Search

Mobile devices are an indispensable part of modern-day lives to support portable computations and context-aware communication. Android applications within a mobile device share data to support application operations and better user experience, which also increases security risks to device’s data integrity and confidentiality. To analyze the security provided by the Android permissions, modern security techniques, based on the programming languages, have been used to enforce best practices for developing the secure Android applications. Android security assessment, based on the language-based techniques in an informal setting without formal tool support, is tedious and error-prone. Furthermore, the lack of proof of the soundness of the language-based techniques raises questions about the validity of the analysis. To enable computer-aided formal verification in Android security domain, we have developed a mathematical model of language-based Android security using computer-based proof assistant Coq. One of the main challenges for mechanizing the language-based security in theorem prover relates to the complexity of variable bindings in language-based security techniques. As the main contributions of the paper: 1) the language-based security, including variable binding, is formalized in theorem prover Coq; 2) a formal type checker is built to type check (capture safe data flows within) Android applications using computer; and 3) the soundness of the language-based security technique (type system) is mechanically verified. The formal model of the Android type system and their proof of soundness are machine-readable, and their correctness can be checked in the computer using Coq proof and type checkers.

To promote online businesses and sales, e-commerce industry focuses to fulfill users' demands by giving them top set of recommendations which are ranked through different ranking measures.Deep learning based auto-encoder models have further improved the performance of recommender systems. A state-of-the-art collaborative denoising auto-encoder (CDAE) models user-item interactions as a corrupted version of users rating inputs. However, this architecture still lacks users' ratings-trend information which is an important parameter to recommend top-N items to users. In this paper, building upon CDAE characteristics, we propose a novel users rating-trend based collaborative denoising auto-encoder (UT-CDAE) which determines user-item correlations by evaluating rating-trend(High or Low) of a user towards a set of items. This inclusion of a user's rating-trend provides additional regularization flexibility which helps to predict improved top-N recommendations. The correctness of the suggested method is verified through different ranking evaluation metrics i.e., (mean reciprocal rank, mean average precision and normalized discounted gain), for various input corruption values, learning rates and regularization parameters.Experiments on standard ML-100K and ML-1M datasets show that suggested model has improved performance overstate-of-the-art denoising auto-encodermodels.

Underwater Wireless Sensor Networks (UWSNs) face numerous challenges due to small bandwidth, long propagation delay, limited energy resources and high deployment cost. Development of efficient routing strategies is, therefore, mandatory and has remained the focus of researchers over the past few years. To address these challenges and to further improve the performance of the existing protocols, many routing protocols have been designed. In Weighting Depth and Forwarding Area Division-Depth Based Routing (WDFAD-DBR), the forwarding decision is based on the weighting depth difference, which is not an efficient way for void hole avoidance. In this paper, we propose a depth-based routing mechanism called Energy Balanced Efficient and Reliable Routing (EBER2) protocol for UWSNs. First, energy balancing among neighbors and reliability are achieved by considering residual energy and the number of Potential Forwarding Nodes (PFNs) of the forwarder node, respectively. Secondly, energy efficiency is enhanced by dividing the transmission range into power levels, and the forwarders are allowed to adaptively adjust their transmission power according to the farthest node in their neighbor list. Thirdly, duplicate packets are reduced by comparing depths, residual energy and PFNs among the neighbors. Moreover, network latency is decreased by deploying two sinks at those areas of the network that have high traffic density. The results of our simulations show that EBER2 has higher Packet Delivery Ratio (PDR), lower energy tax, and lesser duplicate packets than the WDFAD-DBR routing protocol.

The demand for recommender systems in E-commerce industry has increased tremendously. Efficient recommender systems are being proposed by different E-business companies with the intention to give users accurate and most relevant recommendation of products from huge amount of information. To improve the performance of recommender systems, various stochastic variants of gradient descent based algorithms have been reported. The scalability requirement of recommender systems needs algorithms with fast convergence to generate recommendations of specific items. Using the concepts of fractional calculus, an efficient variant of the stochastic gradient descent (SGD) was developed for fast convergence. Such fractional SGD (F-SGD) is further accelerated by adding a momentum term, thus termed as momentum fractional stochastic gradient descent (mF-SGD). The proposed mF-SGD method is shown to offer improved estimation accuracy and convergence rate, as compared to F-SGD and standard momentum SGD for different proportions of previous gradients, fractional orders, learning rates and number of features.

The developments and applications of wireless body area networks (WBANs) for healthcare and remote monitoring have brought a revolution in the medical research field. Numerous physiological sensors are integrated in a WBAN architecture in order to monitor any significant changes in normal health conditions. This monitored data are then wirelessly transferred to a centralized personal server (PS). However, this transferred information can be captured and altered by an adversary during communication between the physiological sensors and the PS. Another scenario where changes can occur in the physiological data is an emergency situation, when there is a sudden change in the physiological values, e.g., changes occur in electrocardiogram (ECG) values just before the occurrence of a heart attack. This paper presents a centralized approach for the detection of abnormalities, as well as intrusions, such as forgery, insertions, and modifications in the ECG data. A simplified Markov model-based detection mechanism is used to detect changes in the ECG data. The features are extracted from the ECG data to form a feature set, which is then divided into sequences. The probability of each sequence is calculated, and based on this probability, the system decides whether the change has occurred or not. Our experiments and analyses show that the proposed scheme has a high detection rate for 5% as well as 10% abnormalities in the data set. The proposed scheme also has a higher true negative rate with a significantly reduced running time for both 5% and 10% abnormalities. Similarly, the receiver operating characteristic (ROC) and ROC convex hull have very promising results.

In today's Internet of Everything (IOE), security and safety hazards have been promptly increasing. The rapidly evolving and complex technological innovations continuously impede preemptive security and safety mechanisms. In such a scenario, well-functioning mechanisms and versatile systems are essential, which can contribute towards advanced safety and security approaches. To avoid the hazardous situation, these efficient systems and mechanisms work on two important principles including security policy to counter threats and the resilience to handle safety hazards. The role of trust to manage safety and security is significant. It is actually the belief that the system is competent enough to function securely, reliably, and dependably in a specific context. Therefore, the unified approach to manage trustworthiness, safety, and security on such systems can be of great importance. By developing trust among different entities involved in such systems/mechanisms, the security of the sensitive resources can be ensured more effectively. In the same way, several trust levels on the systems within a network could enhance cooperation and security by minimizing safety hazards and malicious activities. Moreover, there is a robust requirement of developing generic safety and security management mechanism, which should be reliable enough to handle multiple numbers and types of resources since systems/mechanisms integrate applications of different nature. Each of these applications/systems requires different safety rules, information assurance, and security policies that existing systems are incapable to manage. Therefore, in order to several types of resources, the co-engineering approach should be adopted during development, which addresses both safety hazards and security threats.

Traditional anti-virus technologies have failed to keep pace with proliferation of malware due to slow process of their signatures and heuristics updates. Similarly, there are limitations of time and resources in order to perform manual analysis on each malware. There is a need to learn from this vast quantity of data, containing cyber attack pattern, in an automated manner to proactively adapt to ever-evolving threats. Machine learning offers unique advantages to learn from past cyber attacks to handle future cyber threats. The purpose of this research is to propose a framework for multi-classification of malware into well-known categories by applying different machine learning models over corpus of malware analysis reports. These reports are generated through an open source malware sandbox in an automated manner. We applied extensive pre-modeling techniques for data cleaning, features exploration and features engineering to prepare training and test datasets. Best possible hyper-parameters are selected to build machine learning models. These prepared datasets are then used to train the machine learning classifiers and to compare their prediction accuracy. Finally, these results are validated through a comprehensive 10-fold cross-validation methodology. The best results are achieved through Gaussian Naive Bayes classifier with random accuracy of 96% and 10-Fold Cross Validation accuracy of 91.2%. The said framework can be deployed in an operational environment to learn from malware attacks for proactively adapting matching counter measures.

The goal of SSTM'17 was to attract young researchers, Ph.D. students, practitioners, and industry experts to bring contributions in the area of Security, Safety and Trust Management, especially in the developments of computing, management and programming models, technologies, framework and middleware

High Definition (HD) videos are the most widely used in HD television and mobile phones now a days for transmission and storage. Due to large data size, HD videos require efficient and robust video coding mechanism to enable real-time encoding. Numerous Motion Estimation (ME) algorithms are proposed to reduce the computational complexity of the coding process. In this paper, we present the performance analysis of some famous Block Matching ME Algorithms (BMAs) for HD videos. Different performance measuring parameters are used to evaluate the performance of BMAs, like Peak Signal to Noise Ratio (PSNR), ME time, Mean Square Error (MSE). The simulation results show that the Adaptive Rood Pattern Search (ARPS) ME algorithm outperforms in term of MSE, PSNR and number of search points, for HD (720p) videos, over various search parameters. ARPS, Diamond Search (DS) and Flatted Hexagon Search (FHS) ME algorithms improve the PSNR from 32dB to 48dB for some video sequences, by increasing search range, whereas the number of search points also increased with the same parameter that causes to increased ME time and computational complexity.

Threats to computer networks are numerous and potentially devastating. Intrusion detection techniques provide protection to our data and track unauthorized access. Many algorithms and techniques have been proposed to improve the accuracy and minimize the false positive rate of the intrusion detection system (IDS). Statistical techniques, evolutionary techniques, and data mining techniques have also been used for this purpose. In this paper, we use a centroid-based technique for network intrusion detection in which the centroid is constructed on the basis of diversity. Diversity of a point is the sum of the distances from a point to all other points in a cluster. The point having minimum diversity is chosen as a centroid. The performance of diversity-based centroid shows significant improvement in the classification of intrusions. Experimental results on the KDDCup99 dataset demonstrate that the proposed method shows excellent performance in terms of accuracy, detection rate, and false positive rate.

Arrhythmia is caused by improper and irregular sinus rhythm or heartbeats. In order to diagnose cardiac arrhyth- mia, electrocardiogram (ECG) beat classification and analysis is very necessary. The efficiency and accuracy of any classification model highly depends on selecting the most relevant features. The aim of this study is to classify different arrhythmic beats with a reduced set of relevant-only ECG features. To optimize the ECG feature selection process and increase the classification accuracy, a Mahalanobis-Taguchi System (MTS) based classifica- tion and analysis scheme is proposed. MTS is a multi-dimensional pattern recognition system which dynamically selects important features for further analysis. Arrhythmia can occur at any time and thus requires proper and continuous monitoring of the patient to reduce sudden heart attacks. The proposed MTS- based classification scheme is integrated with a Wireless Body Area Network (WBAN) for pervasive monitoring. The proposed scheme is analyzed and compared with a state-of-the-art scheme in terms of sensitivity, specificity, and accuracy. The results show that the proposed scheme performs significantly better than the other scheme by achieving high sensitivity, specificity, and classification accuracy for different arrhythmic heartbeats i.e., Left Bundle Branch Block (LBBB), Premature Ventricular Contraction (PVC), Right Bundle Branch Block (RBBB), and Atrial Premature Contraction (APC).

Optical waveguide is the backbone of an electrooptic modulator (EOM). In this paper an optical waveguide design is proposed for EOM. The proposed optical waveguide provide better impedance and velocity matching to the electrode as compare to the previous work done on electro-optic modulator. In addition to this, the push pull technique and polymer integration used in the design of EOM reduces required drive voltage and increase the speed of operation. Travelling wave structure is used in the proposed design, which is proved to be advantages for EOM as there is no limitation on modulator speed due to resistance and capacitance as in lumped element design. The electrode parameters are analysed with the help of an electro-optic solver tool.

Information and communication security has gained significant importance due to its wide spread use, increased sophistication and complexity in its deployment. On the other hand, more sophisticated and stealthy techniques are being practiced by the intruder's group to penetrate and exploit the technology and attack detection. One such treacherous threat to all critical assets of an organization is Advanced Persistent Threat (APT). Since APT attack vector is not previously known, consequently this can harm the organization's assets before the patch for this security flaw is released/available. This paper presents a preliminary research effort to counter the APT or zero day attacks at an early stage by detecting malwares. Open Source version of Security Information and Event Management (SIEM) is used to detect denial of service attack launched through remote desktop service. The framework presented in this paper also shows the efficiency of the technique and it can be enhanced with more sophisticated mechanisms for APT attack detection.

A system connected to a network is an open choice for network intrusions unless a powerful intrusion detection or prevention system is implemented. Network security has become a serious issue due to increased unauthorized access and manipulation of network resources. Evolutionary approaches play an important role in identifying attacks with high detection rates and low false discovery rates. In this paper, a binary version of multi-objective particle swarm optimization (PSO) approach is used to detect PROBE attacks in a network. A vector evaluated PSO approach is used in the proposed technique with two objectives i.e., intrusion detection rate and false discovery rate, to guide the process of feature selection. The experiments are performed using the well-known KDD99Cup dataset. Multi-objective PSO approach is used for feature selection from a set of 41 features and Random Forests (RF), a highly accurate and fast algorithm, is used for classification. Empirical results show that the proposed technique outperforms well-known classification and regression techniques in most of the cases.

Location information of sensor nodes deployed in the mission field plays an important role on the performance of Wireless Sensor Networks (WSNs). It is highly desirable to develop localization systems by keeping in mind WSN constraints and its location estimation capability. Optimization algorithms have proven to be good candidates for quality of position estimation. Flip ambiguity is one of the major challenges in such techniques. In this paper two types of constraints are proposed to overcome this problem. Particle Swarm Optimization (PSO) in conjunction with the proposed constraints is used iteratively in distributed manners to localize blind nodes in the WSN. Simulation results show that the proposed technique overcomes the problem of flip ambiguity and is resource efficient as well. The proposed technique mitigates 95 percent (worst-case) to 100 percent (best-case) flips and saves 80 percent (worst-case) to 87 percent (best-case) energy as compared to the previous technique available in the literature.

This paper presents an application for monitoring and detection of fire in coal mines using wireless sensor networks (WSNs). The application uses BDI (Belief, Desire and Intention) based multi-agent model and its implementation on sensor networks. The language used for implementation is interpreted by Jason; an extension of AgentSpeak which is based on the BDI Architecture. The BDI agents are reactive planning systems; systems that are not meant to compute the value of a function and terminate but rather designed to be permanently running and reacting to some form of event. The distributed model of the environment is adopted to overcome the communication overhead, power consumption, network delay and reliability on a centralized base station.

In this paper, we propose a new evolutionary optimization algorithm called Welfare State Optimization (WSO) for solving optimization problems. In this algorithm, we emulate the behavior of welfare states to improve the lives of their citizens. The work is motivated by the fact that the welfare state optimally uses its resources (optimization) and restricts a group to lead the whole nation to a specific direction (local trap). So, the behavior of a welfare state is quite suitable for optimization algorithms. The proposed WSO algorithm is validated using ten standard benchmark functions and its performance is compared with five different variants of Particle Swarm Optimization (PSO) available in the literature. The results of our experiments are very promising and confirm the validity of the proposed approach. Hence, WSO algorithm can be considered as a strong alternative to solve optimization problems.

Group Counseling Optimizer (GCO) is a new heuristic inspired by human behavior in problem solving during counseling within a group. GCO has been found to be successful in case of single-objective optimization problems, but so far it has not been extended to deal with multi-objective optimization problems. In this paper, a Pareto dominance based GCO technique is presented in order to allow this approach to handle multi-objective optimization problems. In order to compute change in decision for each individual, we also incorporate a selfbelief counseling probability operator in the original GCO algorithm that enriches the exploratory capabilities of our algorithm. The proposed Multi-objective Group Counseling Optimizer (MOGCO) is tested using several standard benchmark functions and metrics taken from the literature for multiobjective optimization. The results of our experiments indicate that the approach is highly competitive and can be considered as a viable alternative to solve multi-objective optimization problems.

From the security perspective Mobile Ad hoc Networks (MANETs) are amongst the most challenging research areas and one of the key reasons for this is the ambiguous nature of insider attacks in these networks. In recent years, many attempts have been made to study the intrinsic attributes of these insider attacks but the focus has generally been on the analysis of one or very few particular attacks, or only the survey of various attacks without any performance analysis. Therefore, a major feature that research has lately lacked is a detailed and comprehensive study of the effects of various insider attacks on the overall performance of MANETs. In this paper we investigate, in detail, some of the most severe attacks against MANETs namely the blackhole attack, sinkhole attack, selfish node behavior, RREQ flood, hello flood, and selective forwarding attack. A detailed NS-2 implementation of launching these attacks successfully using Ad hoc On-Demand Distance Vector (AODV) routing protocol has been presented and a comprehensive and comparative analysis of these attacks is performed. We use packet efficiency, routing overhead, and throughput as our performance metrics. Our simulationbased study shows that flooding attacks like RREQ flood and hello flood drastically increase the routing overhead of the protocol. Route modification attacks such as sinkhole and blackhole are deadly and severely affect the packet efficiency and bring down the throughput to unacceptable ranges.

During the past few years, huge amount of network attacks have increased the requirement of efficient network intrusion detection techniques. Different classification techniques for identifying various attacks have been proposed in the literature. In this paper we propose and implement a hybrid classifier based on binary particle swarm optimization (BPSO) and random forests (RF) algorithm for the classification of PROBE attacks in a network. PSO is an optimization method which has a strong global search capability and is used for fine-tuning of the features whereas RF, a highly accurate classifier, is used here for classification. We demonstrate the performance of our technique using KDD99Cup dataset. We also compare the performance of our proposed classifier with eight other well-known classifiers and the results show that the performance achieved by the proposed classifier is much better than the other approaches.

Domestic and international investors are interested in introducing their brands in Pakistan, the question is whether period of high inflation is suitable for introducing new brand in Pakistan. This study learned the consumer behavior during extremely high inflation period in Pakistan. The study conducted in Karachi, financial hub of the country. As population of Pakistan consisted of mainly three income classes, that are upper, middle and lower, the study selected randomly 100 consumers from each class of consumers. The descriptive statistics discovered that inflation caused a compromising change in behavior of consumers from the lower class, the study found a considerable change in terms of the buying capacity as well as selection of the brand among consumers from lower income class. The study further found little change in both these dimensions among the consumers from the middle class. However, the study did not find any large change in the behavior of the consumers from the upper class. It concludes that income level is the real determinant of consumer behavior while inflation is merely a catalyst. Based on such findings, the study contributed that period of high inflation in Pakistan is suitable for promoting new local brands providing economy in the markets where majority of the consumers come from middle and lower classes of income.

In this paper, a new transport protocol called Sensor Reliability and Congestion Control Protocol (SRCP) is proposed for Wireless Sensor Networks (WSNs) which has both the functionalities of providing end-to-end reliability as well as delaybased congestion detection and control. It uses packet sequence numbering and Selective Negative Acknowledgements (SNACKs) for providing reliability in the form of retransmissions. The sink continuously records trip times and receives the rate of incoming packets and observes the delay trend. Congestion is avoided and actively controlled by sending control packets containing packet receiving rate value to the sending sensors if there is an increase trend in the trip times. Exact rate adjustment is performed by the sender according to the receiving rate and then slow additive increase is done after small fixed intervals of time resulting in fairness among sensors. The simulation results show that SRCP saves a lot of energy by scheduling packets according to the channel capacity, resulting in small number of packet losses while maintaining the data fidelity.

It is hard for the cryptanalysts to apply traditional techniques and brute-force attacks against feistel ciphers due to their inherent structure based on high nonlinearity and low autocorrelation. In this paper, we propose a technique for the cryptanalysis of four-rounded Data Encryption Standard (DES) based on Binary Ant Colony Optimization (BACO). A known plaintext attack is used to recover the secret key of the DES cipher. The environment for the ants is a directed graph, which we call search space, is constructed for efficiently searching the secret key. We also develop a heuristic function which measures the quality of a constructed solution. Several optimum keys are computed over different runs on the basis of routes completed by the ants. These optimum keys are then used to find each individual bit of the 56 bit secret key used by DES. The results of our experiments show that ACO is an effective technique for the cryptanalysis of four-rounded DES. To the best of our knowledge, this is the first time that BACO has been used for this specific problem.

This paper presents a location-aware routing protocol for mobile ad hoc networks which uses zones for efficient routing. Unlike other protocols, on-demand routing is performed on part of the route requesting node within a zone where each zone has a leader node which is responsible for making routing decisions.