weighted proposed method can significantly improve the prediction accuracy of these basic indices, and some of them are higher than those of the global indices.
2016, 38(9): 2188-2193.
doi: 10.11999/JEIT151328
Abstract:
To implement the high performance web server, it is a key to exploit fully the performance of multi-core CPUs in web servers. Traditional First Come First Served (FCFS) policy does not consider the characteristic of multi-core processors, and could not fully exploit its performance. To address this problem, a dynamic request scheduling algorithm based on allocation matrix is proposed in this paper. The algorithm fully considers the characteristic of multi-core processors, assigns the same kind of dynamic requests to the same processing core, and improves the speed of handling dynamic requests in web server. The results of the experiment show that the web server that adopts this algorithm is superior to the traditional FCFS algorithm in the aspect of similarity, mean response time and dropped rate of dynamic requests.
To implement the high performance web server, it is a key to exploit fully the performance of multi-core CPUs in web servers. Traditional First Come First Served (FCFS) policy does not consider the characteristic of multi-core processors, and could not fully exploit its performance. To address this problem, a dynamic request scheduling algorithm based on allocation matrix is proposed in this paper. The algorithm fully considers the characteristic of multi-core processors, assigns the same kind of dynamic requests to the same processing core, and improves the speed of handling dynamic requests in web server. The results of the experiment show that the web server that adopts this algorithm is superior to the traditional FCFS algorithm in the aspect of similarity, mean response time and dropped rate of dynamic requests.
2016, 38(9): 2194-2201.
doi: 10.11999/JEIT151374
Abstract:
By reasonably exploiting the collaborative relationships between nodes and the limited cache resources of neighbor nodes, the cache utilization rate and the message delivery probability of opportunistic networks can be effectively improved. A node status aware distributed cooperative cache management mechanism is proposed in the paper, where the importance degree of a given message can be dynamically estimated by perceiving its spreading degree to determine the caching priority. Furthermore, according to the active degree and the relative viscosity, the encounter probability of a given message to its destination node can be obtained for the design of adaptive cooperative cache management strategies for messages in each cache area respectively. When the cache is fully occupied, the collaboration node within the transmission range is dynamically selected, thus the message transferring to the collaboration node and the efficient utilization of caching resources can be ideally achieved. The results show that the proposed caching mechanism can fully exploit the limited cache resources and greatly improve the message delivery rate and the buffer utilization rate.
By reasonably exploiting the collaborative relationships between nodes and the limited cache resources of neighbor nodes, the cache utilization rate and the message delivery probability of opportunistic networks can be effectively improved. A node status aware distributed cooperative cache management mechanism is proposed in the paper, where the importance degree of a given message can be dynamically estimated by perceiving its spreading degree to determine the caching priority. Furthermore, according to the active degree and the relative viscosity, the encounter probability of a given message to its destination node can be obtained for the design of adaptive cooperative cache management strategies for messages in each cache area respectively. When the cache is fully occupied, the collaboration node within the transmission range is dynamically selected, thus the message transferring to the collaboration node and the efficient utilization of caching resources can be ideally achieved. The results show that the proposed caching mechanism can fully exploit the limited cache resources and greatly improve the message delivery rate and the buffer utilization rate.
2016, 38(9): 2202-2207.
doi: 10.11999/JEIT151311
Abstract:
The network performance of WSNs (Wireless Sensor Networks) can be improved significantly by injecting mobile elements. However, the infection process of worm will be greatly accelerated once the mobile element has been captured and become the new infection source. To cope with this new threat, this paper first proposes the infection model for the networks with the mobile worm and designs a heuristic algorithm to identify the boundary of infected area. High risk nodes near the boundary can be found and switched to sleeping states to block the further spreading of the worm. Second, an algorithm with directed-diffusion based anti-worm is designed to repair those infected sensors. Theoretical analysis and experimental results show that the proposed methods can achieve better worm cleaning effect with low cost, which can be applied to energy-limited wireless sensor networks.
The network performance of WSNs (Wireless Sensor Networks) can be improved significantly by injecting mobile elements. However, the infection process of worm will be greatly accelerated once the mobile element has been captured and become the new infection source. To cope with this new threat, this paper first proposes the infection model for the networks with the mobile worm and designs a heuristic algorithm to identify the boundary of infected area. High risk nodes near the boundary can be found and switched to sleeping states to block the further spreading of the worm. Second, an algorithm with directed-diffusion based anti-worm is designed to repair those infected sensors. Theoretical analysis and experimental results show that the proposed methods can achieve better worm cleaning effect with low cost, which can be applied to energy-limited wireless sensor networks.
2016, 38(9): 2208-2214.
doi: 10.11999/JEIT151395
Abstract:
Considering that conventional distributed synchronicity algorithm may lead to a state of partial synchronization separately with small couple coefficient or unsteadiness with large couple coefficient, a multi granularity firefly-inspired synchronicity algorithm is proposed. It lets the phase value couple in multi granularity by its divergence in time and phase, which can relieve the issue of partial synchronization and speed up the convergence process. Its performance is tested by simulation in a non-fully connect network by comparing with the conventional MS algorithm. The result shows that it works better.
Considering that conventional distributed synchronicity algorithm may lead to a state of partial synchronization separately with small couple coefficient or unsteadiness with large couple coefficient, a multi granularity firefly-inspired synchronicity algorithm is proposed. It lets the phase value couple in multi granularity by its divergence in time and phase, which can relieve the issue of partial synchronization and speed up the convergence process. Its performance is tested by simulation in a non-fully connect network by comparing with the conventional MS algorithm. The result shows that it works better.
2016, 38(9): 2215-2220.
doi: 10.11999/JEIT151385
Abstract:
Considering the line-of-sight path interference cancellation in Co-time and Co-frequency Full-Duplex (CCFD), the pre-nulling technique of transceiver is usually used. However, this method forms a subsidence zone of signal power for the remote receiver. The main reason is that the local receiver antenna is not considered in the near field of the transmitter antenna, and the design is based on the plane wave model. In this paper, the spherical wave assumption is adopted, and the local receiver antenna position is optimized by the subsidence capacity maximum while guaranteeing transmitter pre-nulling of self-interference. The procedure of calculating the optimal local receive antenna position is also given. The simulation results show that when the transmitter spacing is half wavelength, the subsidence capacity is improved with the increase of the channel correlation compared with the antennas placement optimization by the plane wave assumption.
Considering the line-of-sight path interference cancellation in Co-time and Co-frequency Full-Duplex (CCFD), the pre-nulling technique of transceiver is usually used. However, this method forms a subsidence zone of signal power for the remote receiver. The main reason is that the local receiver antenna is not considered in the near field of the transmitter antenna, and the design is based on the plane wave model. In this paper, the spherical wave assumption is adopted, and the local receiver antenna position is optimized by the subsidence capacity maximum while guaranteeing transmitter pre-nulling of self-interference. The procedure of calculating the optimal local receive antenna position is also given. The simulation results show that when the transmitter spacing is half wavelength, the subsidence capacity is improved with the increase of the channel correlation compared with the antennas placement optimization by the plane wave assumption.
2016, 38(9): 2221-2226.
doi: 10.11999/JEIT151316
Abstract:
For channel estimation technology of multiple antenna communication system based on cognitive radio architecture, traditional zero correlation zone sequences (which assume the availability of the entire spectral band) can not be used because their orthogonality will be destroyed by the spectrum hole constraint. This paper introduces the channel estimation algorithm in multiple antenna communication system and points out the requirement of the ideal sequence set, then on complex electromagnetic environment, i.e., under the condition of the existence of spectrum holes, joints optimization evaluation indexes of limited spectrum and good cycle correlation degree, designs a suitable sequence set, which can be applied to the channel estimation algorithm of cognitive radio system. The simulation results verify the effectiveness of the new sequence set.
For channel estimation technology of multiple antenna communication system based on cognitive radio architecture, traditional zero correlation zone sequences (which assume the availability of the entire spectral band) can not be used because their orthogonality will be destroyed by the spectrum hole constraint. This paper introduces the channel estimation algorithm in multiple antenna communication system and points out the requirement of the ideal sequence set, then on complex electromagnetic environment, i.e., under the condition of the existence of spectrum holes, joints optimization evaluation indexes of limited spectrum and good cycle correlation degree, designs a suitable sequence set, which can be applied to the channel estimation algorithm of cognitive radio system. The simulation results verify the effectiveness of the new sequence set.
2016, 38(9): 2227-2232.
doi: 10.11999/JEIT151363
Abstract:
Considering the issues of nonlinearity and self-interference in transmitting and receiving channels of a full-duplex MIMO radio respectively, a novel low hardware cost and low software complexity design scheme with transmitting channels linearization and self-interference cancellation by multi-tap RF cancellers and digital cancellers is proposed, where (1) An improved Cross-Talk Cancelling-Digital Pre-Distorter (CTC-DPD) algorithm and common feedback are used for decoupling and digital pre-distortion to make transmitting channels gain linearly and equally; (2) By introducing adjustable attenuators in receiving channels, multi-tap cancellers use received data along with minimum residual self-interference power criterion based multidimensional gradient descent method to search the multi-tap setting; (3) Digital self-interference is reconstructed for cancellation based on channel estimation in frequency domain. In the prototype of 20 MHz bandwidth LTE full-duplex 22 MIMO radio, transmitting channels have more flat in-band spectrum and 30 dB lower out-band noise after linearization. RF and digital self-interference cancellation need 0.17 ms for one turning and provide about 75 dB cancellation together. When two nodes with 16QAM mapping work at full-duplex mode, they achieve a sum of 220 Mbps bit rate, which double the bit rate of 110 Mbps at half-duplex mode and thus double the spectrum efficiency. The prototype demonstrates the feasibility of the proposed design.
Considering the issues of nonlinearity and self-interference in transmitting and receiving channels of a full-duplex MIMO radio respectively, a novel low hardware cost and low software complexity design scheme with transmitting channels linearization and self-interference cancellation by multi-tap RF cancellers and digital cancellers is proposed, where (1) An improved Cross-Talk Cancelling-Digital Pre-Distorter (CTC-DPD) algorithm and common feedback are used for decoupling and digital pre-distortion to make transmitting channels gain linearly and equally; (2) By introducing adjustable attenuators in receiving channels, multi-tap cancellers use received data along with minimum residual self-interference power criterion based multidimensional gradient descent method to search the multi-tap setting; (3) Digital self-interference is reconstructed for cancellation based on channel estimation in frequency domain. In the prototype of 20 MHz bandwidth LTE full-duplex 22 MIMO radio, transmitting channels have more flat in-band spectrum and 30 dB lower out-band noise after linearization. RF and digital self-interference cancellation need 0.17 ms for one turning and provide about 75 dB cancellation together. When two nodes with 16QAM mapping work at full-duplex mode, they achieve a sum of 220 Mbps bit rate, which double the bit rate of 110 Mbps at half-duplex mode and thus double the spectrum efficiency. The prototype demonstrates the feasibility of the proposed design.
2016, 38(9): 2233-2240.
doi: 10.11999/JEIJ151371
Abstract:
A physical layer security transmission protocol in a two-hop relay system is studied. All nodes are equipped with an antenna and have the ability of energy harvesting. There are direct links between the source node and the eavesdropper as well as the relay node and the eavesdropper. Each transmission time slot is divided into two stages, which are respectively used for energy harvesting and data transmitting. The energy harvested is used to send data. The amplify-and-forward protocol is adopted, and the destination node sends artificial noise to protect the information transmitted in the first and second hop. To maximize the secrecy rate, iterative algorithm is used to optimize two variables of the time for energy harvesting and the power of artificial noise. Simulation results show that the optimization algorithm is accurate and the cooperative jamming can effectively improve the secrecy rate. Considering that eavesdropper can intercept the information transmitted in the two hops, the proposed scheme is more practical and can solve a complicate optimization problem.
A physical layer security transmission protocol in a two-hop relay system is studied. All nodes are equipped with an antenna and have the ability of energy harvesting. There are direct links between the source node and the eavesdropper as well as the relay node and the eavesdropper. Each transmission time slot is divided into two stages, which are respectively used for energy harvesting and data transmitting. The energy harvested is used to send data. The amplify-and-forward protocol is adopted, and the destination node sends artificial noise to protect the information transmitted in the first and second hop. To maximize the secrecy rate, iterative algorithm is used to optimize two variables of the time for energy harvesting and the power of artificial noise. Simulation results show that the optimization algorithm is accurate and the cooperative jamming can effectively improve the secrecy rate. Considering that eavesdropper can intercept the information transmitted in the two hops, the proposed scheme is more practical and can solve a complicate optimization problem.
2016, 38(9): 2241-2247.
doi: 10.11999/JEIT160103
Abstract:
Interference alignment has the advantage of eliminating interference, but it needs the perfect Channel State Information (CSI) which is difficult to achieve in practical systems. The advantages and disadvantages of robust interference alignment algorithms are analyzed in this paper. And then a robust interference alignment algorithm based on maximizing the Rayleigh entropy is proposed. The convergence, the degree of freedom and spectrum efficiency are analyzed at the same time. Unlike MAX-SINR algorithm, interference suppression matrix is obtained through maximizing the signal Rayleigh entropy. The unitary form of original interference suppression matrix is regarded as the optimal interference suppression matrix considering the correlation among the data flows. And then, the water-filling power allocation scheme is used to realize the optimal power allocation among user data flows. Meanwhile, the similar process is carried out in reverse communication link based on channel reciprocity. The interference is reduced gradually through alternately computing. Finally, under the conditions of perfect CSI and error CSI, the simulation results verify that the proposed algorithm improves the performance of the system.
Interference alignment has the advantage of eliminating interference, but it needs the perfect Channel State Information (CSI) which is difficult to achieve in practical systems. The advantages and disadvantages of robust interference alignment algorithms are analyzed in this paper. And then a robust interference alignment algorithm based on maximizing the Rayleigh entropy is proposed. The convergence, the degree of freedom and spectrum efficiency are analyzed at the same time. Unlike MAX-SINR algorithm, interference suppression matrix is obtained through maximizing the signal Rayleigh entropy. The unitary form of original interference suppression matrix is regarded as the optimal interference suppression matrix considering the correlation among the data flows. And then, the water-filling power allocation scheme is used to realize the optimal power allocation among user data flows. Meanwhile, the similar process is carried out in reverse communication link based on channel reciprocity. The interference is reduced gradually through alternately computing. Finally, under the conditions of perfect CSI and error CSI, the simulation results verify that the proposed algorithm improves the performance of the system.
2016, 38(9): 2248-2254.
doi: 10.11999/JEIT151286
Abstract:
A transmission mechanism based on the joint management of time, frequency and space domain resources is proposed in the Cooperative Cognitive Radio Network (CCRN), with which the cooperative cognitive users can obtain proper reward under the premise of improving the primary users transmission. The proposed scheme employs multi-antenna cognitive user as a relay. Via adaptive time-slot and bandwidth allocation, bottleneck in the two-hop transmission can be eliminated. Furthermore, a relay selection algorithm taking fairness into account is given for the situation where multiple cooperative cognitive users exist. On one hand, the relay selection utilizes proportional fair to achieve inter-cognitive-user fairness. On the other hand, proper reward is determined by adjusting the incentive factor. Simulation results show that the proposed mechanism can improve the data rate of both primary and cognitive systems, and afford fair reward to the cognitive relays.
A transmission mechanism based on the joint management of time, frequency and space domain resources is proposed in the Cooperative Cognitive Radio Network (CCRN), with which the cooperative cognitive users can obtain proper reward under the premise of improving the primary users transmission. The proposed scheme employs multi-antenna cognitive user as a relay. Via adaptive time-slot and bandwidth allocation, bottleneck in the two-hop transmission can be eliminated. Furthermore, a relay selection algorithm taking fairness into account is given for the situation where multiple cooperative cognitive users exist. On one hand, the relay selection utilizes proportional fair to achieve inter-cognitive-user fairness. On the other hand, proper reward is determined by adjusting the incentive factor. Simulation results show that the proposed mechanism can improve the data rate of both primary and cognitive systems, and afford fair reward to the cognitive relays.
2016, 38(9): 2255-2260.
doi: 10.11999/JEIT151307
Abstract:
This paper deals with blind equalization of sparse multipath channels. A linear model is built under the framework of Reduced Constellation Algorithm (RCA). And the inherent sparse nature of the equalizer is exploited by employing a sparse promoting prior distribution. Then, the sparse Bayesian learning iterative inference method is applied to the proposed model in order to obtain the optimal sparse equalizer. The new proposed algorithm, which belongs to data recycling equalization algorithm domain, can be applied to short packet data applications. Compared with traditional Stochastic Gradient Descent (SGD) method, the new proposed algorithm performs more steadily under different equalizer order and has superior steady-state Symbol-Error-Rate (SER) performance. The effectiveness of the proposed algorithm is verified by simulations.
This paper deals with blind equalization of sparse multipath channels. A linear model is built under the framework of Reduced Constellation Algorithm (RCA). And the inherent sparse nature of the equalizer is exploited by employing a sparse promoting prior distribution. Then, the sparse Bayesian learning iterative inference method is applied to the proposed model in order to obtain the optimal sparse equalizer. The new proposed algorithm, which belongs to data recycling equalization algorithm domain, can be applied to short packet data applications. Compared with traditional Stochastic Gradient Descent (SGD) method, the new proposed algorithm performs more steadily under different equalizer order and has superior steady-state Symbol-Error-Rate (SER) performance. The effectiveness of the proposed algorithm is verified by simulations.
2016, 38(9): 2261-2267.
doi: 10.11999/JEIT151384
Abstract:
Multipath provisioning algorithm outperforms single-path provisioning algorithm in terms of bandwidth blocking probability. However, multipath transmission causes the differential delay among different paths and affects the usage of spectrum resources. To address the problem, a Genetic Multipath Protection Algorithm (GMPA) is proposed based on multi-objectives genetic algorithm. according to traffic requests, the K link-disjoined paths and bandwidth assignments are designed as the population initialization scheme. A vector function is proposed to balance the path-distance difference and network spectrum resources by optimizing population classification and crowding distance sorting. An individual self-cross pattern is introduced and the variation range and constraint conditions of bandwidth gene are designed to improve the algorithm search ability and convergence. Compared with the Multiple Path Protection (MPP) and Primary First-fit Modified Backup Last-fit (PF-MBL), simulation results show that the proposed GMPA algorithm can get lowest bandwidth blocking probability, its spectrum resource utilization is close to the optimal MPP, and the path-distance difference of GMPA is better than that of MPP.
Multipath provisioning algorithm outperforms single-path provisioning algorithm in terms of bandwidth blocking probability. However, multipath transmission causes the differential delay among different paths and affects the usage of spectrum resources. To address the problem, a Genetic Multipath Protection Algorithm (GMPA) is proposed based on multi-objectives genetic algorithm. according to traffic requests, the K link-disjoined paths and bandwidth assignments are designed as the population initialization scheme. A vector function is proposed to balance the path-distance difference and network spectrum resources by optimizing population classification and crowding distance sorting. An individual self-cross pattern is introduced and the variation range and constraint conditions of bandwidth gene are designed to improve the algorithm search ability and convergence. Compared with the Multiple Path Protection (MPP) and Primary First-fit Modified Backup Last-fit (PF-MBL), simulation results show that the proposed GMPA algorithm can get lowest bandwidth blocking probability, its spectrum resource utilization is close to the optimal MPP, and the path-distance difference of GMPA is better than that of MPP.
2016, 38(9): 2268-2273.
doi: 10.11999/JEIT151362
Abstract:
Low-density parity-check code is the one of error-correction codes most approaching Shannon limit, which is adopted as a standard for channel coding by many international communication standard organizations. CCSDS recommends LDPC as channel coding scheme in near earth space and deep space communication. An efficient LDPC coding scheme with low power and low parallel is presented in this paper. By filling 0 and changing the cyclic-matrix structure, the proposed scheme implements a low parallel coding for the LDPC, which is recommended by CCSDS, and of which the size of submatrix of check matrices is odd. By analyzing the coding process, the valid bit 1 among input information bits is coded only, and it decreases obviously the code power. The encoder architecture for 7/8 LDPC is implemented in FPGA. The result shows that encoder achieves a high encoding speed approaching low parallel encoder scheme and a much lower encoding power while increases few hardware overhead.
Low-density parity-check code is the one of error-correction codes most approaching Shannon limit, which is adopted as a standard for channel coding by many international communication standard organizations. CCSDS recommends LDPC as channel coding scheme in near earth space and deep space communication. An efficient LDPC coding scheme with low power and low parallel is presented in this paper. By filling 0 and changing the cyclic-matrix structure, the proposed scheme implements a low parallel coding for the LDPC, which is recommended by CCSDS, and of which the size of submatrix of check matrices is odd. By analyzing the coding process, the valid bit 1 among input information bits is coded only, and it decreases obviously the code power. The encoder architecture for 7/8 LDPC is implemented in FPGA. The result shows that encoder achieves a high encoding speed approaching low parallel encoder scheme and a much lower encoding power while increases few hardware overhead.
2016, 38(9): 2274-2279.
doi: 10.11999/JEIT151376
Abstract:
In this paper, a new ensemble of the polynomial matrix of a time-invariant LDPC convolutional code is proposed. Based on the method of deriving time-invariant LDPC convolutional codes from QC (Quasi-Cyclic)- LDPC block codes, the elements over finite fields are used to generate directly the polynomial parity-check matrices of LDPC convolutional codes. A concrete example of using MDS (Maximum-Distance Separable) convolutional codes to derive the polynomial matrices is given. The proposed method ensures the fast encoding property, maximum encoding memory and designed rate. Simulation results show that the proposed LDPC convolutional codes exhibit low error floor and good decoding performance under BP (Belief Propagation) decoding algorithm over AWGN (Additive White Gaussian Noise) channel.
In this paper, a new ensemble of the polynomial matrix of a time-invariant LDPC convolutional code is proposed. Based on the method of deriving time-invariant LDPC convolutional codes from QC (Quasi-Cyclic)- LDPC block codes, the elements over finite fields are used to generate directly the polynomial parity-check matrices of LDPC convolutional codes. A concrete example of using MDS (Maximum-Distance Separable) convolutional codes to derive the polynomial matrices is given. The proposed method ensures the fast encoding property, maximum encoding memory and designed rate. Simulation results show that the proposed LDPC convolutional codes exhibit low error floor and good decoding performance under BP (Belief Propagation) decoding algorithm over AWGN (Additive White Gaussian Noise) channel.
2016, 38(9): 2280-2286.
doi: 10.11999/JEIT151164
Abstract:
All the existing provable secure proactive threshold RSA signature schemes rely on additive secret sharing, in which all players have to cooperate to produce a signature, valid players secret shares may be exposed, and the computing efficiency is too low. Based on Shoups threshold RSA signature scheme, a proactive threshold RSA signature scheme is proposed by using polynomial secret sharing, and its security and practicability are analyzed. Results show that the proposed scheme is unforgeable and robust under the model of static mobile adversary, and compared with the existing comparable schemes, its communication overhead is lower and computing efficiency is higher.
All the existing provable secure proactive threshold RSA signature schemes rely on additive secret sharing, in which all players have to cooperate to produce a signature, valid players secret shares may be exposed, and the computing efficiency is too low. Based on Shoups threshold RSA signature scheme, a proactive threshold RSA signature scheme is proposed by using polynomial secret sharing, and its security and practicability are analyzed. Results show that the proposed scheme is unforgeable and robust under the model of static mobile adversary, and compared with the existing comparable schemes, its communication overhead is lower and computing efficiency is higher.
2016, 38(9): 2287-2293.
doi: 10.11999/JEIT151044
Abstract:
The existing lattice-based signcryption schemes are based on trapdoor generation algorithm and preimage sample algorithm. However, both algorithms are complex, require a lot of time to run, and affect the efficiency of latticed-based signcryption schemes deeply. To solve this problem, the first lattice-based signcryption scheme without trapdoor generation algorithm and preimage sample algorithm is proposed, with the help of the technique of lattice signatures without trapdoors and the associated signature compression technique, as well as the encryption method based on the learning with errors assumption. The scheme achieves indistinguishability against adaptive chosen ciphertext attacks under the learning with errors assumption. It also achieves existential unforgeability against adaptive chosen message attacks under the small integer solution assumption. The proposed scheme is not only quantum resistant, but also efficient.
The existing lattice-based signcryption schemes are based on trapdoor generation algorithm and preimage sample algorithm. However, both algorithms are complex, require a lot of time to run, and affect the efficiency of latticed-based signcryption schemes deeply. To solve this problem, the first lattice-based signcryption scheme without trapdoor generation algorithm and preimage sample algorithm is proposed, with the help of the technique of lattice signatures without trapdoors and the associated signature compression technique, as well as the encryption method based on the learning with errors assumption. The scheme achieves indistinguishability against adaptive chosen ciphertext attacks under the learning with errors assumption. It also achieves existential unforgeability against adaptive chosen message attacks under the small integer solution assumption. The proposed scheme is not only quantum resistant, but also efficient.
2016, 38(9): 2294-2300.
doi: 10.11999/JEIT160102
Abstract:
Privacy-preserving determination of spatial relationship belongs to spatial geometry problem in secure multiparty computation, which is significant to confidential business, engineering, military, etc. However, most existing schemes transform the original problem into the distance problem or the correspondingly proportional data problem, which makes the computation complexity high and the application range being limited. To deal with these problems, first, the original problem is transformed into whether a point is the solution of equation. Based on the technique, a simple and efficient scalar product protocol is adopted to determine five spatial relationships all at once: point and line, point and plane, line and line, line and plane, and plane and plane. In addition, the security of the proposed protocol is proved with simulation paradigm. The proposed scheme does not employ any public key encryption algorithm so as to achieve the information security. The analysis indicates the trick transformation makes the proposed scheme higher efficient and more applicable than the known schemes.
Privacy-preserving determination of spatial relationship belongs to spatial geometry problem in secure multiparty computation, which is significant to confidential business, engineering, military, etc. However, most existing schemes transform the original problem into the distance problem or the correspondingly proportional data problem, which makes the computation complexity high and the application range being limited. To deal with these problems, first, the original problem is transformed into whether a point is the solution of equation. Based on the technique, a simple and efficient scalar product protocol is adopted to determine five spatial relationships all at once: point and line, point and plane, line and line, line and plane, and plane and plane. In addition, the security of the proposed protocol is proved with simulation paradigm. The proposed scheme does not employ any public key encryption algorithm so as to achieve the information security. The analysis indicates the trick transformation makes the proposed scheme higher efficient and more applicable than the known schemes.
ISAR Imaging of Targets with Complex Motion Based on the Modified Fast Bilinear Parameter Estimation
2016, 38(9): 2301-2308.
doi: 10.11999/JEIT151359
Abstract:
In view of image defocus caused by Doppler frequency shift in ISAR imaging of targets with complex motion, this paper characterizes the azimuth echoes as Cubic Phase Signal (CPS) and proposes an ISAR imaging algorithm for targets with complex motion based on the modified fast bilinear parameter estimation. This algorithm can achieve parameter estimation of CPS and ISAR imaging quickly by employing the cubic phase bilinear function, NonUniform Fast Fourier Transform (NUFFT), scale transform based on Chirp-z transform and Fast Fourier Transform (FFT). The computational cost is low due to the NUFFT and FFT in implementation procedure, and bilinearity guarantees a high anti-noise performance and a good suppression on cross-terms. The theoretical analysis and simulation results demonstrate the effectiveness of the proposed ISAR imaging algorithm.
In view of image defocus caused by Doppler frequency shift in ISAR imaging of targets with complex motion, this paper characterizes the azimuth echoes as Cubic Phase Signal (CPS) and proposes an ISAR imaging algorithm for targets with complex motion based on the modified fast bilinear parameter estimation. This algorithm can achieve parameter estimation of CPS and ISAR imaging quickly by employing the cubic phase bilinear function, NonUniform Fast Fourier Transform (NUFFT), scale transform based on Chirp-z transform and Fast Fourier Transform (FFT). The computational cost is low due to the NUFFT and FFT in implementation procedure, and bilinearity guarantees a high anti-noise performance and a good suppression on cross-terms. The theoretical analysis and simulation results demonstrate the effectiveness of the proposed ISAR imaging algorithm.
2016, 38(9): 2309-2313.
doi: 10.11999/JEIT151319
Abstract:
In order to improve the accuracy of low-angle estimation in meter-wave radars, combined with sparse Bayesian learning, this paper makes use of the Kronecker product and the similarity of the sparse structure between adjacent snapshots to transform the multiple measurement vector model into a single measurement vector model. The angle of the source is obtained by the coefficient matrix of the sensing matrix related to signal and the coefficient matrix is recovered by the continuous iteration in sparse Bayesian learning. Simulation experiments show that the proposed method has better performance than the generalized MUSIC algorithm and M-FOCUSS algorithm, the influence on algorithm performance caused by the snapshot change is obtained.
In order to improve the accuracy of low-angle estimation in meter-wave radars, combined with sparse Bayesian learning, this paper makes use of the Kronecker product and the similarity of the sparse structure between adjacent snapshots to transform the multiple measurement vector model into a single measurement vector model. The angle of the source is obtained by the coefficient matrix of the sensing matrix related to signal and the coefficient matrix is recovered by the continuous iteration in sparse Bayesian learning. Simulation experiments show that the proposed method has better performance than the generalized MUSIC algorithm and M-FOCUSS algorithm, the influence on algorithm performance caused by the snapshot change is obtained.
2016, 38(9): 2314-2320.
doi: 10.11999/JEIT151280
Abstract:
The Local Linear Regression (LLR) estimator is usually used when developing a nonparametric model for radar altimeter Sea State Bias (SSB). However, the conventional LLR estimator contains matrices with high dimension. When a large number of data are used to develop the SSB model, the SSB estimation costs too much time. Therefore, the nonparametric estimation method can hardly be used to develop high-dimensional SSB models. This paper presents an Improved LLR (ILLR) estimator, complexity fromO(N2)toO(N) which can avoid high-dimensional matrix operations. The improved LLR estimator can greatly reduce the time for SSB estimation without affecting the estimated accuracy. So the improved LLR estimator can laid the foundation for high-dimensional SSB models.
The Local Linear Regression (LLR) estimator is usually used when developing a nonparametric model for radar altimeter Sea State Bias (SSB). However, the conventional LLR estimator contains matrices with high dimension. When a large number of data are used to develop the SSB model, the SSB estimation costs too much time. Therefore, the nonparametric estimation method can hardly be used to develop high-dimensional SSB models. This paper presents an Improved LLR (ILLR) estimator, complexity fromO(N2)toO(N) which can avoid high-dimensional matrix operations. The improved LLR estimator can greatly reduce the time for SSB estimation without affecting the estimated accuracy. So the improved LLR estimator can laid the foundation for high-dimensional SSB models.
2016, 38(9): 2321-2329.
doi: 10.11999/JEIT151340
Abstract:
Hyperspectral image contains the properties of much bands and high redundancy, and the research of hyperspectral image classification focuses on feature extraction. To overcome this problem, a Semi-Supervised Sparse Manifold Embedding (S3ME) algorithm is proposed in this paper. The S3ME method makes full use of labeled and unlabeled samples to adaptively reveal the similarity relationship between data with the sparse representation of tangent space. It constructs a semi-supervised similarity graph via the sparse coefficients and enhances the weight between labeled samples from the same class. In a low-dimensional embedding space, S3ME preserves the similarity of graph to minimize the sum of the weighted distance. Then, it obtains a projection matrix for feature extraction. S3ME not only reveals the sparse manifold structure of data but also enhances the compactness of the same class data, which can effectively extract the discriminating feature and improve the classification performance. The overall classification accuracies of the proposed S3ME method respectively reach 84.62% and 88.07% on the PaviaU and Salinas hyperspectral data sets, and the classification performance of land cover is improved compared with the traditional feature extraction methods.
Hyperspectral image contains the properties of much bands and high redundancy, and the research of hyperspectral image classification focuses on feature extraction. To overcome this problem, a Semi-Supervised Sparse Manifold Embedding (S3ME) algorithm is proposed in this paper. The S3ME method makes full use of labeled and unlabeled samples to adaptively reveal the similarity relationship between data with the sparse representation of tangent space. It constructs a semi-supervised similarity graph via the sparse coefficients and enhances the weight between labeled samples from the same class. In a low-dimensional embedding space, S3ME preserves the similarity of graph to minimize the sum of the weighted distance. Then, it obtains a projection matrix for feature extraction. S3ME not only reveals the sparse manifold structure of data but also enhances the compactness of the same class data, which can effectively extract the discriminating feature and improve the classification performance. The overall classification accuracies of the proposed S3ME method respectively reach 84.62% and 88.07% on the PaviaU and Salinas hyperspectral data sets, and the classification performance of land cover is improved compared with the traditional feature extraction methods.
2016, 38(9): 2330-2335.
doi: 10.11999/JEIT151315
Abstract:
The association between multiple targets and echo data is the main problem for multi-target location method, which causes huge calculation and the problem of extracting the targets accurately. The location method based on Bistatic Range Space Projection (BRSP) can be used for the sake of overcoming the data association problem. While there are two problems existing on the location method of BRSP, the huge calculation and the low resolution. In the face of vast calculation in projection imaging localization, this paper utilizes hierarchical strategy to decrease calculation. The possible targets areas are located with low resolution at first. After that, more precision probable areas are pinpointed via higher resolution from these possible areas. In this way, the calculation of areas without targets can be avoided. Furthermore, results of hierarchical processing are used to be the initial position guess for Taylor-series estimation. Positioning errors could be modified by the iterative correction of Taylor-series estimation. Simulation results indicate a significant improvement in the running time and positioning precision of the proposed method.
The association between multiple targets and echo data is the main problem for multi-target location method, which causes huge calculation and the problem of extracting the targets accurately. The location method based on Bistatic Range Space Projection (BRSP) can be used for the sake of overcoming the data association problem. While there are two problems existing on the location method of BRSP, the huge calculation and the low resolution. In the face of vast calculation in projection imaging localization, this paper utilizes hierarchical strategy to decrease calculation. The possible targets areas are located with low resolution at first. After that, more precision probable areas are pinpointed via higher resolution from these possible areas. In this way, the calculation of areas without targets can be avoided. Furthermore, results of hierarchical processing are used to be the initial position guess for Taylor-series estimation. Positioning errors could be modified by the iterative correction of Taylor-series estimation. Simulation results indicate a significant improvement in the running time and positioning precision of the proposed method.
2016, 38(9): 2336-2343.
doi: 10.11999/JEIT151379
Abstract:
To solve the single-observer passive location estimation using illuminators of opportunity, a jointing Direction Of Arrival (DOA) and Time Difference Of Arrival (TDOA) location method based on Regularized Constrained Total Least Squares (RCTLS) algorithm is proposed. Firstly, the DOA and TDOA measurement equations are linearized. Considering the errors in the location equations, the localization problem is established as a RCTLS model. Then the Newtons method is applied to solving the RCTLS model to obtain the target position. The theoretical error of the proposed algorithm is derived and an optimal regularization parameter is chosen by the least mean square error rule. Simulation results show that the proposed RCTLS algorithm has lower mean squares error than the Constrained Total Least Squares (CTLS) algorithm. Moreover, from the Geometric Dilution Of Precision (GDOP) figure, it can be concluded that positions of the target and illuminators are also important factors affecting the localization accuracy.
To solve the single-observer passive location estimation using illuminators of opportunity, a jointing Direction Of Arrival (DOA) and Time Difference Of Arrival (TDOA) location method based on Regularized Constrained Total Least Squares (RCTLS) algorithm is proposed. Firstly, the DOA and TDOA measurement equations are linearized. Considering the errors in the location equations, the localization problem is established as a RCTLS model. Then the Newtons method is applied to solving the RCTLS model to obtain the target position. The theoretical error of the proposed algorithm is derived and an optimal regularization parameter is chosen by the least mean square error rule. Simulation results show that the proposed RCTLS algorithm has lower mean squares error than the Constrained Total Least Squares (CTLS) algorithm. Moreover, from the Geometric Dilution Of Precision (GDOP) figure, it can be concluded that positions of the target and illuminators are also important factors affecting the localization accuracy.
2016, 38(9): 2344-2350.
doi: 10.11999/JEIT151295
Abstract:
Jamming and spoofing are the most common and serious threat intentional interferences for Global Navigation Satellite?System (GNSS). A joint GNSS interference mitigation approach is proposed for jamming and spoofing issue based on multi-antenna array in this paper.?Firstly, jamming is suppressed by the subspace technology. Then, spoofing is detected and mitigated by using the weighting vectors obtained by the despread- respread method.?Finally, for maximizing the output of each authentic signal, the despread-respread method is reutilized to form multiple beams pointing at each of the authentic satellites.? Simulation results show that the proposed algorithm can null jamming and spoofing simultaneously. The interference suppression performance of this method is not sensitive to the array manifold?errors and not limited by the prior knowledge of satellites directions.
Jamming and spoofing are the most common and serious threat intentional interferences for Global Navigation Satellite?System (GNSS). A joint GNSS interference mitigation approach is proposed for jamming and spoofing issue based on multi-antenna array in this paper.?Firstly, jamming is suppressed by the subspace technology. Then, spoofing is detected and mitigated by using the weighting vectors obtained by the despread- respread method.?Finally, for maximizing the output of each authentic signal, the despread-respread method is reutilized to form multiple beams pointing at each of the authentic satellites.? Simulation results show that the proposed algorithm can null jamming and spoofing simultaneously. The interference suppression performance of this method is not sensitive to the array manifold?errors and not limited by the prior knowledge of satellites directions.
2016, 38(9): 2351-2357.
doi: 10.11999/JEIT151377
Abstract:
In this work, considering a radar-communication integrated radio frequency system, a dynamic allocation method of shared aperture using relevant environmental information is proposed. Firstly, the shared aperture allocation task is formulated as a Multi-Objective Optimization (MOO) problem based on Pareto optimality, which uses the peak side-lobe level of radar array pattern and the channel capacity of Multiple Input Multiple Output (MIMO) communication system as its objective function. Then, an improved particle swarm optimization algorithm based on integer encoding is proposed to solve the MOO problem. The iterative algorithm can find out a set of optimal solutions in the form of Pareto front, one of which can be chosen by decision makers as the most satisfactory solution according to mission requirements. Finally, the simulation results verify the effectiveness of the proposed method.
In this work, considering a radar-communication integrated radio frequency system, a dynamic allocation method of shared aperture using relevant environmental information is proposed. Firstly, the shared aperture allocation task is formulated as a Multi-Objective Optimization (MOO) problem based on Pareto optimality, which uses the peak side-lobe level of radar array pattern and the channel capacity of Multiple Input Multiple Output (MIMO) communication system as its objective function. Then, an improved particle swarm optimization algorithm based on integer encoding is proposed to solve the MOO problem. The iterative algorithm can find out a set of optimal solutions in the form of Pareto front, one of which can be chosen by decision makers as the most satisfactory solution according to mission requirements. Finally, the simulation results verify the effectiveness of the proposed method.
2016, 38(9): 2358-2364.
doi: 10.11999/JEIT151365
Abstract:
The accuracy of lung segmented results is important in actual clinical application. However, all kinds of segmentation methods can not be uniform for all the chest CT (Computed Tomography) images because of the irregularities and diversity of lung disease, as well as significant differences in the anatomy of the human chest. Lung parenchyma segmentation studies still have a great challenge. Based on the analysis of domestic and international research, a new lung segmentation method is presented by combining with 3D region growing method and improved convex hull patching algorithm. Firstly, the 3D region growing method is adopted for the rough segmentation of lung CT images. Then the refining work is done to the segmented results. The connected domain labeling and morphological methods are used to remove the trachea and main bronchi to get the pulmonary parenchyma mask. The improved convex hull algorithm is presented to repair and smooth the concavities of lung contour. Finally, the segmented results can be gotten. The improved convex hull algorithm can repair the concavities of lung contour effectively in comparison with the convex hull algorithm and the rolling ball method, and the segmentation precision of results is very high after repairing.
The accuracy of lung segmented results is important in actual clinical application. However, all kinds of segmentation methods can not be uniform for all the chest CT (Computed Tomography) images because of the irregularities and diversity of lung disease, as well as significant differences in the anatomy of the human chest. Lung parenchyma segmentation studies still have a great challenge. Based on the analysis of domestic and international research, a new lung segmentation method is presented by combining with 3D region growing method and improved convex hull patching algorithm. Firstly, the 3D region growing method is adopted for the rough segmentation of lung CT images. Then the refining work is done to the segmented results. The connected domain labeling and morphological methods are used to remove the trachea and main bronchi to get the pulmonary parenchyma mask. The improved convex hull algorithm is presented to repair and smooth the concavities of lung contour. Finally, the segmented results can be gotten. The improved convex hull algorithm can repair the concavities of lung contour effectively in comparison with the convex hull algorithm and the rolling ball method, and the segmentation precision of results is very high after repairing.
2016, 38(9): 2365-2370.
doi: 10.11999/JEIT151430
Abstract:
Traditionally, low dimensional features for partial information are extracted to analyze image quality. Though high dimensional features are difficult to be analyzed, they contain more information to fully analyze image quality. On this condition, this paper proposes an image quality assessment method based on non-local high dimensional feature analysis after optimized data sampling. Firstly, image data is filtered by using block matching method and dimensionally reduced by Principal Component Analysis (PCA). Secondly, Kernel Independent Component Analysis (KICA) is applied to extract high dimensional features. The features are finally synthesized to evaluate image quality based on natural image statistics. The experimental results show that the proposed method keeps accordance with human objective perception.
Traditionally, low dimensional features for partial information are extracted to analyze image quality. Though high dimensional features are difficult to be analyzed, they contain more information to fully analyze image quality. On this condition, this paper proposes an image quality assessment method based on non-local high dimensional feature analysis after optimized data sampling. Firstly, image data is filtered by using block matching method and dimensionally reduced by Principal Component Analysis (PCA). Secondly, Kernel Independent Component Analysis (KICA) is applied to extract high dimensional features. The features are finally synthesized to evaluate image quality based on natural image statistics. The experimental results show that the proposed method keeps accordance with human objective perception.
2016, 38(9): 2371-2376.
doi: 10.11999/JEIT151398
Abstract:
Many of the high-resolution Direction Of Arrival (DOA) estimation methods fail to estimate the DOAs of incoming signals and distinguish the two close signals under small sample support; especially only one snapshot is available. To handle this problem, a novel DOA method based on improved Direct Data Domain (D3) technique is proposed. The basic ideal of the proposed method is to augment the time domain samples by sacrificing the space domain degree. Firstly, by splitting the entire array into many overlapping sub-array, many low-dimension samples can be obtained. Secondly, utilizing the property of far-filed narrowband signal that its amplitude response on each array is the same, a new constraint is imposed. Experimental results indicate that the resolution of the propsed method is superior to the conventional algorithms.
Many of the high-resolution Direction Of Arrival (DOA) estimation methods fail to estimate the DOAs of incoming signals and distinguish the two close signals under small sample support; especially only one snapshot is available. To handle this problem, a novel DOA method based on improved Direct Data Domain (D3) technique is proposed. The basic ideal of the proposed method is to augment the time domain samples by sacrificing the space domain degree. Firstly, by splitting the entire array into many overlapping sub-array, many low-dimension samples can be obtained. Secondly, utilizing the property of far-filed narrowband signal that its amplitude response on each array is the same, a new constraint is imposed. Experimental results indicate that the resolution of the propsed method is superior to the conventional algorithms.
2016, 38(9): 2377-2384.
doi: 10.11999/JEIT151170
Abstract:
In order to extract Frequency-Hopping (FH) communication parameters and provide the necessary information for the communication countermeasure, an online estimation algorithm of 2D-DOA and frequency tracking for multiple FH signals is proposed in this paper. Firstly, the data model of the L-array for FH signals is built and the applicability of Auto Regresive Moving Average (ARMA) model to L-array data is proved. Then, the particle filtering is introduced to conduct the online estimation of manifold matrix and the frequency, and the ARMA model is built based on the frequency estimates, depending on which, the online detection of hop timing is obtained. After that , the precise estimation of 2D-DOA can be gained via manifold matrix estimates and without parameter matching. With the rational method of particle generation and the weight updating, the new method makes the estimates of manifold matrix and the frequency reach to the stable value promptly. Finally the the Monte-Carlo simulation results show the effectiveness of the proposed algorithm.
In order to extract Frequency-Hopping (FH) communication parameters and provide the necessary information for the communication countermeasure, an online estimation algorithm of 2D-DOA and frequency tracking for multiple FH signals is proposed in this paper. Firstly, the data model of the L-array for FH signals is built and the applicability of Auto Regresive Moving Average (ARMA) model to L-array data is proved. Then, the particle filtering is introduced to conduct the online estimation of manifold matrix and the frequency, and the ARMA model is built based on the frequency estimates, depending on which, the online detection of hop timing is obtained. After that , the precise estimation of 2D-DOA can be gained via manifold matrix estimates and without parameter matching. With the rational method of particle generation and the weight updating, the new method makes the estimates of manifold matrix and the frequency reach to the stable value promptly. Finally the the Monte-Carlo simulation results show the effectiveness of the proposed algorithm.
2016, 38(9): 2385-2390.
doi: 10.11999/JEIT151260
Abstract:
This paper proposes a new algorithm to design near-perfect-reconstruction nonuniform Cosine modulated filter banks. Due to the infeasibility of directly controlling the performance of Nonuniform Filter Banks (NUFBs) in the existing combined algorithms, the design problem boils down to an unconstrained optimization problem with respect to the Prototype Filter (PF), which minimizes a weighted sum of the transfer function distortion of the NUFBs and the stopband energy of the PF. The optimization problem can be efficiently solved by utilizing linearizing iterative approach. The theoretical analysis and numerical experiments are carried out to show that compared with the existing methods, the proposed method can lead to NUFBs with better overall performance.
This paper proposes a new algorithm to design near-perfect-reconstruction nonuniform Cosine modulated filter banks. Due to the infeasibility of directly controlling the performance of Nonuniform Filter Banks (NUFBs) in the existing combined algorithms, the design problem boils down to an unconstrained optimization problem with respect to the Prototype Filter (PF), which minimizes a weighted sum of the transfer function distortion of the NUFBs and the stopband energy of the PF. The optimization problem can be efficiently solved by utilizing linearizing iterative approach. The theoretical analysis and numerical experiments are carried out to show that compared with the existing methods, the proposed method can lead to NUFBs with better overall performance.
2016, 38(9): 2391-2396.
doi: 10.11999/JEIT151312
Abstract:
In this paper, a Glitch-PUF circuit technique is proposed by taking into consideration various aspects i.e. the signal transmission theory, race and hazard phenomenon, and Physical Unclonable Functions (PUF). First and foremost, the glitch circuit is obtained under the signal transmission theory. Using the combinational logic circuit propagation delay difference which causes 1-hazard and 0-hazard, this feature is used to form output glitch waveform. This glitch is sampled by multistage delay sampling circuit. Due to the nonlinear characteristics of the Glitch, Glitch-PUF can thwart the modeling attack. Finally, under the TSMC 65 nm CMOS technology, a 128-bit output data Glitch-PUF circuit is designed. Monte Carlo simulation results show that the Glitch PUF circuit has better randomness.
In this paper, a Glitch-PUF circuit technique is proposed by taking into consideration various aspects i.e. the signal transmission theory, race and hazard phenomenon, and Physical Unclonable Functions (PUF). First and foremost, the glitch circuit is obtained under the signal transmission theory. Using the combinational logic circuit propagation delay difference which causes 1-hazard and 0-hazard, this feature is used to form output glitch waveform. This glitch is sampled by multistage delay sampling circuit. Due to the nonlinear characteristics of the Glitch, Glitch-PUF can thwart the modeling attack. Finally, under the TSMC 65 nm CMOS technology, a 128-bit output data Glitch-PUF circuit is designed. Monte Carlo simulation results show that the Glitch PUF circuit has better randomness.
2016, 38(9): 2397-2404.
doi: 10.11999/JEIT151216
Abstract:
In order to break through the bottleneck of the huge cluster area in AIC (And-Inverter Cone) architecture based FPGA, the research on the optimisation of the input crossbar architecture is carried on. A post-pack netlist statistics method is creatively proposed to analyze the utilization of AIC cluster inputs and feedbacks and to guide the input crossbar design. And on the architecture parameter design level, it is firstly proposed to separately design the connective probability of the AIC cluster inputs and feedbacks. Through substantial experiments, optimum connective probability combination is derived. From the circuit implement view, dual-phases multiplexer input crossbar is presented according to the characteristics of AIC. The area of the AIC cluster, optimized through the proposed approach, achieves 21.21% smaller than the original one, the huge area problem is markedly ameliorated. When implementing the MCNC and VTR benchmarks, compared to Stratix IV, LUT based FPGA from Altera, the area-delay product of the AIC FPGA after optimisation is reduced by 48.49% and 26.29%, respectively. Compared to the original AIC-based FPGA architecture, the area-delay product is reduced by 28.48% and 28.37%, respectively.
In order to break through the bottleneck of the huge cluster area in AIC (And-Inverter Cone) architecture based FPGA, the research on the optimisation of the input crossbar architecture is carried on. A post-pack netlist statistics method is creatively proposed to analyze the utilization of AIC cluster inputs and feedbacks and to guide the input crossbar design. And on the architecture parameter design level, it is firstly proposed to separately design the connective probability of the AIC cluster inputs and feedbacks. Through substantial experiments, optimum connective probability combination is derived. From the circuit implement view, dual-phases multiplexer input crossbar is presented according to the characteristics of AIC. The area of the AIC cluster, optimized through the proposed approach, achieves 21.21% smaller than the original one, the huge area problem is markedly ameliorated. When implementing the MCNC and VTR benchmarks, compared to Stratix IV, LUT based FPGA from Altera, the area-delay product of the AIC FPGA after optimisation is reduced by 48.49% and 26.29%, respectively. Compared to the original AIC-based FPGA architecture, the area-delay product is reduced by 28.48% and 28.37%, respectively.
2016, 38(9): 2405-2409.
doi: 10.11999/JEIT151308
Abstract:
Compared with the first-order Markov random fields, higher-order Markov random fields could incorporate more statistical priors, thus have much expressive power of modeling. And the defogged images which based on dark channel prior have much error in sky regions and big white blocks. To solve those problems, this paper proposes a Markov random fields defogging method based on Color Lines. This method corrects the dark channel prior, according to the color lines which has a good robustness to color distortion, then uses the higher-order Markov random fields to optimize the transmission image to obtain final defogged image. The experimental results show that this method could improve the image resolution, while maintaining more image details.
Compared with the first-order Markov random fields, higher-order Markov random fields could incorporate more statistical priors, thus have much expressive power of modeling. And the defogged images which based on dark channel prior have much error in sky regions and big white blocks. To solve those problems, this paper proposes a Markov random fields defogging method based on Color Lines. This method corrects the dark channel prior, according to the color lines which has a good robustness to color distortion, then uses the higher-order Markov random fields to optimize the transmission image to obtain final defogged image. The experimental results show that this method could improve the image resolution, while maintaining more image details.
2016, 38(9): 2410-2414.
doi: 10.11999/JEIT160036
Abstract:
The design and implementation of ultra-high-speed FFT processor is imperative in radar system and prospective wireless communication system. In this paper, the fully-parallel-architecture FFT with bit-serial arithmetic is proposed. This method avoids the complexity of data addressing, access and routing. Based on the high-radix factorization, the multiplication number can be reduced. Out of the reason that twiddle factors are fixed in the design, constant coefficient optimization can be used in multiplications. Besides, bit-serial arithmetic cuts down the hardware cost, and makes the computation elements full-load to get a 100% efficiency. As a result, the presented 512-point FFT processer has 5.97 times gain in speed-throughput ratio while its hardware only accounts for 30% LUTs and 9% registers resource based on Xilinx V7-980t FPGA.
The design and implementation of ultra-high-speed FFT processor is imperative in radar system and prospective wireless communication system. In this paper, the fully-parallel-architecture FFT with bit-serial arithmetic is proposed. This method avoids the complexity of data addressing, access and routing. Based on the high-radix factorization, the multiplication number can be reduced. Out of the reason that twiddle factors are fixed in the design, constant coefficient optimization can be used in multiplications. Besides, bit-serial arithmetic cuts down the hardware cost, and makes the computation elements full-load to get a 100% efficiency. As a result, the presented 512-point FFT processer has 5.97 times gain in speed-throughput ratio while its hardware only accounts for 30% LUTs and 9% registers resource based on Xilinx V7-980t FPGA.
