Biblio

The traditional approach to distributed deep neural network (DNN) training is data-distributed learning, which partitions and distributes data to workers. This approach, although has good convergence properties, has high communication cost, which puts a strain especially on edge systems and increases delay. An emerging approach is model-distributed learning, where a training model is distributed across workers. Model-distributed learning is a promising approach to reduce communication and storage costs, which is crucial for edge systems. In this paper, we design ResPipe, a novel resilient model-distributed DNN training mechanism against delayed/failed workers. We analyze the communication cost of ResPipe and demonstrate the trade-off between resiliency and communication cost. We implement ResPipe in a real testbed consisting of Android-based smartphones, and show that it improves the convergence rate and accuracy of training for convolutional neural networks (CNNs).

Internet of Things (IoT), Cloud and Augmented Reality (AR) are the emerging and developing technologies and are at the horizon and hype of their life cycle. Lots of commercial applications based on IoT, cloud and AR provide unrestricted access to data. The real-time applications based on these technologies are at the cusp of their innovations. The most frequent security attacks for IoT, cloud and AR applications are DDoS attacks. In this paper a detailed account of various DDoS attacks that can be the hindrance of many important sensitive services and can degrade the overall performance of recent services which are purely based on network communications. The DDoS attacks should be dealt with carefully and a set of a new generations of algorithm need to be developed to mitigate the problems caused by non-repudiation kinds of attacks.

With the rapid development of the internet of things and cloud computing, users can instantly transmit a large amount of data to various fields, with the development of communication technology providing convenience for people's life, information security is becoming more and more important. Therefore, it is of great significance to study the technology of image hiding information detection. This paper mainly uses the support vector machine learning algorithm to detect the hidden information of the image, based on a standard image library, randomly selecting images for embedding secret information. According to the bit-plane correlation and the gradient energy change of a single bit-plane after encryption of an image LSB matching algorithm, gradient energy change is selected as characteristic change, and the gradient energy change is innovatively applied to a support vector machine classifier algorithm, and has very good detection effect and good stability on the dense image with the embedding rate of more than 40 percent.

Many techniques are available nowadays, for Intellectual Property(IP) protection of Digital circuits. Out of these techniques, the popular one is watermarking. Similar to the watermarking used in case of text, image and video, watermarking of digital circuits also modifies a digital circuit design in such a way, that only the IP owner of design is able to extract the watermark form the design. In this paper, Multi – Feedback configuration of Linear Feedback Shift Register(LFSR) is used to watermark a FSM based design. This watermarking technique improves the watermark strength of already existing LFSR based watermarking technique. In terms of hardware utilization, it is significantly efficient than some popular watermarking techniques. The proposed technique has been implemented using Verilog HDL in Xilinx ISE and the simulation is done using ModelSim.

Ternary content addressable memories (TCAMs) widely utilized in network systems to enforce the labeling of packets. For example, they are used for packet forwarding, security, and software-defined networks (SDNs). TCAMs are typically deployed as standalone instruments or as an embedded intellectual property component on application-specific integrated circuits. However, field-programmable gate arrays (FPGAs) do not have TCAM bases. However, FPGAs’ versatility allows them to appeal for SDN deployment, and most FPGA vendors have SDN production kits. Those need to help TCAM features and then simulate TCAMs using the FPGA logic blocks. Several methods to reproduction TCAMs on FPGAs have been introduced in recent years. Some of them use a huge multiple storage blocks within modern FPGAs to incorporate TCAMs. A trouble while remembrances are that soft errors that corrupt stored bits can affect them. Memories may be covered by a parity test to identify errors or by an error correction code, although this involves extra bits in a word frame. This brief considers memory security used to simulate TCAMs. It is shown in particular that by leveraging the assumption its part of potential memory information is true, most single-bit errors can be resolved when memoirs are emulated with a parity bit.

We propose and experimentally demonstrate a compressive sensing scheme based on optical coherent receiver that recovers sparse optical arbitrary signals with an analog bandwidth up to 25GHz. The proposed scheme uses 16x lower sampling rate than the Nyquist theorem and spectral resolution of 24.4MHz.

The power cyber-physical system combines the cyber network with the traditional electrical power network, which can monitor and control the operation of the power grid stably and efficiently. Since the system's structure and function is complicated and large, it becomes fragile as a result. Therefore, establishing a reasonable and effective CPS model and discussing its vulnerability performance under external attacks is essential and vital for power grid operation. This paper uses the theory of complex networks to establish a independent system model by IEEE-118-node power network and 200-node scale-free information network, introducing information index to identify and sort important nodes in the network, and then cascade model of the power cyber-physical system based on the node load capacity is constructed and the vulnerability assessment analysis is carried out. The simulation shows that the disintegration speed of the system structure under deliberate attacks is faster than random attacks; And increasing the node threshold can effectively inhibit the propagation of failure.

The paradigm of quantum computation has led to the development of new algorithms as well variations on existing algorithms. In particular, novel cryptographic techniques based upon quantum computation are of great interest. Many classical encryption techniques naturally translate into the quantum paradigm because of their well-structured factorizations and the fact that they can be phased in the form of unitary operators. In this work, we demonstrate a quantum approach to data encryption and decryption based upon the McEliece cryptosystem using Reed-Muller codes. This example is of particular interest given that post-quantum analyses have highlighted this system as being robust against quantum attacks. Finally, in anticipation of quantum computation operating over binary fields, we discuss alternative operator factorizations for the proposed cryptosystem.

This paper considers a multivariate quickest detection problem with false data injection (FDI) attacks in internet of things (IoT) systems. We derive a sequential generalized likelihood ratio test (GLRT) for zero-mean Gaussian FDI attacks. Exploiting the fact that covariance matrices are positive, we propose strategies to detect positive semi-definite matrix additions rather than arbitrary changes in the covariance matrix. The distribution of the GLRT is only known asymptotically whereas quickest detectors deal with short sequences, thereby leading to loss of performance. Therefore, we use a finite-sample correction to reduce the false alarm rate. Further, we provide a numerical approach to estimate the threshold sequences, which are analytically intractable to compute. We also compare the average detection delay of the proposed detector for constant and varying threshold sequences. Simulations showed that the proposed detector outperforms the standard sequential GLRT detector.

We present an alternate-channel chaotic laser secure communication system to enhance information communication security and study its technical solution via combining chaos shift keying (CSK) and chaos masking (CM). Two coupled lasers and other two single lasers are introduced as a novel alternate-channel secure communication system, where one of two coupled lasers is modulated via CSK to encode a digital signal and the other of coupled lasers is used to emit a chaotic carrier to mask an information using CM. The two single lasers are used to decode CSK and CM information, respectively. And such CSK performance results in enhancement of CM secure performance because of in-time variation of the emitter' parameter as secret keys. The obtained numerical results show that the encoding and decoding can be successfully performed. The study is beneficial to chaotic cryptography and optics secure communication.

When biometric databases grow larger, a security breach or leak can affect millions. In order to protect against such a threat, the use of encryption is a natural choice. However, a biometric identification attempt then requires the decryption of a potential huge database, making a traditional approach potentially unfeasible. The use of selective JPEG2000 encryption can reduce the encryption’s computational load and enable a secure storage of biometric sample data. In this paper we will show that selective encryption of face biometric samples is secure. We analyze various encoding settings of JPEG2000, selective encryption parameters on the "Labeled Faces in the Wild" database and apply several traditional and deep learning based face recognition methods.

An open source software package for performing dynamic RMS simulation of small to medium-sized power systems is presented, written entirely in the Python programming language. The main objective is to facilitate fast prototyping of new wide area monitoring, control and protection applications for the future power system by enabling seamless integration with other tools available for Python in the open source community, e.g. for signal processing, artificial intelligence, communication protocols etc. The focus is thus transparency and expandability rather than computational efficiency and performance.The main purpose of this paper, besides presenting the code and some results, is to share interesting experiences with the power system community, and thus stimulate wider use and further development. Two interesting conclusions at the current stage of development are as follows:First, the simulation code is fast enough to emulate real-time simulation for small and medium-size grids with a time step of 5 ms, and allows for interactive feedback from the user during the simulation. Second, the simulation code can be uploaded to an online Python interpreter, edited, run and shared with anyone with a compatible internet browser. Based on this, we believe that the presented simulation code could be a valuable tool, both for researchers in early stages of prototyping real-time applications, and in the educational setting, for students developing intuition for concepts and phenomena through real-time interaction with a running power system model.

The front-end text processing module is considered as an essential part that influences the intelligibility and naturalness of a Mandarin text-to-speech system significantly. For commercial text-to-speech systems, the Mandarin front-end should meet the requirements of high accuracy and low time latency while also ensuring maintainability. In this paper, we propose a universal BERT-based model that can be used for various tasks in the Mandarin front-end without changing its architecture. The feature extractor and classifiers in the model are shared for several sub-tasks, which improves the expandability and maintainability. We trained and evaluated the model with polyphone disambiguation, text normalization, and prosodic boundary prediction for single task modules and multi-task learning. Results show that, the model maintains high performance for single task modules and shows higher accuracy and lower time latency for multi-task modules, indicating that the proposed universal front-end model is promising as a maintainable Mandarin front-end for commercial applications.

In this work, an IF(intermediate frequency) module of a hyperspectral microwave radiometer based on a polyphase filter bank (PFB) and Discrete Fourier Transformation (DFT)is introduced. The IF module is designed with an 800MSPS sampling-rate ADC and a Xilinx Virtex-7 FPGA. The module can achieve 512 channels and a bandwidth of 400M and process all the sampled data in real-time. The test results of this module are given and analyzed, such as linearity, accuracy, etc. It can be used in various applications of microwave remote sensing. The system has strong expandability.

Digital watermarking is the multimedia leading security protection as it permanently escorts the digital content. Image copyright protection is becoming more anxious as the new 5G technology emerged. Protecting images with a robust scheme without distorting them is the main trade-off in digital watermarking. In this paper, a watermarking scheme based on discrete cosine transform (DCT) and singular value decomposition (SVD) using canny edge detector technique is proposed. A binary encrypted watermark is reshaped into a vector and inserted into the edge detected vector from the diagonal matrix of the SVD of DCT DC and low-frequency coefficients. Watermark insertion is performed by using an edge-tracing mechanism. The scheme is evaluated using the Peak Signal to Noise Ratio (PSNR) and Normalized Correlation (NC). Attained results are competitive when compared to present works in the field. Results show that the PSNR values vary from 51 dB to 55 dB.

To share the recorded ECG data with the cardiologist in Golden Hours in an efficient and secured manner via tele-cardiology may save the lives of the population residing in rural areas of a country. This paper proposes an encryption-authentication scheme for secure the ECG data. The main contribution of this work is to generate a one-time padding key and deploying an encryption algorithm in authentication mode to achieve encryption and authentication. This is achieved by a water cycle optimization algorithm that generates a completely random one-time padding key and Triple Data Encryption Standard (3DES) algorithm for encrypting the ECG data. To validate the accuracy of the proposed encryption authentication scheme, experimental results were performed on standard ECG data and various performance parameters were calculated for it. The results show that the proposed algorithm improves security and passes the statistical key generation test.

An oblivious signature is a digital signature with some property. The oblivious signature scheme has two parties, the signer and the receiver. First, the receiver can choose one and get one of n valid signatures without knowing the signer’s private key. Second, the signer does not know which signature is chosen by the receiver. In this paper, we propose the oblivious signature which is combined with blind signature and zero-knowledge set membership. The property of blind signature makes sure that the signer does not know the message of the signature by the receiver chosen, on the other hand, the property of the zero-knowledge set membership makes sure that the message of the signature by the receiver chosen is one of the set original messages.

Federated learning (FL) allows to train a massive amount of data privately due to its decentralized structure. Stochastic gradient descent (SGD) is commonly used for FL due to its good empirical performance, but sensitive user information can still be inferred from weight updates shared during FL iterations. We consider Gaussian mechanisms to preserve local differential privacy (LDP) of user data in the FL model with SGD. The trade-offs between user privacy, global utility, and transmission rate are proved by defining appropriate metrics for FL with LDP. Compared to existing results, the query sensitivity used in LDP is defined as a variable, and a tighter privacy accounting method is applied. The proposed utility bound allows heterogeneous parameters over all users. Our bounds characterize how much utility decreases and transmission rate increases if a stronger privacy regime is targeted. Furthermore, given a target privacy level, our results guarantee a significantly larger utility and a smaller transmission rate as compared to existing privacy accounting methods.

Power quality gains more and more attentions because disturbances in power quality may damage equipment security, power availability and system reliability in power system. Detection and classification of the power quality disturbances is the first step before taking measures to lessen their harmful effects. Common methods to classify power quality disturbances includes signal processing methods, machine learning methods and deep learning methods. Signal processing methods are good at feature extraction, while machine learning methods and deep learning methods are expert in multi-classification tasks. Via combing their respective advantages, this paper proposes a combined method based on variational mode decomposition and convolutional neural networks, which needs a small quantity of samples but achieves high classification precision. The proposed method is proved to be a qualified and competitive scheme for the detection and classification of power quality disturbances.

Security plays a major role in data transmission and reception. Providing high security is indispensable in communication systems. The RSA (Rivest-Shamir-Adleman) cryptosystem is used widely in cryptographic applications as it offers highly secured transmission. RSA cryptosystem uses Montgomery multipliers and it involves modular exponentiation process which is attained by performing repeated modular-multiplications. This leads to high latency and owing to improve the speed of multiplier, highly efficient modular multiplication methodology needs to be applied. In the conventional methodology, Carry Save Adder (CSA) is used in the multiplication and it consumes more area and it has larger delay, but in the suggested methodology, the Reverse Carry Propagate (RCP) adder is used in the place of CSA adder and the obtained output shows promising results in terms of area and latency. The simulation is done with Xilinx ISE design suite. The proposed multiplier can be used effectively in signal processing, image processing and security based applications.

In the current network security situation, the types of network threats are complex and changeable. With the development of the Internet and the application of information technology, the general trend is opener. Important data and important business applications will face more serious security threats. However, with the development of cloud computing technology, the trend of large-scale deployment of important business applications in cloud centers has greatly increased. The development and use of software-defined networks in cloud data centers have greatly reduced the effect of traditional network security boundary protection. How to find an effective way to protect important applications in open multi-step large-scale cloud data centers is a problem we need to solve. Threat intelligence has become an important means to solve complex network attacks, realize real-time threat early warning and attack tracking because of its ability to analyze the threat intelligence data of various network attacks. Based on the research of threat intelligence, machine learning, cloud central network, SDN and other technologies, this paper proposes an active defense method of network security based on threat intelligence for super-large cloud data centers.

Wireless communication systems are inherently vulnerable to adversarial attacks since malevolent jammers might jam and disrupt the legitimate transmission intentionally. Of particular interest are so- called denial-of-service (DoS) attacks in which the jammer is able to completely disrupt the communication. Accordingly, it is of crucial interest for the legitimate users to detect such DoS attacks. Turing machines provide the fundamental limits of today's digital computers and therewith of the traditional signal processing. It has been shown that these are incapable of detecting DoS attacks. This stimulates the question of how powerful the signal processing must be to enable the detection of DoS attacks. This paper investigates the general computation framework of Blum-Shub-Smale machines which allows the processing and storage of arbitrary reals. It is shown that such real number signal processing then enables the detection of DoS attacks.

A method for providing energy and structural secrecy of a signal is presented, which is based on the method of pseudo-random restructuring of the spreading sequence. This method complicates the implementation of the accumulation mode, and therefore the detection of the signal-code structure of the signal in a third-party receiver, due to the use of nested pseudo-random sequences (PRS) and their restructuring. And since the receiver-detector is similar to the receiver of the communication system, it is necessary to ensure optimal signal processing to implement an acceptable level of structural secrecy.

With the widespread deployment of Internet of Things (IoT) devices, hackers can use IoT devices to launch large-scale distributed denial of service (DDoS) attacks, which bring great harm to the Internet. However, how to defend against these attacks remains to be an open challenge. In this paper, we propose a novel prevention method for IoT DDoS attacks based on blockchain and obfuscation of IP addresses. Our observation is that IoT devices are usually resource-constrained and cannot support complicated cryptographic algorithms such as RSA. Based on the observation, we employ a novel authentication then communication mechanism for IoT DDoS attack prevention. In this mechanism, the attack targets' IP addresses are encrypted by a random security parameter. Clients need to be authenticated to obtain the random security parameter and decrypt the IP addresses. In particular, we propose to authenticate clients with public-key cryptography and a blockchain system. The complex authentication and IP address decryption operations disable IoT devices and thus block IoT DDoS attacks. The effectiveness of the proposed method is analyzed and validated by theoretical analysis and simulation experiments.

The Internet of Things (IoT) is clubbed by networking of sensors and other embedded electronics. As more devices are getting connected, the vulnerability of getting affected by various IoT threats also increases. Among the IoT threads, DDoS attacks are causing serious issues in recent years. In IoT, these attacks are challenging to detect and isolate. Thus, an effective Intrusion Detection System (IDS) is essential to defend against these attacks. The traditional IDS is based on manual blacklisting. These methods are time-consuming and will not be effective to detect novel intrusions. At present, IDS are automated and programmed to be dynamic which are aided by machine learning & deep learning models. The performance of these models mainly depends on the data used to train the model. Majority of IDS study is performed with non-compatible and outdated datasets like KDD 99 and NSL KDD. Research on specific DDoS attack datasets is very less. Therefore, in this paper, we first aim to examine the effect of existing datasets in the IoT environment. Then, we propose a real-time data collection framework for DNS amplification attacks in IoT. The generated network packets containing DDoS attack is captured through port mirroring.