In this work we address a problem of active fault detection in an IoT scenario, whereby a monitor probes a remote device in order to detect faults and acquire fresh information. However, probing can have a significant impact on the IoT network’s energy and communication resources. To address this problem we utilize Age of Information as a measure of the freshness of information at the monitor and adopt a semantics-aware communication approach between the monitor and the remote device. In semantics-aware communications, the processes of generating and transmitting information are treated jointly in order to consider the importance of information and the purpose of communication. We formulate the problem as a Partially Observable Markov Decision Process and propose a computationally efficient stochastic approximation algorithm to approximate the optimal policy. Finally, we present numerical results that exhibit the advantage of our approach compared to a conventional delay-based probing policy.

Traditional supply chain traceability involves time-consuming operations where human involvement is required. With the help of IoT, supply chain traceability processes that are complex, time-consuming and error-prone, can be automated and become more reliable. In this paper, we present the work-in-progress of an IoT-based platform for supply chain monitoring that can make feasible the collection of IoT sensory data, jointly with supply chain data formatted, according to GS1 standards.

The Internet-of-Things (IoT) presents itself as an emerging technology, which is able to interconnect a massive number of heterogeneous smart objects. These ubiquitous object-enabled networks, which may operate for several years in variable conditions, are used for supporting complex data-driven applications such as smart cities applications, home automation, health monitoring, etc. Throughout their extended lifetime, the devices forming the IoT networks need to be re-programmed, so that new features are added and software bugs or security vulnerabilities are resolved. In this work, we present work-in-progress on the design of a Compilation-and Remote-Programming-as-a-Service platform that enables cloud-based compilation and build of device firmware, as well as remote firmware updates for deployed IoT devices, in a reliable and secure way. We introduce the functional architecture of our platform and elaborate on the interaction of functional components. Our solution can easily support various embedded operating systems and heterogeneous hardware platforms, as well as different communication patterns for firmware transfer.

Modern IoT solutions require minimisation of the transmitted data during a firmware update, in order to save energy for the constrained devices. To accommodate this requirement, IoT nodes can be updated incrementally using only parts of the current firmware, which is already stored in their flash memory in order to reconstruct the new firmware locally. This the role of the so-called differencing algorithm that executes in a firmware server and aims to detect common segments between the current and the new firmware, producing an encoded small delta script, which is finally transmitted to the IoT nodes. In this work, we present a differencing algorithm that operates in byte-level and can compute optimal, in terms of size, delta scripts in O(nlogn) time and O(n) space complexity.

In the new era of IoT, hundreds or even thousands of interconnected miniature sensors have made feasible the creation of novel applications spanning in multiple areas like e-health, environmental monitoring, on-farming, etc. Despite the technological advances in this domain, the sensors are still severe constrained devices, in terms of memory and processing. These limitations cannot only compromise applications' performance but can also affect trust and security in the IoT ecosystem. Besides security and trust, energy efficiency is also of paramount importance as sensors are often battery-operated. For energy minimisation and data security purposes, several contributions have mainly focused either on data compression, or data encryption; however, considering those as two independent operations. The last few years, the Compressive Sensing theory has shown that compression and encryption can be used simultaneously, given that data are sparse in some domain. As data sparsity cannot be known in advance, here, we present a practical implementation of a compressive sensing system where the data sparsity is estimated, and the compression rate is selected accordingly. Our system consists of two entities: a server implemented in Java running on a powerful machine, and a client that runs in a miniature sensor, developed in C and executing in the Contiki operating system. The evaluation results show the superiority of the proposed scheme against a non-adaptive one.

Internet of Things (IoT) interconnects resource constrained devices for providing smart applications to citizens. These devices have to be able to ensure both a minimum Quality of Service (QoS) and a minimum level of security when gathering and transmitting data. Compressive Sensing (CS) is a relatively new theory that performs simultaneous lightweight compression and encryption and can be used to prolong the battery lifetime of devices. In this paper, we stress the fact that on the contrary with most previous approaches, the sparsity of the signals can change significantly due to their time-varying nature. We propose a rate-adaptive scheme for maintaining a maximum level of reconstruction error at the receiver, and ensure the QoS requirements. This scheme uses a change point detection method, detecting the change in the sparsity, and estimating the optimum compression rate for maintaining a minimum reconstruction error. Performance is evaluated using real experimental data.

We present intrusion detection algorithms to detect physical layer jamming attacks in wireless networks. We compare the performance of local algorithms on the basis of the signal-to-interference-plus-noise ratio SINR executing independently at several monitors, with a collaborative detection algorithm that fuses the outputs provided by these algorithms. The local algorithms fall into two categories: simple threshold that raise an alarm if the output of the SINR-based metrics we consider deviates from a predefined detection threshold and cumulative sum cusum algorithms that raise an alarm if the aggregated output exceeds the predefined threshold. For collaborative detection, we use the Dempster-Shafer theory of evidence algorithm. We collect SINR traces from a real IEEE 802.11 network, and with the use of a new evaluation method, we evaluate both the local and the Dempster-Shafer algorithms in terms of the detection probability, false alarm rate, and their robustness to different detection threshold values, under different attack intensities. The evaluation shows that the cusums achieve higher performance than the simple threshold algorithms under all attack intensities. The Dempster-Shafer algorithm when combined with the simple algorithms, it can increase their performance by more than 80%, but for the cusum algorithms it does not substantially improve their already high performance.Copyright © 2013 John Wiley & Sons, Ltd.

The MESH-WISE project is aiming to address key fundamental issues in wireless mesh networking that span from fundamental performance characterization to prototyping heterogeneous-access mesh networking solutions for emergency response scenarios. In this paper we present the infrastructures that will be leveraged in this project and the latest results that have been acquired from the ones that are deployed. We further discuss the state of the art in academic and industrial resource and network management platforms and discuss how we will use existing know-how towards a centralized solution.

The recent advances in micro-sensor hardware technologies, along with the invention of energy-efficient protocols, have enabled a world-wide spread in wireless sensor networks deployment. These networks are used for a large number of purposes, while having small maintenance and deployment costs. However, as these are usually unattended networks, several security threats have emerged. In this work, we show how an adversary can overhear the encrypted wireless transmissions, and detect the periodic components of the wireless traffic that can further reveal the application used in the sensor network. Traffic analysis is performed in a very energy-efficient way using the compressed sensing principles. Furthermore, the periodic components are detected using the Lomb-Scargle periodogram technique.

The broadcast nature of wireless networks has been widely exploited by adversaries in order to cause severe denial-of-service attacks. Several algorithms are proposed in the literature for the detection and mitigation of such attacks at the physical and medium access layers. In this work, we combine recent advances in compressed sensing theory, along with a cumulative-sum anomaly-based algorithm, for the detection of physical-layer attacks. The algorithm considers a metric based on the Signal-to-Interference-plus-Noise-Ratio (SINR). Compressed sensing makes feasible the use of far fewer SINR measurements for effective intrusion detection. The performance evaluation based on real experimental data shows that attacks are detected with high accuracy using a small number of measurements.

Wireless mesh networks (WMNs) have attracted much attention lately for providing efficiently wireless services with high quality of service (QoS). Metropolitan WMNs are a low-cost solution for providing broadband wireless internet access in large areas. One of the fundamental issues of wireless communications is interference. In WMNs interference can be caused by simultaneous transmissions at links internal to the mesh network or by external sources. In this work we perform extensive measurements in a multi-radio metropolitan WMN deployed in the city of Heraklion, Greece. The basic goal is to investigate the impact of interference on the performance of the multi-path WMN. Towards that goal, we perform measurements with FTP, video streaming and raw data traffic for two scenarios, one with an efficient channel assignment (CA) that accounts for interference and another with a random CA that results in high internal and external interference in the network. The results prove that interference creates severe performance degradation, with regards to high delay, high packet losses, low throughput and low signal-to-interference and noise ratio (SINR). As a result, the metropolitan WMN becomes unable to support multi-path flows and demanding applications with an acceptable QoS.

This paper presents the performance evaluation of a PC-based active router, in terms of packet loss within the router. The architecture of the active router consists of two parts: a software part that is implemented using two Linux hosts and a hardware part that contains a PCI-based FPGA (field programmable gate array) board. A description of the modules comprising the router is given and then the method used to gather the performance evaluation data is described. The main sources for the CPU cycles consumption in the router are highlighted and a method for minimising the packet loss is applied. An example of the active network approach is given by implementing an active secure FTP application. Its performance is compared to that of passive FTP. The performance evaluation of the active secure FTP application reveals several bottlenecks, especially in the hardware extension of the active router