Search

This paper investigates the robust and reliable transmission problem in Internet of Things (IoT) applications, where multiple gateways are deployed. We discover that the reliable routing path with the best link qualities may not always gain the reliable transmission. The main reason is that the majority of existing routing metrics generally do not consider the working state of gateways. And the gateways may significantly reduce the reliability of data transmission at the last hop when they operate at the variable duty cycles (e.g., due to insufficient energy harvesting from ambiance). Last-hop data loss will lead to the inefficient transmission in all previous hops. To address this issue, we propose a novel routing metric ETD (Expected Transmission Direction, ETD), which efficiently selects a proper set of gateways with improved reliability in variable duty-cycled IoT through estimating the working state of gateways. Based on ETD, we design an efficient opportunistic routing protocol PoR to ensure reliable data transmission. Our simulations demonstrate the superior performance of PoR. It is shown that PoR achieves over 98% packet delivery ratio even in the worst network setting, with effective load balancing among selected gateways.

With the fast increasing popularity of smart device communication technologies, the wireless networks on mobile sensing applications have received much attention. Wireless Sensor Networks (WSNs) with a strip structure are ubiquitous in real world deployments, such as pipeline monitoring, water quality monitoring as well as Great Wall monitoring. However, the existing routing methods will select the next-hop node that deviates from the transmission direction to sink node in strip networks with high curvature, leading to the high communication failure rate and energy consumption. To this end, we propose a new geographic routing sensing opportunistic approach, named EasyGo, to cope with the routing problem, i.e., the transmission success rate decreases in the complicated strip networks. Specifically, by investigating the transmission direction, we propose a new candidate selection algorithm SLS, which introduces the concepts of layer slicing and virtual sinks to improve the transmission success rate in strip WSNs. Theoretical analysis and extensive simulations illustrate the high efficiency and transmission performance of the proposed EasyGo strategy for strip WSNs. Furthermore, we implement the EasyGo on the testbed with Z-Stack™ nodes. Compared with the classic algorithms, our EasyGo improves the transmission success rate by up to 10%, reduces the communication overhead and the energy consumption rate by up to 11.8% and 5%, respectively.

Sensing and communication are foundations of the Internet of Things(IoT). Although energy efficiency is an important issue in MAC protocol design for general energy limited sensor networks, throughput is non trivial for some specific sensor networks. In this paper, we propose a new duty cycling scheme called MaxPut that can maximize throughput without sacrificing energy efficiency through the appropriate combination of random and scheduled duty cycling schemes. MaxPut attempts to identify risky nodes and enables risky nodes to maximize the utilization of active periods of their neighbors such that MaxPut can avoid potential buffer overflow due to aggregation of bursty data. We obtain overall throughput in networks with homogeneous and heterogeneous event occurrence processes respectively. Further, we compare MaxPut against a fully random duty cycling scheme. The simulation results show that MaxPut outperforms prior work with respect to the network throughput, while energy consumption is almost equivalent to the existing duty cycling scheme.

Code virtualization built upon virtual machine (VM) technologies is emerging as a viable method for implementing code obfuscation to protect programs against unauthorized analysis. State-of-the-art VM-based protection approaches use a fixed scheduling structure where the program always follows a single, deterministic execution path for the same input. Such approaches, however, are vulnerable in certain scenarios where the attacker can reuse knowledge extracted from previously seen software to crack applications protected with the same obfuscation scheme. This paper presents Dsvmp, a novel VM-based code obfuscation approach for software protection. Dsvmp brings together two techniques to provide stronger code protection than prior VM-based approaches. Firstly, it uses a dynamic instruction scheduler to randomly direct the program to execute different paths without violating the correctness across different runs. By randomly choosing the program execution path, the application exposes diverse behavior, making it much more difficult for an attacker to reuse the knowledge collected from previous runs or similar applications to launch an attack. Secondly, it employs multiple VMs to further obfuscate the mapping from VM opcode to native machine instructions, so that the same opcode could be mapped to different native instructions at runtime, making code analysis even harder. We have implemented Dsvmp in a prototype system and evaluated it using a set of widely used applications. Experimental results show that Dsvmp provides stronger protection with comparable runtime overhead and code size, when it is compared to two commercial VM-based code obfuscation tools.

Generative Adversarial Networks (GANs) have made great progress in cross-domain image translation. In fact, image-to-image translation tasks often encounter structural differences in two domains, such as translation on unpaired Chinese calligraphy dataset. However, existing models can only convert color and texture features and keep the structures unchanged (e.g.: in apples to oranges tasks, these models only convert the color of apples, but maintain the shape of apples). In order to address cross-structure image translation, such as cross-structure translation of Chinese calligraphy, a novel Generative Adversarial Networks (GAN) model, named CS-GAN, is proposed in this paper. In CS-GAN, distribution transform, reparameterization trick and sampling features are used to convert feature maps obtained from domain S to domain T . Then images of domain T are generated through features concatenation. The proposed CS-GAN is verified on three sets of Chinese calligraphic data with structural differences from three famous calligraphers, Yan Zhenqing, Zhao Mengfu and Ouyang Xun. The extensive experimental results show that the proposed CS-GAN successfully transforms the Chinese calligraphy data of different structures and outperforms the state of art models.

Content caching is emerging as a vital mechanism of Wireless Cellular Network. It could reduce the content transmission distance and balance the network load efficiently. Current content caching systems have exploited the content popularity with temporal locality and users’ interests to solving the major questions: when and what to cache? But cannot point out where to cache. Past proposals for addressing these problems rely on online web caching and distributed storage systems, these works focus on wired networks or wireless networks that handsets (also called users) are independent and hence perform inefficiently when faced with complicated correlation in the practical social network, where the request probabilities of the handsets are easily affected by the others. In this paper, we introduce a wireless network content caching algorithm called PreCache that consider the social influence on users’ content requests. Our design exploit social correlation to content popularity, the intuition underlying our design is that the content request probability of handset could be affected by others handsets’ action in a same social network, and thus the content popularity exhibits a feature of spatial locality in the social network, we called the feature as virtual spatial locality. We adopt virtual spatial locality of content popularity for the change of content request probability by using a social influence model. We then design the heuristic algorithm PreCache for the selection of base stations, in which the contents are cached. We evaluate the performance of the PreCache in different network setups. The results show that PreCache could improve the successful content request ratio up to 90 % , and gain an average of 2 hop count reduction compared with ProbCache and Random Caching.

Multi-user concurrent communication has become an increasingly important issue in mobile computing. Emerging vibration-based communication provides a new multi-to-single (M2S) data communication model for face-to-face tasks as opposed to most existing device-to-device (D2D) communication technologies. However, the transmission of multiple concurrent vibration signals often results in a highly coupled received signal, which makes it very difficult to reconstruct the source signals. This paper proposes ViCom, a vibration-based communication system that performs parallel decoding to achieve M2S transmission between mobile devices. The core of ViCom is a Decomposition Transformation Diagonal Algorithm (DTDA) that decomposes the received signal and constructs an estimated set of source signals, which are first decoupled by an autocorrelation function and then diagonalized to further improve the signal reconstruction accuracy. We implement ViCom using vibration motors and accelerometers embedded in existing smart phones and wearable devices, and evaluate its performance over different distances through disparate transmission media. Experimental results show that ViCom achieves effective data transmission of eight users with an accuracy of 98.6% and exhibits a superior overall performance over existing methods.

In optical character recognition, text strings should be extracted from images first. But only the complete text strings can accurately express the meanings of the words, so the extracted individual characters should be grouped into text strings before recognition. There are lots of text strings in topographic maps, and these texts consist of the characters with multi-colored, multi-sized and multi-oriented, and the existing methods cannot effectively group them. In this paper, a dynamic character grouping method is proposed to group the characters into text strings based on four consistency constraints, which are the color, size, spacing and direction respectively. As we know that the characters in the same word have similar colors, sizes and distances between them, and they are also on some curve lines with a certain bending, but the characters in different words are not. Based on these features of the characters, the background pixels around the characters are expanded to link the characters into text strings. In this method, due to the introduction of the color consistency constraint, the characters with different colors can be grouped well. And this method can deal with curved character strings more accurately by the improved direction consistency constraint. The final experimental results show that this method can group the characters more efficiently, especially for the case in which the beginning or the end characters of words are close to the characters of the other words.

By restoring the program into an easier understandable form, deobfuscation is an important technique for detecting and analyzing malicious software. To enable deobfuscation, one must know if the target program is obfuscated and what types of obfuscation schemes may be used. However, obtaining such information is challenging without having access to the original program source code. This paper presents a new way to estimate the obfuscation scheme of a compiled binary. It achieves this by using semantic information of the disassembled binary to predict if the program has been obfuscated and if so, what type of obfuscation scheme may be used. At the core of our approach is a set of deep neural networks that can effectively characterize and leverage the contextual information available in the assembly code. Our models are first trained offline, and the learned models can then be applied to new previously unseen obfuscated binaries. We evaluate our approach by applying it to a large dataset of over 277,000 obfuscated samples with different individual obfuscation schemes and their combinations. Experimental results show that our approach is highly effective in identifying the obfuscation scheme, with a prediction accuracy of at least 83% (up to 98%).

In this paper, a multi-direction virtual array transformation algorithm termed MVATA for two dimension (2D) direction of arrival (DOA) estimation is proposed. A kind of multi-direction rotation invariant virtual subarrays is firstly constructed. Secondly, by singular value decomposition (SVD) of the response matrix of the real array to signals from a sector, the interpolation matrices can be estimated and from which the real and virtual signal subspaces are further estimated. Thirdly, a rotational invariant factor matrix is derived from the real and virtual subspaces base on least squares. Subsequently, the azimuth and elevation can be solved through the rotational invariant factor matrix without spectral peak search. Theoretical calculation and time-consuming statistics show that the proposed MVATA has the advantage of computational complexity compared with the conventional interpolated methods. Moreover, the proposed MVATA can be applied to arbitrary structure array, and the estimated 2D angles can be paired automatically. Simulation results demonstrate the proposed MVATA possesses significantly accuracy improvement especially in low SNR and small snapshots. Proposed algorithm need not the support of pairing algorithm.Complexity is reduced without spectrum search and fractal dimension processing.Accuracy is obviously improved especially in low SNR and small snapshots.2D angels have the same azimuth or elevation can be distinguished.Proposed can be applied not only to UCA but also to arbitrary structure arrays.

In topographic maps, only the complete text strings can accurately express the properties of the geographic elements, so individual characters should be grouped into text strings before recognition. This paper presents a novel character grouping method based on the graph model. In this method, undirected graphs are used to describe different words, where the color and size of the characters are served as the properties of the nodes, while the distance and angle between the characters are served as the weights of the edges connecting pairs of characters. Therefore, the nodes can be connected to construct undirected graphs according to their properties. Then the constructed graphs are simplified according to the weights of the edges. Finally, we can get the final results corresponding to the grouped characters. Experimental results show that this method can especially group the characters with significant wide spacing. Moreover, it has higher efficiency with graphic processing instead of image processing.

Image segmentation for topographic maps is challenging due to their low quality, high degrees of mixed and false coloring. Besides, many pixels cannot be explicitly separated from each other because of the fuzziness of their colors. Therefore the algorithms based on fuzzy theory are suitable to process such images utilizing their ability to deal with the blurring effect. However, there are still some problems of large-scale data, high time complexity and inaccurate classification. In order to overcome these problems, we propose a novel algorithm for segmenting large topographic maps based on the ideas of fuzzy theory, randomized sampling and multilevel image fusion. In this algorithm, the large topographic map is randomly sampled first. Then the optimal clustering centers are acquired by fuzzy C-means (FCM) clustering. Further, the map is segmented with the fuzzy classification method. Finally, multilevel image fusion is used to fuse the segmented images into the final segmentation maps. Randomized sampling is used to reduce the amount of data, and improve the efficiency of image segmentation. Multilevel fusion can make the final segmentation maps more accurate. The experiments show that our method has higher efficiency and accuracy than the traditional ones. It provides a reliable image segmentation method for large topographic maps.

Infringing intellectual property by reverse analysis is a severe threat to Android applications. By replacing the program instructions with virtual instructions that an adversary is unfamiliar with, code obfuscation based on virtualization is a promising way of protecting Android applications against reverse engineering. However, the current code virtualization approaches for Android only target at the DEX bytecode level. The DEX file with the open file format and more semantic information makes the decode-dispatch pattern easier to expose, which has been identified as a severe vulnerability of security and can be exploited by various attacks. Further, decode-dispatch interpretation frequently uses indirect branches in this structure to introduce extra overhead. This paper presents a novel approach to transfer code virtualization from DEX level to native level, which possesses strong security strength and good stealth, with only modest cost. Our approach contains two components: pre-compilation and compile-time virtualization. Pre-compilation is designed for performance improvement by identifying and decompiling the critical functions which consume a significant fraction of execution time. Compile-time virtualization builds upon the widely used LLVM compiler framework. It automatically translates the DEX bytecode into the common LLVM intermediate representations where a unified code virtualization pass can be applied for DEX code. We have implemented a working prototype Dex2VM of our technique and applied it to eight representative Android applications. Our experimental results show that the proposed approach can effectively protect the target code against a state-of-the-art code reverse engineering tool that is specifically designed for code virtualization, and it achieves good stealth with only modest cost.