Ad-Hoc, Mobile, and Wireless Networks

Telecom operators recently started to integrate Software-Defined Networking facilities for controlling and managing their optical transport networks. Here, the management of forwarding rules into the resulting Transport Software-Defined Networking (T-SDN) architecture has to be addressed by taking into account the energy and quality of service requirements. While the most of works in the literature studied these aspects separately, the few contributions that simultaneously take care of energy and quality of service requirements present latency, scalability, or control communication issues. Starting from these considerations, this paper formulates a novel methodology for the dynamic and reactive management of forwarding rules in a (potentially large-scale) T-SDN network, based on the knowledge of network topology, the power consumption of optical switches, the expected volume of traffic, and the variability of the actual traffic load. First, the expected volume of traffic and the estimated power consumption of optical switches are exploited to select the minimum number of nodes and transport links to activate, which enable the communication among any source and destination pairs declared within a given traffic matrix. Then, the bandwidth consumption of activated transport links is periodically monitored by a centralized controller and, in case of congestion, a new set of optical switches and transport links are quickly turned on for addressing the growth of the traffic load. The effectiveness of the proposed approach has been investigated through experimental tests and compared against another reference scheme which considers the energy issue only. Obtained results demonstrate its ability to offer higher levels of quality of service to end-users, at the expense of a limited decrease of the registered energy-saving.

Thanks to recent advancements in edge computing, the traditional centralized cloud-based approach to deploy Artificial Intelligence (AI) techniques will be soon replaced or complemented by the so-called edge AI approach. By pushing AI at the network edge, close to the large amount of raw input data, the traffic traversing the core network as well as the inference latency can be reduced. Despite such neat benefits, the actual deployment of edge AI across distributed nodes raises novel challenges to be addressed, such as the need to enforce proper addressing and discovery procedures, to identify AI components, and to chain them in an interoperable manner. Named Data Networking (NDN) has been recently argued as one of the main enablers of network and computing convergence, which edge AI should build upon. However, the peculiarities of such a new paradigm entails to go a step further. In this paper we disclose the potential of NDN to support the orchestration of edge AI. Several motivations are discussed, as well as the challenges which serve as guidelines for progress beyond the state of the art in this topic.

Network Function Virtualization (NFV) can lower the CAPEX and/or OPEX for service providers and allows the deployment of services quickly. The main challenge in the use of Virtualized Network Functions (VNF) is optimally placing them in the physical network in terms of deployment cost and resource consumption. The critical problem of VNF placement is inherently NP-hard and the available optimal solutions do not scale with respect to the network size. The problem of NFV placement is even more challenging in wireless networks as we are facing the issue of scarcity of BW due to the presence of interference. Therefore, this paper aims to solve the problem of VNF placement in wireless multi-hop networks by considering BW limitations and scalability. We tackle both issues at once by limiting the search space to the shortest paths. We search for the placement solution along shortest paths to minimize the BW consumption and at the same time reduce the search space to the nodes and links along the shortest path. The results are compared to a mathematical optimization model and a comparable heuristic model. They show that our proposed heuristic greatly decreases the execution time in comparison to the mathematical model and the alternative heuristic while keeping the acceptance ratio close to the optimal solution.

Public transport networks (PTN) can serve as a basis to establish low-cost communication solutions by using Delay- and Disruption-tolerant Networking (DTN) technologies. Since vehicles move according to a schedule, DTN protocols able to leverage topological information are expected to perform well in such a setup. Anyway, it has not been evaluated, if deterministic protocols perform best or if opportunistic protocols like PRoPHET or Spray and Wait can outperform them in some scenarios if appropriate parameter tuning is performed. In this paper, the performance of state-of-the-art DTN routing protocols, namely Epidemic Routing, Spray and Wait, PRoPHET, MaxProp, and CGR is compared with respect to their use in PTN. The performance comparison takes delivery probability, average latency, buffer utilization, and network overhead into account. The ONE was extended to simulate vehicle movement according to tracks and the schedule of the PTNs of Helsinki, Freiburg, and Prague. Our evaluation results demonstrate that protocol parameters should be selected carefully to achieve the best performance in different scenarios. The most efficient parameterization of protocols is described, and their influence on different performance metrics is discussed.

This paper extends an existing decisional framework for the navigation of Micro Aerial Vehicle (MAV) swarms. The work finds inspiration in the geocaching outdoor game. It leverages place recognition methods, information sharing and collaborative work between MAVs. It is unique in that a priori none of the MAVs knows the trajectory, waypoints and destination. The MAVs collectively solve a series of problems that involve the recognition of physical places and determination of their GPS coordinates. Our algorithm builds upon various methods that had been created for place recognition. The need for a decisional framework comes from the fact that all methods are fallible and make place recognition errors. In this paper, we augment the navigation algorithm with a decisional framework resolving conflicts resulting from errors made by place recognition methods. The errors divide the members of a swarm with respect to the location of waypoints (i.e., some members continue the trip following the proper itinary; others follow a wrong one). We propose four decisional algorithms to resolve conflicts among members of a swarm due to place recognition errors. The performance of the decisional algorithms is modeled and analyzed.

The Internet of Drones (IoD) is an emerging network architecture that leverages bidirectional communications between Unmanned Aerial Vehicles (UAVs) and a ground infrastructures. Given its versatility, the IoD is a key enabler for the evolution of sophisticated verticals, thus including Smart Cities, Industry 4.0 and Smart Transportation, to name a few. As a matter of fact, a swarm of drones can carry out real-time data gathering, industrial plants monitoring and wide area surveying. Since drones are battery-supplied systems, energy efficiency, communication reliability, and flight formation management routines become critical aspects of the IoD design. This work discusses the communication facets in a swarm of drones, leveraging the joint adoption of Bluetooth and Millimeter-Wave technologies. The proposal enables the exchange of heterogeneous data with a ground infrastructure while facilitating coordinated flight of the swarm. The resulting dual mode system optimizes energy resources, while maximizing reliability and effectiveness of drones’ communications.

Multiple Unmanned Aerial Vehicles (multi-UAVs) applications are recently growing in several fields, ranging from military and rescue missions, remote sensing, and environmental surveillance, to meteorology, logistics, and farming. Overcoming the limitations on battery lifespan and on-board processor capabilities, the coordinated use of multi-UAVs is indeed more suitable than employing a single UAV in certain tasks. Hence, the research on swarm of UAVs is receiving increasing attention, including multidisciplinary aspects, such as coordination, aggregation, network communication, path planning, information sensing, and data fusion. The focus of this paper is on defining novel control strategies for the deployment of multi-UAV systems in a distributed time-varying set-up, where UAVs rely on local communication and computation. In particular, modeling the dynamics of each UAV by a discrete-time integrator, we analyze the main swarm intelligence strategies, namely flight formation, swarm tracking, and social foraging. First, we define a distributed control strategy for steering the agents of the swarm towards a collection point. Then, we cope with the formation control, defining a procedure to arrange agents in a family of geometric formations, where the distance between each pair of UAVs is predefined. Subsequently, we focus on swarm tracking, defining a distributed mechanism based on the so-called leader-following consensus to move the entire swarm in accordance with a predefined trajectory. Moreover, we define a social foraging strategy that allows agents to avoid obstacles, by imposing on-line a time-varying formation pattern. Finally, through numerical simulations we show the effectiveness of the proposed algorithms.

An effort is being made by authorities worldwide to improve the safety of the transportation of goods while preserving efficiency. Vehicle inspections are important for safety but not very frequent. When they do happen, vehicles are selected on the roadside and authorities spend a long time retrieving the relevant information while the vehicle is stopped. In this paper, we present and evaluate STOP, a road transportation vehicle inspection support system with tamper-proof records to prevent location spoofing attacks. To the best of our knowledge, it is the first such system described in literature. The STOP system uses mobile devices and a central server to allow authorities to select and notify vehicles for inspection while retrieving the needed information to prepare the procedure beforehand. The location chain for each vehicle can be verified and signed by the inspectors. We implemented a prototype in the Android platform and tested it with real users. We evaluated the system’s location retrieval accuracy, response times, and Bluetooth communication during inspection.

This work presents the ongoing development of a simulation/emulation framework for real-time multimedia transmissions in multi-channel scenarios. The proposed software engine can be used to simulate the aggregation of multiple network links delivering a video flow from a video source, such as a camera mounted on-board an Unmanned Aerial Vehicle (UAV), to the indented receiver, such as the pilot on the ground. The software engine can also be used in real testbeds exploiting multiple physical links, such as 4G/5G cellular connections. The main aim of this novel software framework is to support the design, development, and test of different scheduling strategies to achieve real-time, good quality and fluidity, energy-efficient multimedia transmissions, by selecting the optimal subset of network channels able to meet the target Quality of Experience (QoE) at the receiver.

In recent years, the scientific community has found increasing interest in technological systems for the evaluation of the mobility performance of the elderly population. The reduced quantity of dataset for gait and balance analysis of elderly people is a serious issue in studying the link between cognitive impairment and motor dysfunction, particularly in people suffering from neurodegenerative diseases. In this context, this work aims to provide a dataset with skeletal information of people aged 60 years and older, while they perform well-established tests for stability assessment. 27 healthy people and 20 patients affected by neurodegenerative diseases, housed at two different nursing institutes, have been selected for the stability analysis. Subjects have been observed and evaluated by clinical therapists while executing three motion tests, namely balance, sit-to-stand and walking. The stability postural and gait control of each subject has been analyzed using a video-based system, made of three low-cost cameras, without the need for wearable and invasive sensors. The dataset provided in this work contains the skeletal information and highly-discriminant features of the balance, sit-to-stand and walking tests performed by each subject. To evaluate the efficiency of the balance dataset, the estimated risk of fall of the subjects has been processed considering the extracted features, and compared with the expected one. Final results have proven a good estimation of the risk of fall of the people under analysis, underlining the effectiveness of the dataset.

Conserving energy is probably the most important requirement in wireless sensor networks. In TSCH, this goal is obtained by subdividing time into slots, and by switching the communication interface of Internet of Things devices (frequently referred to as motes) off when, at any given time, neither transmissions nor receptions are scheduled for them. Nevertheless, in this kind of networks a considerable amount of energy may still be wasted due to idle listening. This occurs every time a cell is scheduled for frame reception but no transmissions are performed in the related slot and channel.

In this paper, Proactive Reduction of Idle Listening (PRIL) techniques are introduced, which aim at lowering the energy wasted because of the above phenomenon. In particular, here we focus on a simplified mechanism that only considers the first hop (PRIL-F). A relevant feature of this kind of techniques is that they cannot worsen performance in any way. On the contrary, in those cases where they can be applied, they may only bring benefits. Results obtained through a simulation campaign show a tangible reduction in energy consumption, especially for periodic traffic generation, in application contexts based on either a star topology (wireless sensor and actuator networks) or a two-level topology (wireless sensor networks).

During the last decade, the research on nanotechnology and wireless communications in the terahertz band supported the design of pioneering biomedical applications. To counteract the very scarce amount of energy available for nano-devices, a current challenge is to develop energy-aware and energy harvesting mechanisms enabling long-lasting communications at the nanoscale. Many contributions in this direction envisage exploiting piezoelectric nanogenerators to retrieve energy from external vibrations (i.e., the human heartbeat) and use it for transmission purposes. Indeed, in line with the recent scientific achievements in this context, this paper investigates a power control mechanism based on the feedback control theory. The control law is conceived for managing the communication in human tissues, where nano-devices are equipped with a piezoelectric nanogenerator and transmit information messages through electromagnetic waves in the terahertz band. The amount of energy spent to transmit an information message is dynamically tuned by a proportional controller in a closed-loop control scheme which simultaneously considers harvesting and discharging processes. The whole system is analytically described with a nonlinear state equation. As well, it is presented the acceptable range of values of the proportional gain guaranteeing technological constraints and its asymptotic stability. Finally, a numerical evaluation shows the behavior of the proposed approach in a conceivable biomedical scenario.

In 2015, the IEEE 802.15.4 standard was expanded by the Deterministic and Synchronous Multi-Channel Extension (DSME) to increase reliability, scalability and energy-efficiency in industrial applications. The extension offers a TDMA/FDMA-based channel access, where time is divided into two alternating phases, a contention access period (CAP) and a contention free period (CFP). During the CAP, transmission slots can be allocated offering an exclusive access to the shared medium during the CFP. The fraction \(\tau \) of CFP’s time slots in a dataframe is a critical value, because it directly influences agility and throughput. A high throughput demands that the CFP is much longer than the CAP, i.e., a high value of \(\tau \), because application data is only sent during the CFP. High agility is given if the expected waiting time to send a CAP message is short and that the length of the CAPs are long enough to accommodate necessary GTS negotiations, i.e., a low value of \(\tau \). Once DSME is configured according to the needs of an application, \(\tau \) can only assume one of two values and cannot be changed at run-time. In this paper, we propose two extensions of DSME that allow to adopt \(\tau \) to the current traffic pattern. We show theoretically and through simulations that the proposed extensions provide a high degree of responsiveness to traffic fluctuations while keeping the throughput high.

The continuous spreading of innovative applications and services, based on the emerging Internet of Things (IoT) paradigm, leads to the need of even more efficient network architectures/infrastructures, in order to support the huge amount of information to be transmitted in real-time. Hence, new protocols and mechanisms must be conceived to allow the IoT network to be more reactive towards environmental changes and in promptly satisfying the IoT users’ requests. Aiming at dealing with the emerged issues, the paper presents an efficient IoT platform, which, thanks to fog computing principles, acts as a middleware layer between data producers and consumers; it adopts a security-aware publish&subscribe protocol, based on MQTT, coupled with a network of brokers, for efficiently sharing the processed information with end-users. Transmitted data are kept secure under an enforcement framework based on sticky policies. A test campaign is conducted on a prototypical implementation of the just mentioned platform, for preliminary evaluating its efficiency, in terms of computing effort and latency.

With the popularity of web applications, cyber security is becoming more and more important. The most common web attack is cross-site scripting (XSS), which can be easily constructed in malicious URLs. However, the existing methods of detecting XSS attacks are suffering from the lack of labeled data, and some semi-supervised methods still have the problem of mislabeling. In this paper, we propose a novel XSS attack detection model based on semi-supervised learning algorithm with weighted neighbor purity. Semi-supervised learning can make best use of little labeled data, and a simple mechanism of neighbor purity using weighted-kNN is applied to rectify mislabeled samples, improving classification accuracy. To verify the feasibility of our solution in real-world scenario, we collected real HTTP requests in the China Education and Research Network (CERNET) as training data. The comparison experiment shows that proposed method performs better than a well-known semi-supervised algorithm and a recently published ensemble learning method in different initially labeled rates.

The Internet of Things (IoT) is a heterogeneous network of constrained devices connected both to each other and to the Internet. Since the significance of IoT has risen remarkably in recent years, a considerable amount of research has been conducted in this area, and especially on, new mechanisms and protocols suited to such complex systems. Routing Procotol for Lower-Power and Lossy Networks (RPL) is one of the well-accepted routing protocols for IoT. Even though RPL has defined some specifications for its security, it is still vulnerable to insider attacks. Moreover, lossy communication links and resource-constraints of devices introduce a challenge for developing suitable security solutions for such networks. Therefore, in this study, a new intrusion detection system based on neural networks is proposed for detecting specific attacks against RPL. Besides features collected from the routing layer, the effects of link layer-based features are investigated on intrusion detection. To the best of our knowledge, this study presents the first cross-layer intrusion detection system in the literature.

LoRa (Long Range) is a low-power wide-area network technology well-suited for Internet of Things (IoT) applications. In this paper, LoRa is used in a cattle monitoring application where an ad-hoc mesh network is configured to collect GPS and accelerometer data from cattle-worn sensors and relay the collected data to a base station. Free-range cattle monitoring is a challenging application since the battery-powered sensors must be small and energy efficient, and enable data communications over long distances from unpredictable locations. We propose novel changes to the existing LoRa mesh network protocols that minimize energy consumption by using global time synchronization enabled by GPS sensors and a concurrent transmission property unique to LoRa. The mesh routing phase efficiently occurs during every data collection period, making this approach ideal for networking highly mobile sensors. We integrate efficient authentication and encryption techniques in the data exchange operations to prevent spoofing and to provide confidentiality in the message exchanges between the sensors and the base station. The performance of the proposed secure implementation is compared to an equivalent insecure implementation. Multiple cattle distribution scenarios are constructed and compared to evaluate the energy consumption of the proposed scheme.

The ubiquitousness of smartphones, wearables and other mobile devices, coupled with the increasing number of communications infrastructure present in smart cities, has led to the rise of location-based services. Many of these services do not verify the location information they consume and are vulnerable to spoofing attacks. Location proof systems aim to solve this by allowing devices to interact with location specific resources and later prove that they were at the location.

In this paper we describe and evaluate CROSS, a system that performs location verification using techniques compatible with off-the-shelf Android smartphones. We present three strategies to produce location proofs with increasing tamper-resistance. We designed our system with user privacy and security in mind, minimizing the number of connections between devices. We implemented a prototype application to assess the feasibility and reliability of the proof strategies. The application allows rewarding users who complete a touristic route with proofs of visit collected along the way. Our evaluation, which included experiments with 30 users, showed that we can use the system in real-world scenarios, providing adequate security guarantees for the use case.

We consider a relay based full-duplex wireless powered cooperative communication network which consists of a hybrid access point (HAP), N users and K decode-and-forward relays with energy harvesting capability. We propose an optimization framework for relay selection with the objective of minimizing the total transmission time subject to energy causality and user traffic demand constraints. The formulated optimization problem is a mixed integer non-linear programming problem, which is difficult to solve for the global optimal solution in polynomial-time. As a solution strategy, we decompose the proposed optimization problem into two sub-problems: time allocation problem and relay selection problem. We derive the optimal solution of the time allocation problem by using convex optimization techniques. For the relay selection problem, based on the optimality analysis, we propose a polynomial-time heuristic algorithm, which minimizes the total transmission time by allocating the best relay to each user. Through simulations, we illustrate that the proposed algorithm outperforms the conventional predetermined relay allocation scheme and performs very close to the optimal solution for different network densities, HAP power values, and initial battery levels.

This paper presents a novel data dissemination strategy called Geographic Energy-aware Epidemic Routing (GEER) for Mobile Opportunistic Networks. This routing scheme considers the residual node energy and the node degree to dynamically decide if forwarding or not data to encountered mobile nodes. Moreover, a buffer management policy is applied to preserve buffer space reducing the Time To Live (TTL) of data sent on nodes with higher degree centrality. A node density estimation differentiated for geo-graphic area is proposed to improve the data forwarding and a buffer data discarding policy has been applied to manage packets with different sizes. GEER has been compared with others schemes such as Energy Aware Epidemic Routing (EAER) and EpSoc routing scheme in terms of Data Packet delivery ratio, overhead and energy consumption.

In this paper we use a “frugal innovation approach” to propose an efficient and generic solution to provide support to the deployment of IoT system in rural areas. Our proposal includes an MQTT (Message Queuing Telemetry Transport) proxy to integrate generic low-cost and low-power sensor devices in a messaging system based on LoRa (Long Range) technology. MQTT allows these data to be provided to external “data lakes” so that they can be used for tasks such as reporting, visualization, advanced analytic, and machine learning. LoRa technology provides long wireless links that can be used to connect villages and towns.

Through a REST-based interface and using JSON as a lightweight data-interchange format, we show how our platform can be used to distribute generic sensor information from rural communities. Finally, we demonstrate through experimental evaluation that this solution provides stable data transfers over links of various kilometers with a minimal utilization of resources.

One of the most widely used communication protocol in the Internet of Things, for collecting information in cities, are LoRa and the LoRaWAN protocol. Thanks to star topologies, operators can, therefore, collect electricity, water or gas consumption remotely and automatically. Unfortunately, some equipment is placed in difficult environments such as tunnels, cellars or wastewater drains. In this paper, we present the 1st fault-tolerant LoRa LoRaWAN relay solution working in LoRaWAN class A and our experimental results that we have been running for almost 1 year on SPOT (https://​spot.​objenious.​com/​) and The Things Network platforms. This solution solves the problem of sensor access to a gateway with our relay while respecting the LoRaWAN standard. The access to the medium is managed by our solution in order to avoid collisions.

Due to uncontrolled interferences, including the self-induced multi-path fading, deterministic networking is difficult to achieve on wireless links. The radio conditions may change much faster than a centralized routing paradigm can adapt and reprogram, in particular when the controller is distant and connectivity is slow and limited. Reliable and Available Wireless (RAW) separates the routing time scale at which a complex path is recomputed from the forwarding time scale at which the forwarding decision is taken for an individual packet. RAW operates at the forwarded time scale. The RAW problem is to decide, within the redundant solutions that are proposed by the routing plane, which will be used for each individual packet to provide a Deterministic Networking (DetNet) service while minimizing the waste of resources. A solution would consist of a set of protocols that evaluate the media in real time and another that controls the use of redundancy and diversity attributes that are available along the path. In this paper, we first introduce the motivation behind this approach along with the industrial use cases that requires RAW characteristics. We then give an overview of the ongoing related works at the Internet Engineering Task Force (IETF). Finally, we present the RAW problem statement.