Privacy-Preserving Content-Aware Search Based on Two-Level Index

Nowadays,cloud computing is used more and more widely,more and more people prefer to using cloud server to store data.So,how to encrypt the data efficiently is an important problem.The search efficiency of existed search schemes decreases as the index increases.For solving this problem,we build the two-level index.Simultaneously,for improving the semantic information,the central word expansion is combined.The purpose of privacy-preserving content-aware search by using the two-level index(CKESS)is that the first matching is performed by using the extended central words,then calculate the similarity between the trapdoor and the secondary index,finally return the results in turn.Through experiments and analysis,it is proved that our proposed schemes can resist multiple threat models and the schemes are secure and efficient.

Dynamic Resource Scheduling in Emergency Environment

Nowadays,emergency accidents could happen at any time.The accidents occur unpredictably and the accidents requirements are diversely.The accidents happen in a dynamic environment and the resource should be cooperative to solve the accidents.Most methods are focusing on minimizing the casualties and property losses in a static environment.However,they are lack in considering the dynamic and unpredictable event handling.In this paper,we propose a representative environmental model in representation of emergency and dynamic resource allocation model,and an adaptive mathematical model based on Genetic Algorithm(GA)to generate an optimal set of solution domain.The experimental results show that the proposed algorithm can get a set of better candidate solutions.

An LSTM-Based Malware Detection Using Transfer Learning

Mobile malware occupies a considerable proportion of cyberattacks.With the update of mobile device operating systems and the development of software technology,more and more new malware keep appearing.The emergence of new malware makes the identification accuracy of existing methods lower and lower.There is an urgent need for more effective malware detection models.In this paper,we propose a new approach to mobile malware detection that is able to detect newly-emerged malware instances.Firstly,we build and train the LSTM-based model on original benign and malware samples investigated by both static and dynamic analysis techniques.Then,we build a generative adversarial network to generate augmented examples,which can emulate the characteristics of newly-emerged malware.At last,we use the augmented examples to retrain the 4th and 5th layers of the LSTM network and the last fully connected layer so that it can discriminate against newly-emerged malware.Actual experiments show that our malware detection achieved a classification accuracy of 99.94%when tested on augmented samples and 86.5%with the samples of newly-emerged malware on real data.

Multi-UAV Cooperative GPS Spoofing Based on YOLO Nano

In recent years,with the rapid development of the drone industry,drones have been widely used in many fields such as aerial photography,plant protection,performance,and monitoring.To effectively control the unauthorized flight of drones,using GPS spoofing attacks to interfere with the flight of drones is a relatively simple and highly feasible attack method.However,the current method uses ground equipment to carry out spoofing attacks.The attack range is limited and the flexibility is not high.Based on the existing methods,this paper proposes a multi-UAV coordinated GPS spoofing scheme based on YOLO Nano,which can launch effective attacks against target drones with autonomous movement:First,a single-attack drone based on YOLO Nano is proposed.The target tracking scheme achieves accurate tracking of the target direction on a single-attack drone;then,based on the single-UAV target tracking,a multi-attack drone coordinated target tracking scheme based on the weighted least squares method is proposed to realize the target drone Finally,a new calculation method for false GPS signals is proposed,which adaptively adjusts the flight trajectory of the attacking drone and the content of the false GPS signal according to the autonomous movement of the target drone.

Privacy-Preserving Genetic Algorithm Outsourcing in Cloud Computing

Genetic Algorithm(GA)has been widely used to solve various optimization problems.As the solving process of GA requires large storage and computing resources,it is well motivated to outsource the solving process of GA to the cloud server.However,the algorithm user would never want his data to be disclosed to cloud server.Thus,it is necessary for the user to encrypt the data before transmitting them to the server.But the user will encounter a new problem.The arithmetic operations we are familiar with cannot work directly in the ciphertext domain.In this paper,a privacy-preserving outsourced genetic algorithm is proposed.The user’s data are protected by homomorphic encryption algorithm which can support the operations in the encrypted domain.GA is elaborately adapted to search the optimal result over the encrypted data.The security analysis and experiment results demonstrate the effectiveness of the proposed scheme.

A Two-Stage Highly Robust Text Steganalysis Model

With the development of natural language processing,deep learning,and other technologies,text steganography is rapidly developing.However,adversarial attack methods have emerged that gives text steganography the ability to actively spoof steganalysis.If terrorists use the text steganography method to spread terrorist messages,it will greatly disturb social stability.Steganalysis methods,especially those for resisting adversarial attacks,need to be further improved.In this paper,we propose a two-stage highly robust model for text steganalysis.The proposed method analyzes and extracts anomalous features at both intra-sentential and inter-sentential levels.In the first phase,every sentence is first transformed into word vectors.To obtain a high dimensional sentence vector,we use Bi-LSTM to obtain feature information for all words in the sentence while retaining strong correlations.In the second phase,we input multiple sentences vectors into the GNN,from which we extract inter-sentential anomaly features and make a judgment as to whether the text contains secret messages.In addition,to improve the robustness of the model,we add adversarial examples to the training set to improve the robustness and generalization of the steganalysis model.Theoretically,our proposed method is more robust and more accurate in detection compared to existing methods.

High Visual Quality Image Steganography Based on Encoder-Decoder Model

Nowadays,with the popularization of network technology,more and more people are concerned about the problem of cyber security.Steganography,a technique dedicated to protecting peoples’private data,has become a hot topic in the research field.However,there are still some problems in the current research.For example,the visual quality of dense images generated by some steganographic algorithms is not good enough;the security of the steganographic algorithm is not high enough,which makes it easy to be attacked by others.In this paper,we propose a novel high visual quality image steganographic neural network based on encoder-decoder model to solve these problems mentioned above.Firstly,we design a novel encoder module by applying the structure of U-Net++,which aims to achieve higher visual quality.Then,the steganalyzer is heuristically added into the model in order to improve the security.Finally,the network model is used to generate the stego images via adversarial training.Experimental results demonstrate that our proposed scheme can achieve better performance in terms of visual quality and security.

A Review of Object Detectors in Deep Learning

Object detection is one of the most fundamental,longstanding and significant problems in the field of computer vision,where detection involves object classification and location.Compared with the traditional object detection algorithms,deep learning makes full use of its powerful feature learning capabilities showing better detection performance.Meanwhile,the emergence of large datasets and tremendous improvement in computer computing power have also contributed to the vigorous development of this field.In the paper,many aspects of generic object detection are introduced and summarized such as traditional object detection algorithms,datasets,evaluation metrics,detection frameworks based on deep learning and state-of-the-art detection results for object detectors.Finally,we discuss several promising directions for future research.

Semantic and secure search over encrypted outsourcing cloud based on BERT

Searchable encryption provides an effective way for data security and privacy in cloud storage.Users can retrieve encrypted data in the cloud under the premise of protecting their own data security and privacy.However,most of the current content-based retrieval schemes do not contain enough semantic information of the article and cannot fully reflect the semantic information of the text.In this paper,we propose two secure and semantic retrieval schemes based on BERT(bidirectional encoder representations from transformers)named SSRB-1,SSRB-2.By training the documents with BERT,the keyword vector is generated to contain more semantic information of the documents,which improves the accuracy of retrieval and makes the retrieval result more consistent with the user’s intention.Finally,through testing on real data sets,it is shown that both of our solutions are feasible and effective.