Email alert
2020 Vol. 42, No. 9
Display Method:
2020, 42(9): 2065-2081.
doi: 10.11999/JEIT191055
Abstract:
Energy critical infrastructure has undergone transformative rapid development in the context of the rapid development of information technology, and has been deeply integrated with new technologies such as Artificial Intelligence (AI), big data, and the Internet of Things. While information technology significantly improves the efficiency and performance of energy critical infrastructure, it also brings new types of security threats that are more persistent and covert. An urgent problem is how to establish a systematic and intelligent security defense system for energy critical infrastructure. This paper starts with the development trend of energy critical infrastructure, and analyzes the mechanism of the traditional and new security threat mechanisms it faces. On this basis, insightful analysis on the research status and evolution trends of defense technologies for energy critical infrastructures is made.
Energy critical infrastructure has undergone transformative rapid development in the context of the rapid development of information technology, and has been deeply integrated with new technologies such as Artificial Intelligence (AI), big data, and the Internet of Things. While information technology significantly improves the efficiency and performance of energy critical infrastructure, it also brings new types of security threats that are more persistent and covert. An urgent problem is how to establish a systematic and intelligent security defense system for energy critical infrastructure. This paper starts with the development trend of energy critical infrastructure, and analyzes the mechanism of the traditional and new security threat mechanisms it faces. On this basis, insightful analysis on the research status and evolution trends of defense technologies for energy critical infrastructures is made.
2020, 42(9): 2082-2094.
doi: 10.11999/JEIT200009
Abstract:
With the prosperous of Android applications, Android malware has been scattered everywhere, which raises the serious security risk to users. On the other hand, the rapid developing of deep learning fires the combat between the two sides of malware detection. Inducing deep learning technologies into Android malware detection becomes the hottest topic of society. This paper summarizes the existing achievements of malware detection from four aspects: Data collection, feature construction, network structure and detection performance. Finally, the current limitations and facing challenges followed by the future researches are discussed.
With the prosperous of Android applications, Android malware has been scattered everywhere, which raises the serious security risk to users. On the other hand, the rapid developing of deep learning fires the combat between the two sides of malware detection. Inducing deep learning technologies into Android malware detection becomes the hottest topic of society. This paper summarizes the existing achievements of malware detection from four aspects: Data collection, feature construction, network structure and detection performance. Finally, the current limitations and facing challenges followed by the future researches are discussed.
2020, 42(9): 2095-2107.
doi: 10.11999/JEIT191056
Abstract:
Attack path discovery is an important research direction in automated penetration testing area. This paper introduces the research progress of domain independent intelligent planning technology and its application to the field of automated penetration testing oriented attack paths discovery. Firstly, the basic concept of attack path discovery is introduced and the related algorithms are divided into domain-specific and domain-independent intelligent planning based attack path discovery algorithms separately. Secondly, the domain-independent planning algorithms are classified into deterministic planning, uncertain planning and game planning, where each of which is described from principle, development and application aspect in detail. Thirdly, this paper summarizes the characteristics of automated penetration testing and compares the advantages and disadvantages of domain independent intelligent planning algorithms adopted in automated penetration testing. Lastly, the development of automated penetration testing oriented attack path discovery is prospected. It is hoped that this paper could contribute future research works on attack path discovery.
Attack path discovery is an important research direction in automated penetration testing area. This paper introduces the research progress of domain independent intelligent planning technology and its application to the field of automated penetration testing oriented attack paths discovery. Firstly, the basic concept of attack path discovery is introduced and the related algorithms are divided into domain-specific and domain-independent intelligent planning based attack path discovery algorithms separately. Secondly, the domain-independent planning algorithms are classified into deterministic planning, uncertain planning and game planning, where each of which is described from principle, development and application aspect in detail. Thirdly, this paper summarizes the characteristics of automated penetration testing and compares the advantages and disadvantages of domain independent intelligent planning algorithms adopted in automated penetration testing. Lastly, the development of automated penetration testing oriented attack path discovery is prospected. It is hoped that this paper could contribute future research works on attack path discovery.
2020, 42(9): 2108-2116.
doi: 10.11999/JEIT191036
Abstract:
Android system is now increasingly used in different kinds of smart devices, such as smart phones, smart watches, smart TVs and smart cars. Unfortunately, reverse attacks against Android applications are also emerging, which not only violates the intellectual right of application developers, but also brings security risks to end users. Existing Android application protection methods such as naming obfuscation, dynamic loading, and code hiding can protect Java code and native (C/C++) code, but are relatively simple and easy to be bypassed. A more promising method is to use instruction virtualization, but previous binary-based methods target specific architecture (x86), and cannot be applied to protect Android devices with different architectures. An architecture-independent instruction virtualization method is proposed, a prototype named Virtual Machine Packing Protection (VMPP) to protect Android native code is designed and implemented. VMPP includes a register-based fix-length instruction set, an interpreter to execute virtualized instructions, and a set of tool-chains for developers to use to protect their code. VMPP is tested on a large number of C/C++ code and real-world Android applications. The results show that VMPP can effectively protect the security of Android native code for different architectures with low overhead.
Android system is now increasingly used in different kinds of smart devices, such as smart phones, smart watches, smart TVs and smart cars. Unfortunately, reverse attacks against Android applications are also emerging, which not only violates the intellectual right of application developers, but also brings security risks to end users. Existing Android application protection methods such as naming obfuscation, dynamic loading, and code hiding can protect Java code and native (C/C++) code, but are relatively simple and easy to be bypassed. A more promising method is to use instruction virtualization, but previous binary-based methods target specific architecture (x86), and cannot be applied to protect Android devices with different architectures. An architecture-independent instruction virtualization method is proposed, a prototype named Virtual Machine Packing Protection (VMPP) to protect Android native code is designed and implemented. VMPP includes a register-based fix-length instruction set, an interpreter to execute virtualized instructions, and a set of tool-chains for developers to use to protect their code. VMPP is tested on a large number of C/C++ code and real-world Android applications. The results show that VMPP can effectively protect the security of Android native code for different architectures with low overhead.
2020, 42(9): 2117-2125.
doi: 10.11999/JEIT190695
Abstract:
In view of many limitations of current client anti plug-in methods, an anti-cheat method based on kernel events is proposed, and the network game anti-cheat system called CheatBlocker is implemented. This method uses the kernel event monitoring provided by Windows to intercept the abnormal access between processes and the injection of abnormal modules. At the same time, the anti-cheat Dynamic Loaded Library (DLL) injected from the kernel can block the simulation of the mouse keyboard. The experimental results show that CheatBlocker can defend against process module injection cheating and user input simulation cheating, and has low performance overhead. Moreover, CheatBlocker does not need to modify the kernel data or code which ensures the integrity of the kernel and is more compatible than the current anti-cheat systems.
In view of many limitations of current client anti plug-in methods, an anti-cheat method based on kernel events is proposed, and the network game anti-cheat system called CheatBlocker is implemented. This method uses the kernel event monitoring provided by Windows to intercept the abnormal access between processes and the injection of abnormal modules. At the same time, the anti-cheat Dynamic Loaded Library (DLL) injected from the kernel can block the simulation of the mouse keyboard. The experimental results show that CheatBlocker can defend against process module injection cheating and user input simulation cheating, and has low performance overhead. Moreover, CheatBlocker does not need to modify the kernel data or code which ensures the integrity of the kernel and is more compatible than the current anti-cheat systems.
2020, 42(9): 2126-2133.
doi: 10.11999/JEIT191059
Abstract:
Machine learning is widely used in malicious code detection and plays an important role in malicious code detection products. Constructing adversarial samples for malicious code detection machine learning models is the key to discovering defects in malicious code detection models, evaluating and improving malicious code detection systems. This paper proposes a method for generating malicious code adversarial samples based on genetic algorithms. The generated samples combat effectively the malicious code detection model based on machine learning, while ensuring the consistency of the executable and malicious behavior of malicious code samples, and improving effectively the authenticity of the generated adversarial samples and the accuracy of the model adversarial evaluation are presented. The experiments show that the proposed method of generating adversarial samples reduces the detection accuracy of the MalConv malicious code detection model by 14.65%, and can directly interfere with four commercial machine-based malicious code detection engines in VirusTotal. Among them, the accuracy rate of Cylance detection is only 53.55%.
Machine learning is widely used in malicious code detection and plays an important role in malicious code detection products. Constructing adversarial samples for malicious code detection machine learning models is the key to discovering defects in malicious code detection models, evaluating and improving malicious code detection systems. This paper proposes a method for generating malicious code adversarial samples based on genetic algorithms. The generated samples combat effectively the malicious code detection model based on machine learning, while ensuring the consistency of the executable and malicious behavior of malicious code samples, and improving effectively the authenticity of the generated adversarial samples and the accuracy of the model adversarial evaluation are presented. The experiments show that the proposed method of generating adversarial samples reduces the detection accuracy of the MalConv malicious code detection model by 14.65%, and can directly interfere with four commercial machine-based malicious code detection engines in VirusTotal. Among them, the accuracy rate of Cylance detection is only 53.55%.
2020, 42(9): 2134-2141.
doi: 10.11999/JEIT200008
Abstract:
Coordinated Cross Plane Session Termination (CXPST) repeatedly implements Low rate Denial of Service (LDoS) attacks on multiple target critical links, causing the cascading failure of the inter-domain routing system and the collapse of the internet. In the early stages of an attack, accurately locating the critical link under attack and carrying out targeted defense can prevent the occurrence of cascading failures. The research on existing locating methods is mainly based on the single-source hypothesis, and does not consider the impact of simultaneous failure of multiple target links on path withdrawal, so the locating accuracy is limited. To solve the above problems, a locating method is proposed based on Weighted Statistical Fit Score (WSFS). Using the target link selection strategy of cascading failure attack as inferring basis, scores are weighted by the reciprocal of the length of the withdrawal paths. The simulation results based on the actual network topology and vantage point location show that WSFS can improve the average accuracy rate by 5.45% compared with the current optimal method. Experimental results prove that WSFS is more suitable for locating target failure links in inter-domain routing system cascading failure than other locating methods.
Coordinated Cross Plane Session Termination (CXPST) repeatedly implements Low rate Denial of Service (LDoS) attacks on multiple target critical links, causing the cascading failure of the inter-domain routing system and the collapse of the internet. In the early stages of an attack, accurately locating the critical link under attack and carrying out targeted defense can prevent the occurrence of cascading failures. The research on existing locating methods is mainly based on the single-source hypothesis, and does not consider the impact of simultaneous failure of multiple target links on path withdrawal, so the locating accuracy is limited. To solve the above problems, a locating method is proposed based on Weighted Statistical Fit Score (WSFS). Using the target link selection strategy of cascading failure attack as inferring basis, scores are weighted by the reciprocal of the length of the withdrawal paths. The simulation results based on the actual network topology and vantage point location show that WSFS can improve the average accuracy rate by 5.45% compared with the current optimal method. Experimental results prove that WSFS is more suitable for locating target failure links in inter-domain routing system cascading failure than other locating methods.
2020, 42(9): 2142-2149.
doi: 10.11999/JEIT191054
Abstract:
Due to the proliferation of geographic location virtual software and the easy simulation or tampering of civil satellite positioning signals, it is difficult to realize the trusted authentication of geographic location. In view of the security risk of single-point failure in the existing location certification scheme using centralized architecture, a zero-knowledge location certification method based on blockchain is proposed, combining with zero knowledge certification protocol, to achieve a decentralized, privacy protected, highly accurate, review offset geographic location certification service, so as to ensure the accuracy of the location provided by users. This method not only ensures the confidentiality of the location data, but also proves that the location data can not tamper once it is linked. The results of the test analysis show that the average performance of the whole proving process is about 5 s/time, and the total time of proof generation and verification is 50.5～55.5 ms. Therefore, the algorithm has better performance overhead, which can meet the actual application requirements.
Due to the proliferation of geographic location virtual software and the easy simulation or tampering of civil satellite positioning signals, it is difficult to realize the trusted authentication of geographic location. In view of the security risk of single-point failure in the existing location certification scheme using centralized architecture, a zero-knowledge location certification method based on blockchain is proposed, combining with zero knowledge certification protocol, to achieve a decentralized, privacy protected, highly accurate, review offset geographic location certification service, so as to ensure the accuracy of the location provided by users. This method not only ensures the confidentiality of the location data, but also proves that the location data can not tamper once it is linked. The results of the test analysis show that the average performance of the whole proving process is about 5 s/time, and the total time of proof generation and verification is 50.5～55.5 ms. Therefore, the algorithm has better performance overhead, which can meet the actual application requirements.
2020, 42(9): 2150-2157.
doi: 10.11999/JEIT190690
Abstract:
To solve the problem that the existing algorithms for recognition of RS codes need to transform the code characters among different domains and poor performance, a new algorithm based on soft decision is proposed. Firstly, starting from the definition of RS codes, the equivalent conversion mode of the check relation of RS code from GF (2m) to GF (2) is given, which avoids the complex symbol transformation in different domains. Secondly, the average check conformity which can measure the validity of the check relationship is introduced and based on its statistical characteristics and minimax decision criteria, the possible code length and corresponding m-level primitive polynomials are traversed to match the initial code root, as the results, the code length and primitive polynomial are recognized. Finally, under the identified code length and the primitive polynomial, the GF (2m) is constructed, and the continuous code root matching decision is made, then the generation polynomial is recognized. The simulation results show that the derived statistical characteristics of the average check conformity are consistent with the actual situation, and the proposed algorithm can effectively recognize parameter under low Signal-to-Noise Ratio (SNR). At the same time, the proposed algorithm has good adaptability to low SNR. At SNR of 6 dB, the recognition rate of common RS codes in engineering can reach more than 90%. Compared with the existing methods, the performance of this algorithm is better than hard-decision algorithm, besides, it is improved by more than 1 dB compared by traditional algorithms.
To solve the problem that the existing algorithms for recognition of RS codes need to transform the code characters among different domains and poor performance, a new algorithm based on soft decision is proposed. Firstly, starting from the definition of RS codes, the equivalent conversion mode of the check relation of RS code from GF (2m) to GF (2) is given, which avoids the complex symbol transformation in different domains. Secondly, the average check conformity which can measure the validity of the check relationship is introduced and based on its statistical characteristics and minimax decision criteria, the possible code length and corresponding m-level primitive polynomials are traversed to match the initial code root, as the results, the code length and primitive polynomial are recognized. Finally, under the identified code length and the primitive polynomial, the GF (2m) is constructed, and the continuous code root matching decision is made, then the generation polynomial is recognized. The simulation results show that the derived statistical characteristics of the average check conformity are consistent with the actual situation, and the proposed algorithm can effectively recognize parameter under low Signal-to-Noise Ratio (SNR). At the same time, the proposed algorithm has good adaptability to low SNR. At SNR of 6 dB, the recognition rate of common RS codes in engineering can reach more than 90%. Compared with the existing methods, the performance of this algorithm is better than hard-decision algorithm, besides, it is improved by more than 1 dB compared by traditional algorithms.
2020, 42(9): 2158-2165.
doi: 10.11999/JEIT190345
Abstract:
For security of stereoscopic video, a video encryption and information hiding algorithm based on entropy coding is proposed. Firstly, with the analysis of Multi-view Video Coding (MVC), the physical mechanism of error drift is investigated. By applying the stereoscopic masking effect, the frames to be encrypted and the frames to be embedded are determined. Secondly, by using a Context-based Adaptive Binary Arithmetic Coding (CABAC) bin-string substitution technique, the encryption and data hiding of stereoscopic video are implemented. Experimental implementation reveals that the video stream has the format compatibility and the bit-rate remains unchanged after encryption and data hiding. The video quality degradation is negligible. The algorithm has significant advantages in computational complexity and rate increase.
For security of stereoscopic video, a video encryption and information hiding algorithm based on entropy coding is proposed. Firstly, with the analysis of Multi-view Video Coding (MVC), the physical mechanism of error drift is investigated. By applying the stereoscopic masking effect, the frames to be encrypted and the frames to be embedded are determined. Secondly, by using a Context-based Adaptive Binary Arithmetic Coding (CABAC) bin-string substitution technique, the encryption and data hiding of stereoscopic video are implemented. Experimental implementation reveals that the video stream has the format compatibility and the bit-rate remains unchanged after encryption and data hiding. The video quality degradation is negligible. The algorithm has significant advantages in computational complexity and rate increase.
2020, 42(9): 2166-2172.
doi: 10.11999/JEIT190543
Abstract:
The mapping strategy of virtual network has important effect on the resource availability and survivability of the Elastic Optical Network (EON). A survivable virtual optical network Coordinated Mapping based on the Distance and Spectrum Dispersion Awareness (CM-DSDA) between nodes is proposed in the paper. A physical node weighted sorting strategy is studied, which not only considers the number of physical node computing resources, but also considers the location centrality of the physical nodes in the EON topology. And a method of spectrum dispersion is designed to evaluate the link’s spectrum fragmentation. During the virtual link’s survivability mapping, the working and protection optical paths adjacent the position of the mapped physical nodes with the minimum number of spectrum usage and the lowest frequency spectrum dispersion are selected to coordinated mapping the virtual optical networks. Simulation results show that the CM-DSDA can effectively increase the EON’s spectrum utilization and reduce bandwidth blocking probability.
The mapping strategy of virtual network has important effect on the resource availability and survivability of the Elastic Optical Network (EON). A survivable virtual optical network Coordinated Mapping based on the Distance and Spectrum Dispersion Awareness (CM-DSDA) between nodes is proposed in the paper. A physical node weighted sorting strategy is studied, which not only considers the number of physical node computing resources, but also considers the location centrality of the physical nodes in the EON topology. And a method of spectrum dispersion is designed to evaluate the link’s spectrum fragmentation. During the virtual link’s survivability mapping, the working and protection optical paths adjacent the position of the mapped physical nodes with the minimum number of spectrum usage and the lowest frequency spectrum dispersion are selected to coordinated mapping the virtual optical networks. Simulation results show that the CM-DSDA can effectively increase the EON’s spectrum utilization and reduce bandwidth blocking probability.
2020, 42(9): 2173-2179.
doi: 10.11999/JEIT190545
Abstract:
To improve the service experience provided by the operator network, this paper studies the online migration of Service Function Chain(SFC). Based on the Markov Decision Process(MDP), modeling analysis is performed on the migration of multiple Virtual Network Functions(VNF) in SFC. By combining reinforcement learning and deep neural networks, a double Deep Q-Network(double DQN) based service function chain migration mechanism is proposed. This method can make online migration decisions and avoid over-estimation. Experimental result shows that when compared with the fixed deployment algorithm and the greedy algorithm, the double DQN based SFC migration mechanism has obvious advantages in end-to-end delay and network system revenue, which can help the mobile operator to improve the quality of experience and the efficiency of resources usage.
To improve the service experience provided by the operator network, this paper studies the online migration of Service Function Chain(SFC). Based on the Markov Decision Process(MDP), modeling analysis is performed on the migration of multiple Virtual Network Functions(VNF) in SFC. By combining reinforcement learning and deep neural networks, a double Deep Q-Network(double DQN) based service function chain migration mechanism is proposed. This method can make online migration decisions and avoid over-estimation. Experimental result shows that when compared with the fixed deployment algorithm and the greedy algorithm, the double DQN based SFC migration mechanism has obvious advantages in end-to-end delay and network system revenue, which can help the mobile operator to improve the quality of experience and the efficiency of resources usage.
2020, 42(9): 2180-2185.
doi: 10.11999/JEIT190654
Abstract:
To solve the problem of increasing the digital currency of mobile terminals based on limited residual resources of the system, a task offloading scheme is proposed based on node residual resources and network delay in the joint blockchain and fog computing system. In order to offload the task in an optimal way, the expected revenue of mobile terminals is firstly analyzed based on the amount of tasks. Secondly, based on the remaining computing resources, storage resources, power resources and network delays, the expenditure of mobile terminal is analyzed. Then, a mathematical optimization model is established to maximize the digital currency income of the mobile terminal. The Simulated Annealing (SA) algorithm is employed to deal with the suboptimal model, respectively. Simulation results demonstrate the effectiveness of the proposed scheme.
To solve the problem of increasing the digital currency of mobile terminals based on limited residual resources of the system, a task offloading scheme is proposed based on node residual resources and network delay in the joint blockchain and fog computing system. In order to offload the task in an optimal way, the expected revenue of mobile terminals is firstly analyzed based on the amount of tasks. Secondly, based on the remaining computing resources, storage resources, power resources and network delays, the expenditure of mobile terminal is analyzed. Then, a mathematical optimization model is established to maximize the digital currency income of the mobile terminal. The Simulated Annealing (SA) algorithm is employed to deal with the suboptimal model, respectively. Simulation results demonstrate the effectiveness of the proposed scheme.
2020, 42(9): 2186-2193.
doi: 10.11999/5EIT190649
Abstract:
To ensure the security of MIMO heterogeneous networks while facing the active multi-antenna eavesdropping, a robust secure transmission scheme based on artificial noise is proposed. Firstly, considering that the eavesdropper sends uplink pilot interference signal, the influence of uplink pilot interference on the channel estimation of legitimate users is studied. Then, based on the channel estimation results, the precoding matrix of downlink data and noise signals of macro base station and micro base station are designed, respectively. The expression of system security rate in this case is derived; After that, for maximizing the system security rate, the transmission power of downlink data and noise signal of base station are optimized, and a solution method based on one-dimensional linear search is proposed. Furthermore, considering the situation that the eavesdropper sends the uplink pilot interference and then sends the noise to interfere with the downlink communication of the legitimate user, a discrete zero-sum game method is proposed to obtain the optimal transmission power design. The simulation results verify the security and robustness of the proposed scheme.
To ensure the security of MIMO heterogeneous networks while facing the active multi-antenna eavesdropping, a robust secure transmission scheme based on artificial noise is proposed. Firstly, considering that the eavesdropper sends uplink pilot interference signal, the influence of uplink pilot interference on the channel estimation of legitimate users is studied. Then, based on the channel estimation results, the precoding matrix of downlink data and noise signals of macro base station and micro base station are designed, respectively. The expression of system security rate in this case is derived; After that, for maximizing the system security rate, the transmission power of downlink data and noise signal of base station are optimized, and a solution method based on one-dimensional linear search is proposed. Furthermore, considering the situation that the eavesdropper sends the uplink pilot interference and then sends the noise to interfere with the downlink communication of the legitimate user, a discrete zero-sum game method is proposed to obtain the optimal transmission power design. The simulation results verify the security and robustness of the proposed scheme.
2020, 42(9): 2194-2200.
doi: 10.11999/JEIT190662
Abstract:
The most critical issue in the deployment of Mobile Wireless Sensor Networks (MWSN) is how to provide maximum regional coverage.To solve the problem that the existing coverage control algorithm has unsatisfactory coverage, low deployment efficiency and high energy consumption, an efficient deployment strategy is proposed.The first stage uses the Voronoi diagram to obtain the coverage hole of the entire network, and detects the uncovered area in the Voronoi polygon, and provides virtual force to drive the sensor movement, and uses the dynamic adjustment strategy to change the moving step size, thereby reducing energy loss;The second stage proposes a detection mechanism that uses a Delaunay triangulation to detect local coverage holes between sensors and repair them.The simulation results show that the algorithm accelerates the convergence speed while improving the network coverage, and provides a new solution for deploying mobile wireless sensor networks.
The most critical issue in the deployment of Mobile Wireless Sensor Networks (MWSN) is how to provide maximum regional coverage.To solve the problem that the existing coverage control algorithm has unsatisfactory coverage, low deployment efficiency and high energy consumption, an efficient deployment strategy is proposed.The first stage uses the Voronoi diagram to obtain the coverage hole of the entire network, and detects the uncovered area in the Voronoi polygon, and provides virtual force to drive the sensor movement, and uses the dynamic adjustment strategy to change the moving step size, thereby reducing energy loss;The second stage proposes a detection mechanism that uses a Delaunay triangulation to detect local coverage holes between sensors and repair them.The simulation results show that the algorithm accelerates the convergence speed while improving the network coverage, and provides a new solution for deploying mobile wireless sensor networks.
2020, 42(9): 2201-2207.
doi: 10.11999/JEIT190691
Abstract:
Device to Device (D2D) communications can effectively offload base station traffic. In D2D networks, the popular content is not only needs to be shared, but also the individual content is needs to be cached. In this paper, the problem of cache content selection is researched. A Content Social Value Prediction(CSVP) method based on feature perception is proposed. Value prediction can not only reduce latency, but also reduce the number of cache replacements and reduce cache costs. Firstly, the current value of content is calculated by combining user features and content features, and then future value is calculated through user social relationships. The small base station provides the user with a personalized content caching service according to the value of the content, and the base station selects a content with a larger value in the individual cache content of each small base station as popular content. Simulation results show that the caching strategy based on the proposed method can alleviate the base station traffic effectively, and reduce the delay by about 20%～40%.
Device to Device (D2D) communications can effectively offload base station traffic. In D2D networks, the popular content is not only needs to be shared, but also the individual content is needs to be cached. In this paper, the problem of cache content selection is researched. A Content Social Value Prediction(CSVP) method based on feature perception is proposed. Value prediction can not only reduce latency, but also reduce the number of cache replacements and reduce cache costs. Firstly, the current value of content is calculated by combining user features and content features, and then future value is calculated through user social relationships. The small base station provides the user with a personalized content caching service according to the value of the content, and the base station selects a content with a larger value in the individual cache content of each small base station as popular content. Simulation results show that the caching strategy based on the proposed method can alleviate the base station traffic effectively, and reduce the delay by about 20%～40%.
2020, 42(9): 2208-2215.
doi: 10.11999/JEIT190320
Abstract:
For blind modulation recognition of Multiple Input Multiple Output (MIMO) signals in non-cooperative communication, a modulation recognition method based on Independent Component Analysis (ICA) and feature extraction is proposed. According to the signal independence of each transmitting antenna in space division multiplexing MIMO system, the ICA algorithm is used to separate the transmitting signal from the received mixed signal. In order to realize modulation recognition under completely blind condition, the Minimum Description Length (MDL) criterion is used to estimate the number of transmitting antennas before ICA separation. After obtaining the transmitted signal, four characteristic parameters are constructed by using six-order cumulant, cyclic spectrum and fourth-power spectrum algorithm, and then the modulation type of the signal is identified by using hierarchical neural network classifier. The simulation results show that the proposed method can effectively recognize {2PSK, 2ASK, 2FSK, 4PSK, 4ASK, MSK, 8PSK, 16QAM} eight MIMO signals at low SNR. When the number of transmitting antennas is 2, the number of receiving antennas is 5 and the SNR is 2dB, the recognition rate can reach more than 98%.
For blind modulation recognition of Multiple Input Multiple Output (MIMO) signals in non-cooperative communication, a modulation recognition method based on Independent Component Analysis (ICA) and feature extraction is proposed. According to the signal independence of each transmitting antenna in space division multiplexing MIMO system, the ICA algorithm is used to separate the transmitting signal from the received mixed signal. In order to realize modulation recognition under completely blind condition, the Minimum Description Length (MDL) criterion is used to estimate the number of transmitting antennas before ICA separation. After obtaining the transmitted signal, four characteristic parameters are constructed by using six-order cumulant, cyclic spectrum and fourth-power spectrum algorithm, and then the modulation type of the signal is identified by using hierarchical neural network classifier. The simulation results show that the proposed method can effectively recognize {2PSK, 2ASK, 2FSK, 4PSK, 4ASK, MSK, 8PSK, 16QAM} eight MIMO signals at low SNR. When the number of transmitting antennas is 2, the number of receiving antennas is 5 and the SNR is 2dB, the recognition rate can reach more than 98%.
2020, 42(9): 2216-2222.
doi: 10.11999/JEIT190505
Abstract:
Grant-free Non-Orthogonal Multiple Access (NOMA) combined with Multi-User Detection (MUD) technology can meet the requirements of large connection volume, low signaling overhead and low latency transmission in massive Machine Type Communications (mMTC) scenarios. In the MUD algorithm based on Compressed Sensing (CS), the number of active users is often used as known information, but it is difficult to accurately estimate in the actual communication system. Based on this, this paper proposes a multi-user algorithm (Modified Sparsity Adaptive Matching Pursuit MUD, MSAMP-MUP) to improve the adaptive matching of sparsity. Firstly, the algorithm uses the generalized Dice coefficient matching criterion to select the atom that best matches the residual, and updates the user support set. When the residual energy is close to the noise energy, the iteration is terminated to obtain the final support set; Otherwise, the above criteria are used to update the user support set, and the estimation accuracy of the active users in the support set is improved. In the iteration process, different iteration steps are selected according to the ratio of the last two residual energies, so as to reduce the number of detection iterations. The simulation results show that, compared with the traditional CS-based MUD algorithm, the proposed algorithm reduces the bit error rate by about 9% and the number of iterations by about 10%.
Grant-free Non-Orthogonal Multiple Access (NOMA) combined with Multi-User Detection (MUD) technology can meet the requirements of large connection volume, low signaling overhead and low latency transmission in massive Machine Type Communications (mMTC) scenarios. In the MUD algorithm based on Compressed Sensing (CS), the number of active users is often used as known information, but it is difficult to accurately estimate in the actual communication system. Based on this, this paper proposes a multi-user algorithm (Modified Sparsity Adaptive Matching Pursuit MUD, MSAMP-MUP) to improve the adaptive matching of sparsity. Firstly, the algorithm uses the generalized Dice coefficient matching criterion to select the atom that best matches the residual, and updates the user support set. When the residual energy is close to the noise energy, the iteration is terminated to obtain the final support set; Otherwise, the above criteria are used to update the user support set, and the estimation accuracy of the active users in the support set is improved. In the iteration process, different iteration steps are selected according to the ratio of the last two residual energies, so as to reduce the number of detection iterations. The simulation results show that, compared with the traditional CS-based MUD algorithm, the proposed algorithm reduces the bit error rate by about 9% and the number of iterations by about 10%.
2020, 42(9): 2223-2230.
doi: 10.11999/JEIT190493
Abstract:
Under certain environmental conditions, the unknown system errors often occur and yield to larger filtering errors when the unverified or uncalibrated measurement equation is used. Incremental equation can be introduced, which can effectively solve the problem of state estimation for the systems under poor observation condition. In this paper, the linear discrete incremental system with unknown noise statistics is considered. Firstly, a noise statistics estimation algorithm is proposed based on innovation. The unbiased estimation of system noise statistics can be obtained. Furthermore, a new incremental system adaptive Kalman filtering algorithm is proposed. Compared with the existing adaptive incremental filtering algorithm, the state estimation accuracy of the proposed algorithm is higher. Two simulation examples prove its effectiveness and feasibility.
Under certain environmental conditions, the unknown system errors often occur and yield to larger filtering errors when the unverified or uncalibrated measurement equation is used. Incremental equation can be introduced, which can effectively solve the problem of state estimation for the systems under poor observation condition. In this paper, the linear discrete incremental system with unknown noise statistics is considered. Firstly, a noise statistics estimation algorithm is proposed based on innovation. The unbiased estimation of system noise statistics can be obtained. Furthermore, a new incremental system adaptive Kalman filtering algorithm is proposed. Compared with the existing adaptive incremental filtering algorithm, the state estimation accuracy of the proposed algorithm is higher. Two simulation examples prove its effectiveness and feasibility.
2020, 42(9): 2231-2238.
doi: 10.11999/JEIT190327
Abstract:
In order to estimate the length of moving ship targets in Alpha stable distribution noise, a moving target length estimation method based on Generalized Time-Frequency Analysis (G-TFA) and least squares estimation is proposed, which utilizes nonlinear transform to suppress impulsive noise and Doppler effect to estimate target motion characteristics. The method uses G-TFA to obtain the Doppler frequency of moving targets in a stable distributed noise environment. Then, the least squares method is used to estimate the target speed and the closest point of approach time of different positions. Finally, the target length is calculated using the above estimation results. Taking Generalized Winger-Ville Distribution (G-WVD) as an example, the ability of G-TFA to extract Doppler features in Alpha stable distributed noise is theoretically derived. The robustness of the proposed method under low-to-medium mixed signal-to-noise ratio is verified by simulation experiments. Compared with the existing methods, the proposed method does not need to estimate the noise characteristic index, and the performance is better than the methods based on the traditional time-frequency analysis.
In order to estimate the length of moving ship targets in Alpha stable distribution noise, a moving target length estimation method based on Generalized Time-Frequency Analysis (G-TFA) and least squares estimation is proposed, which utilizes nonlinear transform to suppress impulsive noise and Doppler effect to estimate target motion characteristics. The method uses G-TFA to obtain the Doppler frequency of moving targets in a stable distributed noise environment. Then, the least squares method is used to estimate the target speed and the closest point of approach time of different positions. Finally, the target length is calculated using the above estimation results. Taking Generalized Winger-Ville Distribution (G-WVD) as an example, the ability of G-TFA to extract Doppler features in Alpha stable distributed noise is theoretically derived. The robustness of the proposed method under low-to-medium mixed signal-to-noise ratio is verified by simulation experiments. Compared with the existing methods, the proposed method does not need to estimate the noise characteristic index, and the performance is better than the methods based on the traditional time-frequency analysis.
2020, 42(9): 2239-2245.
doi: 10.11999/JEIT190539
Abstract:
Considering the requirement for weak signal detection in non-Gaussian background interference, after conceptions and ideas summarizing for Gaussianization processing and extended matched filter, two Gaussianizaiton filters and corresponding detections are proposed based on symmetric alpha-stable distribution modeling, comparing with those under Gaussian mixture modeling proposed earlier. All these Gaussianization filters and extended matched filters are realized in simulation. Their performances, such as Gaussianzing effect, detecting capability and running time are studies in system. Some conclusions are reached: Gaussianization can improve the detecting performance of the succeeding matched filter because of its restraining of big impulsive samples; Gaussianization filters under symmetric alpha-stable distribution modeling have higher operatiing efficiency while their peformance are similar to those under Gaussian mixture modeling.
Considering the requirement for weak signal detection in non-Gaussian background interference, after conceptions and ideas summarizing for Gaussianization processing and extended matched filter, two Gaussianizaiton filters and corresponding detections are proposed based on symmetric alpha-stable distribution modeling, comparing with those under Gaussian mixture modeling proposed earlier. All these Gaussianization filters and extended matched filters are realized in simulation. Their performances, such as Gaussianzing effect, detecting capability and running time are studies in system. Some conclusions are reached: Gaussianization can improve the detecting performance of the succeeding matched filter because of its restraining of big impulsive samples; Gaussianization filters under symmetric alpha-stable distribution modeling have higher operatiing efficiency while their peformance are similar to those under Gaussian mixture modeling.
2020, 42(9): 2246-2252.
doi: 10.11999/JEIT190687
Abstract:
The problem that single-channel Synthetic Aperture Radar (SAR) can not effectively suppress FM slope mismatch jamming is studied in this paper. According to the difference between the jamming spectrum and the real echo spectrum, a method for suppressing the mismatch jamming of SAR FM slope based on matched filtering in the frequency domain is proposed. Based on the accurate measurement of the slope of jamming FM, the sparse super-resolution estimation of the jamming position is carried out to obtain a more accurate phase of jamming delay by utilizing the sparsity of the jamming signal and the characteristics of low Signal-to-Interference Ratio (SIR). Then, the spectrum of the interference signal is reconstructed by the obtained slope of interference frequency modulation and phase delay. Based on this, the orthogonal matched filter is designed to suppress the interference signal in the frequency domain and reconstruct the undistorted scene image. Finally, computer simulation experiments verify the effectiveness of the proposed method.
The problem that single-channel Synthetic Aperture Radar (SAR) can not effectively suppress FM slope mismatch jamming is studied in this paper. According to the difference between the jamming spectrum and the real echo spectrum, a method for suppressing the mismatch jamming of SAR FM slope based on matched filtering in the frequency domain is proposed. Based on the accurate measurement of the slope of jamming FM, the sparse super-resolution estimation of the jamming position is carried out to obtain a more accurate phase of jamming delay by utilizing the sparsity of the jamming signal and the characteristics of low Signal-to-Interference Ratio (SIR). Then, the spectrum of the interference signal is reconstructed by the obtained slope of interference frequency modulation and phase delay. Based on this, the orthogonal matched filter is designed to suppress the interference signal in the frequency domain and reconstruct the undistorted scene image. Finally, computer simulation experiments verify the effectiveness of the proposed method.
2020, 42(9): 2253-2260.
doi: 10.11999/JEIT190356
Abstract:
A motion compensation method based on Kirchhoff migration imaging algorithm for MIMO array is proposed. In the method, the traditional Kirchhoff intergral equation is introduced into the MIMO array and the forward Green function is replaced by the backward Green function to solve the problem of the large view field of the small array. The echo signal of the reference unit is directly recieved by adding a reference channel to estimate the position change information of the moving target, while the signal of the receiving units is sampled through the microwave switch time divition multiplexing. Then 3D imaging is performed after motion compensation for the data of the MIMO channel, which avoids effectively the problem of defocusing due to target migration and the deviation of the imaging result from the real position.
A motion compensation method based on Kirchhoff migration imaging algorithm for MIMO array is proposed. In the method, the traditional Kirchhoff intergral equation is introduced into the MIMO array and the forward Green function is replaced by the backward Green function to solve the problem of the large view field of the small array. The echo signal of the reference unit is directly recieved by adding a reference channel to estimate the position change information of the moving target, while the signal of the receiving units is sampled through the microwave switch time divition multiplexing. Then 3D imaging is performed after motion compensation for the data of the MIMO channel, which avoids effectively the problem of defocusing due to target migration and the deviation of the imaging result from the real position.
2020, 42(9): 2261-2268.
doi: 10.11999/JEIT190674
Abstract:
The system transfer function of high-squint SAR system with curved trajectory has complex multi-dimensional spatial variability. The existing efficient frequency-domain echo simulation algorithms are difficult to achieve high-precision echo simulation of extended scenes. Therefore, a fast echo simulation method based on sub-aperture Keystone transform is proposed for maneuvering SAR. Based on the time-domain range compression function after range cell migration correction, the method calculates efficiently the range compression echo of scene, and then realizes the echo simulation of extended scene by high precision range inverse processing. In order to reduce the influence of residual range cell migration on the accuracy of echo simulation, the method of the Keystone transform of sub-aperture is used in range processing to achieve accurate correction of space-variant range cell migration in extended scenes. The theoretical analysis and simulation results show that the proposed method can quickly and accurately simulate the original echo data of extended scenes for high-squint SAR mounted on a maneuvering platform.
The system transfer function of high-squint SAR system with curved trajectory has complex multi-dimensional spatial variability. The existing efficient frequency-domain echo simulation algorithms are difficult to achieve high-precision echo simulation of extended scenes. Therefore, a fast echo simulation method based on sub-aperture Keystone transform is proposed for maneuvering SAR. Based on the time-domain range compression function after range cell migration correction, the method calculates efficiently the range compression echo of scene, and then realizes the echo simulation of extended scene by high precision range inverse processing. In order to reduce the influence of residual range cell migration on the accuracy of echo simulation, the method of the Keystone transform of sub-aperture is used in range processing to achieve accurate correction of space-variant range cell migration in extended scenes. The theoretical analysis and simulation results show that the proposed method can quickly and accurately simulate the original echo data of extended scenes for high-squint SAR mounted on a maneuvering platform.
Deep Convolutional Neural Network for Parking Space Occupancy Detection Based on Non-local Operation
2020, 42(9): 2269-2276.
doi: 10.11999/JEIT190349
Abstract:
With the intelligent development of urban traffic, accurate and efficient access to available parking spaces is essential to solve the increasingly difficult problem of parking difficulties. Therefore, this paper proposes a deep convolutional neural network parking occupancy detection algorithm based on non-local operation. For the image characteristics of parking spaces, non-local operations are introduced, the similarity between distant pixels is measured, and the high-frequency features of the edges are directly obtained. The local details are obtained by using small convolution kernels, and the network is trained in an end-to-end manner. In the experiment, the network structure is optimized by setting different convolution kernel sizes and non-local module layers. The experimental results show that compared with the traditional texture feature-based parking space occupancy detection algorithm, the proposed algorithm has significant advantages in both prediction accuracy and generalization performance of the model. At the same time, compared with the currently widely used convolutional neural network based on local feature extraction, the algorithm also has great advantages. In real scenes, the algorithm also has high precision and has practical application value.
With the intelligent development of urban traffic, accurate and efficient access to available parking spaces is essential to solve the increasingly difficult problem of parking difficulties. Therefore, this paper proposes a deep convolutional neural network parking occupancy detection algorithm based on non-local operation. For the image characteristics of parking spaces, non-local operations are introduced, the similarity between distant pixels is measured, and the high-frequency features of the edges are directly obtained. The local details are obtained by using small convolution kernels, and the network is trained in an end-to-end manner. In the experiment, the network structure is optimized by setting different convolution kernel sizes and non-local module layers. The experimental results show that compared with the traditional texture feature-based parking space occupancy detection algorithm, the proposed algorithm has significant advantages in both prediction accuracy and generalization performance of the model. At the same time, compared with the currently widely used convolutional neural network based on local feature extraction, the algorithm also has great advantages. In real scenes, the algorithm also has high precision and has practical application value.
2020, 42(9): 2277-2284.
doi: 10.11999/JEIT190393
Abstract:
Co-saliency object detection aims to discover common and salient objects in an image group which contains two or more relevant images. In this paper, a method of using machine learning is proposed to detect co-saliency objects. Firstly, some simple images are selected to form a simple image set based on four scoring indicators. Secondly, positive and negative samples are extracted from the simple images set based on co-coherence characteristics, and high-dimensional semantic features are extracted by the deep learning model which receives RGBD four-channels input. Thirdly, the co-saliency classifier is trained by positive and negative samples, and co-saliency maps are generated by testing all the superpixels in the images by the co-saliency classifier. Finally, a smooth fusion operation is adopted to generate the final co-saliency map. Experimental results on the public benchmark dataset show that the proposed algorithm is superior to the state-of-the-art methods in terms of accuracy and efficiency, and it is robust.
Co-saliency object detection aims to discover common and salient objects in an image group which contains two or more relevant images. In this paper, a method of using machine learning is proposed to detect co-saliency objects. Firstly, some simple images are selected to form a simple image set based on four scoring indicators. Secondly, positive and negative samples are extracted from the simple images set based on co-coherence characteristics, and high-dimensional semantic features are extracted by the deep learning model which receives RGBD four-channels input. Thirdly, the co-saliency classifier is trained by positive and negative samples, and co-saliency maps are generated by testing all the superpixels in the images by the co-saliency classifier. Finally, a smooth fusion operation is adopted to generate the final co-saliency map. Experimental results on the public benchmark dataset show that the proposed algorithm is superior to the state-of-the-art methods in terms of accuracy and efficiency, and it is robust.
2020, 42(9): 2285-2292.
doi: 10.11999/JEIT190755
Abstract:
Rainy days and other severe weather will seriously affect the image quality, thus affecting the performance of vision processing algorithms. In order to improve the imaging quality of rain images, a rain removal algorithm based on multi-channel multi-scale convolution neural network to extract rain line features is proposed. Firstly, the rain images are decomposed by wavelet threshold-guided bilateral filtering to obtain high-frequency rain line images and low-frequency background images with high contour preservation. Then, in order to make the rain line information in the high-frequency part of the image more obvious and reduce the background misjudgment in the high-frequency image during the rain line feature learning, the obtained high-frequency rain line image is passed through a filter again to obtain a higher-frequency rain line image with reduced background information and enhanced rain line information. Secondly, in view of the large amount of raindrop imprint left on the low-frequency background image, it is proposed to send the low-frequency background image and the higher-frequency rain line image together into the convolution neural network for feature learning, in which multi-scale feature information is extracted from the image, finally, a more complete restoration image with rain line removal is obtained. At the same time, when constructing the network model, hole convolution is used instead of standard convolution to extract the feature information of the image, thus obtaining richer image features and improving the rain removal performance of the algorithm. From the experimental results, after removing rain, the image is clear and the detail retention is high.
Rainy days and other severe weather will seriously affect the image quality, thus affecting the performance of vision processing algorithms. In order to improve the imaging quality of rain images, a rain removal algorithm based on multi-channel multi-scale convolution neural network to extract rain line features is proposed. Firstly, the rain images are decomposed by wavelet threshold-guided bilateral filtering to obtain high-frequency rain line images and low-frequency background images with high contour preservation. Then, in order to make the rain line information in the high-frequency part of the image more obvious and reduce the background misjudgment in the high-frequency image during the rain line feature learning, the obtained high-frequency rain line image is passed through a filter again to obtain a higher-frequency rain line image with reduced background information and enhanced rain line information. Secondly, in view of the large amount of raindrop imprint left on the low-frequency background image, it is proposed to send the low-frequency background image and the higher-frequency rain line image together into the convolution neural network for feature learning, in which multi-scale feature information is extracted from the image, finally, a more complete restoration image with rain line removal is obtained. At the same time, when constructing the network model, hole convolution is used instead of standard convolution to extract the feature information of the image, thus obtaining richer image features and improving the rain removal performance of the algorithm. From the experimental results, after removing rain, the image is clear and the detail retention is high.
2020, 42(9): 2293-2301.
doi: 10.11999/JEIT190723
Abstract:
Affected by the micro-lens geometric calibration accuracy of the light field camera, the decoding error of the 4D light field in the angular direction will cause the edge information loss of the integrated refocused image, which will reduce the accuracy of the all-in-focus image fusion. In this paper, a light field all-in-focus image fusion algorithm based on edge-enhanced guided filtering is proposed. Through multi-scale decomposition of the digital refocused images and guided filtering optimization of the feature layer decision map, the final all-in-focus image is obtained. Compared with the traditional fusion algorithm, the edge information loss caused by the 4D light field calibration error is compensated in the presented method. In the step of multi-scale decomposition of the refocused image, the edge layer extraction is added to accomplish the high-frequency information enhancement. Then the multi-scale evaluation model is established to optimize the edge layer’s guided filtering parameters to obtain a better light field all-in-focus image. The experimental results show that the edge intensity and the perceptual sharpness of the all-in-focus image can be improved without significantly reducing the similarity between the all-in-focus image and the original image.
Affected by the micro-lens geometric calibration accuracy of the light field camera, the decoding error of the 4D light field in the angular direction will cause the edge information loss of the integrated refocused image, which will reduce the accuracy of the all-in-focus image fusion. In this paper, a light field all-in-focus image fusion algorithm based on edge-enhanced guided filtering is proposed. Through multi-scale decomposition of the digital refocused images and guided filtering optimization of the feature layer decision map, the final all-in-focus image is obtained. Compared with the traditional fusion algorithm, the edge information loss caused by the 4D light field calibration error is compensated in the presented method. In the step of multi-scale decomposition of the refocused image, the edge layer extraction is added to accomplish the high-frequency information enhancement. Then the multi-scale evaluation model is established to optimize the edge layer’s guided filtering parameters to obtain a better light field all-in-focus image. The experimental results show that the edge intensity and the perceptual sharpness of the all-in-focus image can be improved without significantly reducing the similarity between the all-in-focus image and the original image.
2020, 42(9): 2302-2311.
doi: 10.11999/JEIT190608
Abstract:
The color-ring resistor is one of the most commonly used electronic components in Printed Circuit Board (PCB). It is featured by sequential color rings, which often brings assembling errors, however. Manual detection of color-ring resistors has low efficiency and high false detection rate. Traditional image-based automatic detection methods have difficulties in dealing with PCB images under various illuminations, imaging distance and views. To solve this problem, an automatic detection and localization method for PCB color-ring resistor is proposed based on Convolution Neural Network (CNN). Firstly, the encoder-decoder CNN model is established and trained using weighted cross-entropy loss function. With CNN, color-ring resistors are segmented from PCB images with complex illumination and scenes. Secondly, each color-ring resistor is localized using minimum area bounding rectangle, and its position is adjusted to the vertical direction by affine transformation. Finally, the localization of color rings on the resistor is achieved by Gaussian template matching. The proposed method is tested and verified by 1270 PCB images, and the result is compared with that of the traditional method (method based on geometric contour, and method based on template matching). It is shown that the proposed method has obvious advantages in performance indices, including recall rate, precision, and intersection of unions, which can meet the requirements of practical applications.
The color-ring resistor is one of the most commonly used electronic components in Printed Circuit Board (PCB). It is featured by sequential color rings, which often brings assembling errors, however. Manual detection of color-ring resistors has low efficiency and high false detection rate. Traditional image-based automatic detection methods have difficulties in dealing with PCB images under various illuminations, imaging distance and views. To solve this problem, an automatic detection and localization method for PCB color-ring resistor is proposed based on Convolution Neural Network (CNN). Firstly, the encoder-decoder CNN model is established and trained using weighted cross-entropy loss function. With CNN, color-ring resistors are segmented from PCB images with complex illumination and scenes. Secondly, each color-ring resistor is localized using minimum area bounding rectangle, and its position is adjusted to the vertical direction by affine transformation. Finally, the localization of color rings on the resistor is achieved by Gaussian template matching. The proposed method is tested and verified by 1270 PCB images, and the result is compared with that of the traditional method (method based on geometric contour, and method based on template matching). It is shown that the proposed method has obvious advantages in performance indices, including recall rate, precision, and intersection of unions, which can meet the requirements of practical applications.
2020, 42(9): 2312-2318.
doi: 10.11999/JEIT190592
Abstract:
A 4.5 bit sub-stage circuit for high speed high precision charge domain pipelined Analog-to-Digital Converter (ADC) is proposed. Instead of the high-performance opamps used in traditional switched-capacitor pipelined ADCs, charge transfer and residue charge calculation is realized with Boosted Charge Transfer (BCT) circuit in the proposed 4.5 bit sub-stage. Therefore, the power consumption of the 4.5 bit sub-stage circuit can be reduced remarkably. The proposed 4.5 bit sub-stage circuit is used as the 1st stage circuit for a 14 bit 210 MS/s charge domain pipelined ADC and realized in a 1P6M 0.18 μm CMOS process. Test results show the 14 bit 210 MS/s ADC achieves the signal-to-noise ratio of 71.5 dBFS and the spurious free dynamic range of 85.4 dB, with 30.1 MHz input single tone signal at 210 MS/s, while the ADC core consumes the power consumption of 205 mW and occupies an area of 3.2 mm2.
A 4.5 bit sub-stage circuit for high speed high precision charge domain pipelined Analog-to-Digital Converter (ADC) is proposed. Instead of the high-performance opamps used in traditional switched-capacitor pipelined ADCs, charge transfer and residue charge calculation is realized with Boosted Charge Transfer (BCT) circuit in the proposed 4.5 bit sub-stage. Therefore, the power consumption of the 4.5 bit sub-stage circuit can be reduced remarkably. The proposed 4.5 bit sub-stage circuit is used as the 1st stage circuit for a 14 bit 210 MS/s charge domain pipelined ADC and realized in a 1P6M 0.18 μm CMOS process. Test results show the 14 bit 210 MS/s ADC achieves the signal-to-noise ratio of 71.5 dBFS and the spurious free dynamic range of 85.4 dB, with 30.1 MHz input single tone signal at 210 MS/s, while the ADC core consumes the power consumption of 205 mW and occupies an area of 3.2 mm2.
