Rustem Dautov

Distributed Data Fusion for the Internet of Things

The ubiquitous Internet of Things is underpinned by the recent advancements in the wireless networking technology, which enabled connecting previously scattered devices into the global network. IoT engineers, however, are required to handle current limitations and find the right balance between data transferring range, throughput, and power consumption of wireless IoT devices. As a result, existing IoT systems, based on collecting data from a distributed network of edge devices, are limited by the amount of data they are able to transfer over the network. This means that some sort of data fusion mechanism has to be introduced, which would be responsible for filtering raw data before sending them further to a next node through the network. As a potential way of implementing such a mechanism, this paper proposes utilising Complex Event Processing and introduces a hierarchical distributed architecture for enabling data fusion at various levels.

Property-Based Network Discovery of IoT Nodes Using Bloom Filters

As the number of IoT devices is exponentially growing, and IoT networks are expanding in their size and complexity, timely device discovery is becoming a pressing concern. The extreme (and constantly growing) number of network nodes, dynamically connecting to and disconnecting from a network, renders existing routing techniques, such as multicasting and broadcasting, unscalable, especially when using the IPv6 128-bit addresses. To address this limitation, this paper discusses the potential of implementing the IoT device discovery, based on device properties, such as type, functionality, location, etc., and presents an approach to enable property-based access to IoT nodes using Bloom filters. The proposed approach demonstrates space- and network-efficient characteristics, as well as provides an opportunity to perform device discovery at various granularity levels.

Metropolitan intelligent surveillance systems for urban areas by harnessing IoT and edge computing paradigms: Metropolitan Intelligent Surveillance Systems

Recent technological advances led to the rapid and uncontrolled proliferation of intelligent surveillance systems (ISSs), serving to supervise urban areas. Driven by pressing public safety and security requirements, modern cities are being transformed into tangled cyber‐physical environments, consisting of numerous heterogeneous ISSs under different administrative domains with low or no capabilities for reuse and interaction. This isolated pattern renders itself unsustainable in city‐wide scenarios that typically require to aggregate, manage, and process multiple video streams continuously generated by distributed ISS sources. A coordinated approach is therefore required to enable an interoperable ISS for metropolitan areas, facilitating technological sustainability to prevent network bandwidth saturation. To meet these requirements, this paper combines several approaches and technologies, namely the Internet of Things, cloud computing, edge computing and big data, into a common framework to enable a unified approach to implementing an ISS at an urban scale, thus paving the way for the metropolitan intelligent surveillance system (MISS). The proposed solution aims to push data management and processing tasks as close to data sources as possible, thus increasing performance and security levels that are usually critical to surveillance systems. To demonstrate the feasibility and the effectiveness of this approach, the paper presents a case study based on a distributed ISS scenario in a crowded urban area, implemented on clustered edge devices that are able to off‐load tasks in a “horizontal” manner in the context of the developed MISS framework. As demonstrated by the initial experiments, the MISS prototype is able to obtain face recognition results 8 times faster compared with the traditional off‐loading pattern, where processing tasks are pushed “vertically” to the cloud.

Three-level hierarchical data fusion through the IoT, edge, and cloud computing

The Internet of Things (IoT) has embraced a vertical off-loading model, where avalanches of raw data generated by numerous edge devices are continuously pushed through the network to a remote processing location, such as a datacenter or a cloud. In this rather unbalanced architecture, edge devices are typically not expected to perform sophisticated data processing and analytics, and data fusion takes place remotely from the original source of data. As a result, the underlying network and the remote datacenter have to handle increased amounts of unstructured raw data, which, in turn, may affect the overall performance and decrease reaction times. As a potential solution to these shortcomings, this paper introduces a distributed hierarchical data fusion architecture for the IoT networks, consisting of edge devices, network and communications units, and cloud platforms. According to the proposed approach, different data sources are combined at each level of the IoT hierarchy to produce timely and accurate results by utilising computational capabilities of intermediate nodes. This way, mission-critical decisions, as demonstrated by the presented smart healthcare scenario, are taken with minimum time delay, as soon as necessary information is generated and collected. The initial evaluation suggests that the proposed approach enables fine-grained decision taking at different data fusion levels, and, as a result, improves the overall performance and reaction time.

Semantic Web Languages for Policy Enforcement in the Internet of Things.

To enable device compatibility, interoperability and integration in the Internet of Things (IoT), several ontological frameworks have been developed, using the Semantic Web technologies – a common and widely-adopted toolkit for addressing the heterogeneity issues in complex IT systems. These ontologies aim to provide a common vocabulary of terms to be universally adopted by the IoT community. Defined using the Web Ontology Language – a language underpinned by the Description Logics – these vocabularies, however, seem to neglect the automated reasoning support, which comes along with this semantic approach to model IoT environments. To bridge this gap, this paper builds upon the existing work in the area of semantic modelling for the IoT, and proposes utilising IoT ontologies to define and enforce policies, thus benefiting from the built-in support for automated reasoning.

Crowdsourcing and Stigmergic Approaches for (Swarm) Intelligent Transportation Systems

In the last decades, the impact of Information and Communication Technologies (ICT) on transportation systems radically changed them, identifying in the Intelligent Transportation Systems (ITS) a new research area. A problem often addressed in ITS is vehicle routing, for which plenty of solutions have been already defined in literature. Vehicle routing problems are usually NP hard, therefore these are mainly heuristic solutions. A requirement for them is to be deployed and run in navigation systems, ready to react to sudden changes in a (quasi) real-time way. Hence, to reduce the latency is still an open issue, not only depending on the complexity of the solution but also on other parameters, such as the traffic update latency in traffic-aware vehicle routing. A way to solve them is by exploiting distributed, collaborative approaches, establishing a proper collaboration platform and algorithms able to use it. Mobile Crowdsensing, on the one hand, and collective and swarm intelligence approaches, on the other, can fill this gap. This paper is a first attempt in this direction, aiming at defining a new class of (swarm) the Intelligent Transportation Systems (SITS), on top of a crowdsourcing-based infrastructure.

Towards a Global Intelligent Surveillance System

Recent technological advances have led to the rapid development of Intelligent Surveillance Systems (ISSs), ubiquitously present in modern urban spaces are constantly generating streams of raw data. As most of the actual Internet traffic is nowadays constituted by visual data streams, often originated by ISSs, it is important to properly manage these avalanches of data so as to support sustainability of this technological trend, which will very likely saturate the current network bandwidth in few years. This paper aims to combine existing technologies and paradigms from the Internet of Things, Cloud, Edge Computing and Big Data into a common framework to enable a shared approach for ISSs at a wide geographical scale, thus envisioning a Global ISS. The proposed solution is based on the idea of pushing data processing tasks as close to data sources as possible, thus increasing security and performance levels, usually critical to surveillance systems. To demonstrate the feasibility and the effectiveness of the proposed approach, the paper presents a case study based on a distributed ISS scenario in a crowded area, implemented on clustered edge devices able to offload tasks in a horizontal manner.

Finding Correlations Between Driver Stress and Traffic Accidents: An Experimental Study

As the number of people getting injured or killed on the roads is constantly growing, it is crucial to identify and prevent potential factors causing traffic accidents. This paper focuses on one of such factors – namely, the drivers’ stress, which is known to be one of the main causes of traffic accidents, and timely detection of such situations becomes an important challenge. The paper aims to find a potential correlation between the driver stress when riding through a specific urban location and the recorded history of traffic accidents in that specific location. If proven, such a correlation can help to prevent traffic accidents and re-design urban spaces in a safer manner. To achieve this goal, the paper combines cross-disciplinary techniques from Computer Science and Physiology to measure drivers’ stress levels using physiological sensors during city rides, and match these experimental results against a map of previously recorded traffic accidents. As a result, the conducted study indicates that the correlation indeed exists, and measuring drivers’ stress levels using physiological sensors is a promising approach to minimise the amount of traffic accidents.

Targeted Content Delivery to IoT Devices Using Bloom Filters.

The increasing number of smart interactive devices connected to the network opens new business opportunities for digital content and advertisement providers, interested in reaching out to new customer audiences. To this end, they employ various device discovery and data collection techniques to gather user- and device-specific information in order to build a user profile and deliver targeted content accordingly. However, the extreme (and constantly growing) number of smart devices, dynamically connecting to and disconnecting from a network in the IoT scenario, renders existing routing techniques, such as multicasting and broadcasting, unscalable, especially when using the IPv6 128-bit addresses. Moreover, these existing solutions can hardly provide information about technical capabilities of end devices. To address this limitation, this paper discusses the potential of implementing the IoT device discovery for device-specific content delivery, based on device properties, such as screen size and resolution, network connectivity, presence of speakers, supported languages, etc., and presents an approach to enable property-based access to IoT nodes using Bloom filters. The proposed approach demonstrates space- and network-efficient characteristics, as well as provides an opportunity to perform device discovery at various granularity levels.

Vs-Driven Big Data Process Development

Big Data solutions aim to cope with the overwhelming amount of data generated by various domains, such as social networks and the Internet of Things, thereby enabling a new generation of data-intensive applications (DIAs) and services. At the same time, to facilitate DIA design and development processes and address (Big) data management requirements, proper techniques and tools are requested. To this purpose, this paper proposes an approach, which takes into account the established Big Data V-attributes, (i.e. Volume, Velocity, and Variety) to model and predict computational demands at design time. To do so, the approach relies on annotating Big Data process workflows (and their individual elements) with relevant V-attribute values, which are then mapped into resource requirements and used in a performance model.