arXiv subjects

Issa Khalil

Publications and source records attributed to Issa Khalil.

46 records · Page 3Linked to original sources

arXiv2019

ZK-GanDef: A GAN based Zero Knowledge Adversarial Training Defense for Neural Networks

Neural Network classifiers have been used successfully in a wide range of applications. However, their underlying assumption of attack free environment has been defied by adversarial examples. Researchers tried to develop defenses; however, existing approaches are still far from providing effective solutions to this evolving problem. In this paper, we design a generative adversarial net (GAN) based zero knowledge adversarial training defense, dubbed ZK-GanDef, which does not consume adversarial examples during training. Therefore, ZK-GanDef is not only efficient in training but also adaptive to new adversarial examples. This advantage comes at the cost of small degradation in test accuracy compared to full knowledge approaches. Our experiments show that ZK-GanDef enhances test accuracy on adversarial examples by up-to 49.17% compared to zero knowledge approaches. More importantly, its test accuracy is close to that of the state-of-the-art full knowledge approaches (maximum degradation of 8.46%), while taking much less training time.

cs.LG↗

arXiv2019

GanDef: A GAN based Adversarial Training Defense for Neural Network Classifier

Machine learning models, especially neural network (NN) classifiers, are widely used in many applications including natural language processing, computer vision and cybersecurity. They provide high accuracy under the assumption of attack-free scenarios. However, this assumption has been defied by the introduction of adversarial examples -- carefully perturbed samples of input that are usually misclassified. Many researchers have tried to develop a defense against adversarial examples; however, we are still far from achieving that goal. In this paper, we design a Generative Adversarial Net (GAN) based adversarial training defense, dubbed GanDef, which utilizes a competition game to regulate the feature selection during the training. We analytically show that GanDef can train a classifier so it can defend against adversarial examples. Through extensive evaluation on different white-box adversarial examples, the classifier trained by GanDef shows the same level of test accuracy as those trained by state-of-the-art adversarial training defenses. More importantly, GanDef-Comb, a variant of GanDef, could utilize the discriminator to achieve a dynamic trade-off between correctly classifying original and adversarial examples. As a result, it achieves the highest overall test accuracy when the ratio of adversarial examples exceeds 41.7%.

cs.LG↗

arXiv2018

A Survey on Malicious Domains Detection through DNS Data Analysis

Malicious domains are one of the major resources required for adversaries to run attacks over the Internet. Due to the important role of the Domain Name System (DNS), extensive research has been conducted to identify malicious domains based on their unique behavior reflected in different phases of the life cycle of DNS queries and responses. Existing approaches differ significantly in terms of intuitions, data analysis methods as well as evaluation methodologies. This warrants a thorough systematization of the approaches and a careful review of the advantages and limitations of every group. In this paper, we perform such an analysis. In order to achieve this goal, we present the necessary background knowledge on DNS and malicious activities leveraging DNS. We describe a general framework of malicious domain detection techniques using DNS data. Applying this framework, we categorize existing approaches using several orthogonal viewpoints, namely (1) sources of DNS data and their enrichment, (2) data analysis methods, and (3) evaluation strategies and metrics. In each aspect, we discuss the important challenges that the research community should address in order to fully realize the power of DNS data analysis to fight against attacks leveraging malicious domains.

cs.CR↗

arXiv2018

Unmanned Aerial Vehicles: A Survey on Civil Applications and Key Research Challenges

The use of unmanned aerial vehicles (UAVs) is growing rapidly across many civil application domains including real-time monitoring, providing wireless coverage, remote sensing, search and rescue, delivery of goods, security and surveillance, precision agriculture, and civil infrastructure inspection. Smart UAVs are the next big revolution in UAV technology promising to provide new opportunities in different applications, especially in civil infrastructure in terms of reduced risks and lower cost. Civil infrastructure is expected to dominate the more that $45 Billion market value of UAV usage. In this survey, we present UAV civil applications and their challenges. We also discuss current research trends and provide future insights for potential UAV uses. Furthermore, we present the key challenges for UAV civil applications, including: charging challenges, collision avoidance and swarming challenges, and networking and security related challenges. Based on our review of the recent literature, we discuss open research challenges and draw high-level insights on how these challenges might be approached.

cs.RO↗

arXiv2017

Killing Two Birds with One Stone: Malicious Domain Detection with High Accuracy and Coverage

Inference based techniques are one of the major approaches to analyze DNS data and detecting malicious domains. The key idea of inference techniques is to first define associations between domains based on features extracted from DNS data. Then, an inference algorithm is deployed to infer potential malicious domains based on their direct/indirect associations with known malicious ones. The way associations are defined is key to the effectiveness of an inference technique. It is desirable to be both accurate (i.e., avoid falsely associating domains with no meaningful connections) and with good coverage (i.e., identify all associations between domains with meaningful connections). Due to the limited scope of information provided by DNS data, it becomes a challenge to design an association scheme that achieves both high accuracy and good coverage. In this paper, we propose a new association scheme to identify domains controlled by the same entity. Our key idea is an in-depth analysis of active DNS data to accurately separate public IPs from dedicated ones, which enables us to build high-quality associations between domains. Our scheme identifies many meaningful connections between domains that are discarded by existing state-of-the-art approaches. Our experimental results show that the proposed association scheme not only significantly improves the domain coverage compared to existing approaches but also achieves better detection accuracy. Existing path-based inference algorithm is specifically designed for DNS data analysis. It is effective but computationally expensive. As a solution, we investigate the effectiveness of combining our association scheme with the generic belief propagation algorithm. Through comprehensive experiments, we show that this approach offers significant efficiency and scalability improvement with only minor negative impact of detection accuracy.

cs.CR↗

arXiv2017

Efficient 3D Placement of a UAV Using Particle Swarm Optimization

Unmanned aerial vehicles (UAVs) can be used as aerial wireless base stations when cellular networks go down. Prior studies on UAV-based wireless coverage typically consider an Air-to-Ground path loss model, which assumes that the users are outdoor and they are located on a 2D plane. In this paper, we propose using a single UAV to provide wireless coverage for indoor users inside a high-rise building under disaster situations (such as earthquakes or floods), when cellular networks are down. We assume that the locations of indoor users are uniformly distributed in each floor and we propose a particle swarm optimization algorithm to find an efficient 3D placement of a UAV that minimizes the total transmit power required to cover the indoor users.

cs.IT↗

arXiv2017

The Indoor Mobile Coverage Problem Using UAVs

Unmanned aerial vehicles (UAVs) can be used as aerial wireless base stations when cellular networks are not operational due to natural disasters. They can also be used to supplement the ground base station in order to provide better coverage and higher data rates for the users. Prior studies on UAV based wireless coverage typically consider an Air-to-Ground path loss model, which assumes that the users are outdoor and located on a 2D plane. In this paper, we propose using UAVs to provide wireless coverage for indoor users inside a high-rise building. First, we present realistic Outdoor-Indoor path loss models and describe the tradeoff introduced by these models. Then we study the problem of efficient placement of a single UAV, where the objective is to minimize the total transmit power required to cover the entire high-rise building. The formulated problem is non-convex and is generally difficult to solve. To that end, we consider three cases of practical interest and provide efficient solutions to the formulated problem under these cases. Then we study the problem of minimizing the number of UAVs required to provide wireless coverage to high rise buildings and prove that this problem is NP-complete. Due to the intractability of the problem, we use clustering to minimize the number of UAVs required to cover the indoor users. We demonstrate through simulations that the method that clusters the building into regular structures and places the UAVs in each cluster requires 80% more number of UAVs relative to our clustering algorithm.

cs.IT↗

arXiv2017

On The Continuous Coverage Problem for a Swarm of UAVs

Unmanned aerial vehicles (UAVs) can be used to provide wireless network and remote surveillance coverage for disaster-affected areas. During such a situation, the UAVs need to return periodically to a charging station for recharging, due to their limited battery capacity. We study the problem of minimizing the number of UAVs required for a continuous coverage of a given area, given the recharging requirement. We prove that this problem is NP-complete. Due to its intractability, we study partitioning the coverage graph into cycles that start at the charging station. We first characterize the minimum number of UAVs to cover such a cycle based on the charging time, the traveling time, and the number of subareas to be covered by the cycle. Based on this analysis, we then develop an efficient algorithm, the cycles with limited energy algorithm. The straightforward method to continuously cover a given area is to split it into N subareas and cover it by N cycles using N additional UAVs. Our simulation results examine the importance of critical system parameters: the energy capacity of the UAVs, the number of subareas in the covered area, and the UAV charging and traveling times.We demonstrate that the cycles with limited energy algorithm requires 69%-94% fewer additional UAVs relative to the straightforward method, as the energy capacity of the UAVs is increased, and 67%-71% fewer additional UAVs, as the number of subareas is increased.

cs.IT↗

arXiv2015

Joint Link Scheduling and Brightness Control for Greening VLC-based Indoor Access Networks

Demands for broadband wireless access services is expected to outstrip the spectrum capacity in the near-term - "spectrum crunch". Deploying additional femotocells to address this challenge is cost-inefficient, due to the backhaul challenge and the exorbitant system maintenance. According to an Alcatel-Lucent report, most of the mobile Internet access traffic happens indoor. Leveraging power line communication and the available indoor infrastructure, visible light communication (VLC) can be utilized with small one-time cost. VLC also facilitates the great advantage of being able to jointly perform illumination and communications, and little extra power beyond illumination is required to empower communications, thus rendering wireless access with small power consumption. In this study, we investigate the problem of minimizing total power consumption of a general multi-user VLC indoor network while satisfying users' traffic demands and maintaining an acceptable level of illumination. We utilize the column generation method to obtain an $ε$-bounded solution. Several practical implementation issues are integrated with the proposed algorithm, including different configurations of light source and ways of resolving the interference among VLC links. Through extensive simulations, we show that our approach reduces the power consumption of the state-of-art VLC-based scheduling algorithms by more than 60\% while maintaining the required illumination.

cs.NI↗

arXiv2014

Asymptotically-Optimal Incentive-Based En-Route Caching Scheme

Content caching at intermediate nodes is a very effective way to optimize the operations of Computer networks, so that future requests can be served without going back to the origin of the content. Several caching techniques have been proposed since the emergence of the concept, including techniques that require major changes to the Internet architecture such as Content Centric Networking. Few of these techniques consider providing caching incentives for the nodes or quality of service guarantees for content owners. In this work, we present a low complexity, distributed, and online algorithm for making caching decisions based on content popularity, while taking into account the aforementioned issues. Our algorithm performs en-route caching. Therefore, it can be integrated with the current TCP/IP model. In order to measure the performance of any online caching algorithm, we define the competitive ratio as the ratio of the performance of the online algorithm in terms of traffic savings to the performance of the optimal offline algorithm that has a complete knowledge of the future. We show that under our settings, no online algorithm can achieve a better competitive ratio than $Ω(\log n)$, where $n$ is the number of nodes in the network. Furthermore, we show that under realistic scenarios, our algorithm has an asymptotically optimal competitive ratio in terms of the number of nodes in the network. We also study an extension to the basic algorithm and show its effectiveness through extensive simulations.

cs.NI↗