Anomaly detection in communication networks is the first step challenging task of securing a network, as anomalies may indicate suspicious behaviors, attacks, network malfunctions, or failures. In this paper, we address problem not only detecting anomalous events but also attributing anomaly to flows causing it. To end, develop new statistical decision theoretic framework for temporally correlated traffic via Markov chain modeling. We formulate optimal generalized likelihood ratio test...

We develop new and optimal algorithms for distributed detection in sensor networks over fading channels with multiple receive antennas at the Fusion Centre (FC). Sensors observe a hidden physical phenomenon transmit their observations using amplify-and-forward scheme to FC which is equipped antennas. derive decision rules associated probabilities of false alarm three scenarios Channel State Information (CSI) availability. For most difficult case unknown CSI, we two rule. The first based on...

The rise of geo-spatial location-based applications for the Internet Things introduces new location spoofing security risks. To overcome threat malicious attacks, we develop Secure Location (SLOT) framework which extends current state-of-the-art methods and is able to cope with such threats. SLOT incorporates another piece information has not been utilized so far, audibility information. This indicates whether a node able/unable communicate target. By leveraging on this available...

Climate change could have both positive and negative effects on the energy consumption of buildings. Today, it is not clear what extent these be at multiple spatial scales including building sectors, cities, climate zones. More importantly, uncertainty mathematical models used to estimate well understood. This knowledge gap makes difficult evaluate decisions about buildings, even technologies should prioritized in race mitigate change. Moreover, this lack for researchers build limitations...

We develop new and novel algorithms for spatial field reconstruction exceedance level estimation. consider physical phenomena which are partially observed by a wireless sensor network. focus on the practical scenario in which, due to bandwidth power constraints, observations at sensors quantized transmitted over imperfect channels. first Spatial-Best Linear Unbiased Estimator (S-BLUE) algorithm then an that is based multivariate series expansion approach resulting Saddle-point type...

We develop new algorithms for geo-spatial location estimation Internet of Things (IoT) networks by utilizing a one way time arrival (OW-TOA) approach. first demonstrate the limitations current OW-TOA IoT when some anchor nodes (ANs) are outside communication range with source, which results in ambiguous level-sets likelihood function, causing inaccurate source localization. then show that networks, addition to using conventional noisy ranging measurements from range, we can make use...

IoT market is rapidly changing the cyber threat landscape. The challenges to security and privacy arise not only because devices are large in number, but also heterogeneous type functionality. Machine learning algorithms attractive methods solve various problems such as device identification, anomaly detection, attack detection. Often, all available features extracted from network traffic fed input train models, which practice regarded best approach. Associated with different kinds of cost,...

We develop a new Location Verification System (LVS) focussed on network-based Intelligent Transport Systems and vehicular ad hoc networks. The algorithm we is based an information-theoretic framework which uses the received signal strength (RSS) from network of base-stations claimed position. Based this information derive optimal decision regarding verification user's location. Our in sense maximizing mutual between its input output data. approach practical scenario non-colluding malicious...

We develop a new location spoofing detection algorithm for geo-spatial tagging and location-based services in the Internet of Things (IoT), called enhanced using audibility (ELSA), which can be implemented at backend server without modifying existing legacy IoT systems. ELSA is based on statistical decision theory framework uses two-way time-of-arrival (TW-TOA) information between user's device anchors. In addition to TW-TOA information, exploits implicit (or outage information) improve...

Building energy demand impacts a myriad of interconnected economic, societal, and environmental aspects. As result, Buildings Energy Models (BEM) play an important role in the process urban design planning. While previous studies have investigated effects building interventions on efficiency, their applicability may be limited due to BEM’s high computational complexity. This limits ability systematically study aspects large scale. The development Machine Learning (MLM) allows required...

We develop new algorithms for spatial field reconstruction, exceedance level estimation and classification in heterogeneous (mixed analog & digital sensors) Wireless Sensor Networks (WSNs). consider physical phenomena which are observed by a WSN, meaning that it consists partially of sparsely deployed high-quality sensors low-quality sensors. The transmit their (continuous) noisy observations to the Fusion Centre (FC), while first perform simple thresholding operation then binary values over...

The dynamic and unpredictable nature of energy harvesting sources available for wireless sensor networks, the time variation in network statistics like packet transmission rates link qualities, necessitate use adaptive duty cycling techniques. Such control allows nodes to achieve long-run neutrality, where supply demand are balanced a environment such that function continuously. In this paper, we develop new framework enabling an scheme networks takes into account node battery level, ambient...

In this paper, we address the challenging problem of multiple source localization in wireless sensor networks (WSN). We develop an efficient statistical algorithm, based on novel application sequential Monte Carlo (SMC) sampler methodology, that is able to deal with unknown number sources given quantized data obtained at fusion center from different sensors imperfect channels. also derive posterior Cramér-Rao bound (PCRB) location estimate. The PCRB used analyze accuracy proposed SMC...

Any machine exposed to the Internet today is at risk of being attacked and compromised. Detecting attack attempts, be they successful or not, important for securing networks (servers, end-hosts other assets) as well forensic analysis. In this context, we focus on problem evidence gathering by detecting fundamental patterns in network traffic related suspicious activities. anomalous necessary a solution able detect many types attacks malicious activities possible. Our framework correlates...

Privacy-preserving routing protocols in wireless networks frequently utilize additional artificial traffic to hide the identities of communicating source-destination pairs. Usually, addition is done heuristically with no guarantees that transmission cost, latency, and so on, are optimized every network topology. In this paper, we explicitly examine privacy-utility tradeoff problem for develop a novel privacy-preserving algorithm called optimal privacy enhancing (OPERA). OPERA uses...

We develop a new class of event detection algorithms in Wireless Sensor Networks where the sensors are randomly deployed spatially. formulate problem as binary hypothesis testing and design optimal decision rules for two scenarios, namely Poisson Point Process Binomial random deployments. To calculate intractable marginal likelihood density, we three types series expansion methods which based on an Askey-orthogonal polynomials. In addition, novel framework to provide guidance is most...

We address the two fundamental problems of spatial field reconstruction and sensor selection in heterogeneous networks. consider case where types sensors are deployed: first consists expensive, high quality sensors; second, cheap low sensors, which activated only if intensity exceeds a pre-defined activation threshold (e.g., wind sensors). In addition, these powered by means energy harvesting their time varying status impacts on accuracy measurement that may be obtained. then following...

The verification of the location information utilized in wireless communication networks is a subject growing importance. In this work we formally analyze, for first time, performance Location Verification System (LVS) under realistic setting spatially correlated shadowing. Our analysis illustrates that anticipated levels shadowing can lead to dramatic improvement Received Signal Strength (RSS)-based LVS. We also analyze an LVS utilizes Differential (DRSS), proving rather counter-intuitive...

We develop a new algorithm for geospatial tagging Internet-of-Things (IoT) type applications, which we denote as location-of-things (LoT). The underlying idea of LoT applications is to use low-cost off-the-shelf two-way time-of-arrival (TW-ToA) ranging devices perform localization tags. first demonstrate how conventional TW-ToA algorithms may experience performance degradation in cases where some the access points (APs) are outside communication range then show that can make audibility...

We present a new paradigm to perform spatial spectrum sensing in cooperative cognitive radio networks. In these networks, nodes cooperate infer information on spectral and occupancy. Following the recently proposed interference temperature metric, we formulate problem via semi-parametric Gaussian Process modelling. This allows for development of flexible probabilistic framework inference function space. Our methodology enables reusable zones, by means collaborative scheme estimate...

In this paper, we propose and examine location verification systems (LVSs) for vehicular ad hoc networks (VANETs) in the realistic setting of Rician fading channels. our LVSs, a single authorized base station (BS) equipped with multiple antennas aims to detect malicious vehicle that is spoofing its claimed location. We first determine optimal attack strategy vehicle, which, turn, allows us analyze LVS performance as function K- factor channel between BS legitimate vehicle. Our analysis also...

We present two novel distributed event detection algorithms based on a statistical approach that tolerate Byzantine attacks where malicious (compromised) sensors send false sensing data to the gateway leading increased alarm rate. study problem of attack function optimization and decision threshold consider practical cases in our algorithms. In first case, Channel State Information (CSI) between generating source is unknown while CSI known. second as well are unknown. develop an optimal rule...

Abstract Interference is an important limitation in many communication systems. It has been shown situations that the popular Gaussian approximation not adequate and interference exhibits impulsive behavior. This paper surveys different statistical models proposed for such interference, can generally be unified using class of sub-exponential family distributions, its impact on receiver design. Visualizing optimal decision boundaries allows one to show non linear effect induced by noise...