Ad Hoc Networks

Diabetes is among the major chronic disease around the world since the Glucose level could change drastically and lead to critical conditions reaching to death sometimes. To avoid this, diabetes patient are always advised to track their glucose level at least three times a day. Fingertip pricking - as the traditional method for glucose level tracking - leads patients to be distress and it might infect the skin. In some cases, tracking the glucose level might be a hard job especially if the patient is a child. In this manuscript, we present an optimum solution to this drawback by adopting the Wireless Sensor Network (WSN)-based non-invasive strategies. Near-Infrared (NIR) -as an optical method of the non-invasive technique - has been adopted to help diabetic patients in continuously monitoring their blood without pain. The proposed solution will alert the patients’ parents or guardians of their situation when they about to reach critical conditions specially at night by sending alarms and notifications by Short Messages (SMS) along with the patients current location to up to three people.

In this paper, we compare and contrast model-based and machine learning approaches for designing caching and routing strategies to improve cache network performance (e.g., delay, hit rate). We first outline the key principles used in the design of model-based strategies and discuss the analytical results and bounds obtained for these approaches. By conducting experiments on real-world traces and networks, we identify the interplay between content popularity skewness and request stream correlation as an important factor affecting cache performance. With respect to routing, we show that the main factors impacting performance are alternate path routing and content search. We then discuss the applicability of multiple machine learning models, specifically reinforcement learning, deep learning, transfer learning and probabilistic graphical models for the caching and routing problem.

In the 90s, Q-routing assisted by reinforcement learning was introduced by Boyan and Littman with interesting results in terms of quality of service. Some recent works continue to promote the idea through improvement of the algorithm or specialized extensions. In this paper, we propose a simple modification to workaround the greedy behaviour of Q-routing by considering epoch notion. In comparison with the original Q-routing and the standard OLSRv2 under Qualnet simulator, we show that our extension provides an interesting improvement in terms of packet delivery ratio on the original irregular grid of Boyan and Littman with wireless links.

The Internet Protocol (IP) architecture could not fully satisfy Vehicular Ad-hoc Networks (VANETs) needed efficiency due to their dynamic topology and high mobility. This paper presents a technique to update the software of Electronic Control Units (ECUs) in vehicles using Information Centric Network (ICN) architecture. The proposed technique replaces Flashing Over The Air (FOTA) using IP with FOTA using ICN. The importance of FOTA is illustrated as well as the impact of applying the ICN architecture on VANETs. Through our experiments, we compare between the known FOTA over IP and the newly introduced FOTA technique over ICN.

Unmanned Aerial Vehicles (UAVs) have been extensively utilized to provide wireless connectivity in rural and under-developed areas, enhance network capacity and provide support for peaks or unexpected surges in user demand, mainly due to their fast deployment, cost-efficiency and superior communication performance resulting from Line of Sight (LoS)-dominated wireless channels. In order to exploit the benefits of UAVs as base stations or relays in a mobile network, a major challenge is to determine the optimal UAV placement and relocation strategy with respect to the mobility and traffic patterns of the ground network nodes. Moreover, considering that the UAVs form a multi-hop aerial network, capacity and connectivity constraints have significant impacts on the end-to-end network performance. To this end, we formulate the joint UAV placement and routing problem as a Mixed Integer Linear Program (MILP) and propose an approximation that leads to a LP rounding algorithm and achieves a balance between time-complexity and optimality.

This paper examines the performance of a strategy for mapping the topology of a mobile ad hoc network (MANET), providing insight for network defenders to understand how much information an adversary could discern about a target network. Using this topology inference strategy, a network eavesdropper collects frame emission start- and end-times and uses these to detect the presence of link layer acknowledgements between devices and ultimately constructs a network topology. We show how the performance of this simple strategy varies as a function of the amount of data collected by the eavesdropper over time, the size of the target network, the speed of the nodes, and the nodes’ data generation rate. We derive analytical results that allow for the rapid computation of expected true positive rate and false positive rate for topology inference in a MANET; these are compared against simulation results. The analytical results are used to derive a sensible window of observation over which to perform inference, with guidance on when to discard stale data. The results are also used to recommend strategies for network defenders to frustrate the performance of an adversary’s network inference.

Vehicular Ad Hoc Networks (VANETs) represent a significant leap forward in the deployment of intelligent transport systems. These networks enable vehicles to instantly exchange traffic information with the aim of smoothing traffic flows and intensifying drivers comfort. In this context, this study addresses the issue of traffic congestion description and detection in multi-lane highways. By making use of collected information, a Markov chain based mobility model is proposed to predict the future road traffic states. Based on the obtained stationary distribution probabilities, performance criteria in steady-state are inferred and computed for different road configurations. The numerical results validate the model demonstrated in the paper.

Wireless Rechargeable Sensor Networks (WRSNs) have become more and more popular thanks to the advances in wireless power transfer and battery material. The strategy followed by the charger to decide which sensor to be recharged next, is considered effective if only few sensing holes exist at any time, and their duration is short-lived. Ideally, the strategy will allow the system to be immortal; that is, all sensors are operational at all times. A recharging strategy is said to be flexible if it is effective for a wide range of parameters (i.e., for different applications).

In this paper, we analyze a simple decentralized recharging strategy which is based on local learning, operates without any a-priori knowledge of the network, has small memory requirements, and uses only local communication. We study the effectiveness and the flexibility of such a technique under a variety of ranges of the network parameters, showing its applicability to various contexts. We focus on three classes of applications that differ in network size (number of sensors), level of sensitivity of collected data, transmission rate, battery capacity, and type of mobile charger used to replenish energy. Our experiments show that in all these different settings, this simple local learning strategy is highly effective, achieving total immortality or near immortality in all cases.

A novel, holistic approach to cooperation incentivization in multihop wireless ad hoc networks is introduced. The concept utilizes a reputation metric to tailor a response to rational nodes’ behavior and thus promote cooperation while balancing network traffic among network nodes and protecting well-behaving nodes from excessive network load. The solution aims at deterring selfish optimization techniques which may result in network-wide Path Delivery Ratio (PDR) degradation.

In an SDN architecture, solving the controller placement problem (CPP) in a multi-controller environment plays an important role on network performance in terms of delay, reliability, control overhead, etc. In this architecture, control overhead, referred to as the network cost in this paper, consists of controller-device communications to discover the network topology, and exchange configurations and set up flow tables as well as inter-controller communications, if needed, to synchronize different network views and achieve the global view of the network. In software defined multihop wireless networking (SDMWN), because of the capacity limitation and the effect of interference on wireless links, and an in-band architecture in some types of networks to exchange both data and control traffic, it is important to solve the CPP while minimizing control overhead to reduce energy consumption, have lower packet losses and improve reliability. In this paper, the objective is to solve the CPP in SDMWN while minimizing the number of required control packets to be exchanged in the control plane. The novelty of our work is that we consider the characteristics of SDMWN and the capacity of wireless links to solve the CPP and select routes among network devices and controllers in the network. Our results demonstrate the impact of different factors such as the number of controllers, the capacity of wireless links and the arrival rate of new flows in devices on control overhead in SDMWN.

Hierarchical network architectures are widely deployed to reduce routing overheads and increase scalability. In our work, we are interested in large-scale Mobile Ad-Hoc Networks (MANETs) which are formed by interconnecting smaller clusters through a backbone. To support end-to-end routing in such networks, we employ a hierarchical approach as follows. The clusters are MANETs, running OLSR locally. Each cluster has a gateway, and the gateways are interconnected through a backbone. In this paper, we study four different solutions to provide end-to-end connectivity through the backbone: flooding all data packets through the backbone, modifying an ad-hoc routing protocol such as OLSR and AODV, or using a P2P overlay for routing purposes. Running extensive simulations in OMNeT++, our results highlight the strengths and weaknesses of each approach. Flooding, albeit a very simple approach, appears to be quite competitive with more complex routing solutions, with good performance and low overheads.

The frequent dissemination of safety-related beacons among neighboring vehicles in VANET is fundamental for cooperative awareness. Nevertheless, this has over the years raised a major security concern hence the current state-of-the-art requires all safety-related beacons to carry a certificate and a digital signature as a security mechanism to ensure authenticity and integrity. Unfortunately, this security mechanism is characterized by an increase in the size of a beacons payload which as a result, induces an overhead in communication under dense traffic conditions.

Several works have been published in the literature investigating how to reduce this overhead without compromising the level of security achieved, as well as vehicle cooperative awareness. The Neighbor-based Certificate Omission scheme, which conveys the general idea of a vehicle attaching a certificate to its beacon based on changes it observes from its neighboring table was proposed to address this issue. However, on evaluating the scheme under a dense traffic scenario, it was observed that the scheme reduced the level of achieved cooperative awareness among vehicles as it was unable to obtain a fair balance between the number of incurred cryptographic packet loss (packets dropped because the vehicle had no corresponding certificate to verify it) and network packet loss (packets dropped because of network channel congestion).

In this paper, we propose a Transmission Power-control Certificate Omission scheme, which seeks to achieve a better balance between the number of incurred cryptographic packet loss (CPL) and network packet loss (NPL) to maximize vehicle cooperative awareness even under dense traffic conditions. Unlike previously proposed schemes, we efficiently control channel load by adopting a congestion detection and congestion control algorithm in our scheme. The simulation results indicate that our proposed scheme can achieve a better balance between the number of incurred CPL and NPL and can maximize vehicle cooperative awareness even under dense traffic conditions.

This paper aims to use driving data to create a profile of the driver behavior, which can be then added as an additional layer to the Local Dynamic Map of the vehicle. The main contribution of the paper consists of using the Spherical KMeans Clustering, an unsupervised clustering algorithm for multidimensional datasets, to segment the continuous driving data into multiple segments (hyperspheres). Unlike the state of the art, this helps in studying the behavior since all the data will be processed at the same time regardless of the number of features. The generated hyperspheres are an abstract form of the initial numerical values, and can be contribute to a better representation of the driver behavior. We used the UAH Dataset [9] to present the proposed approach, and the cross-validation technique to evaluate the segmentation results.

Internet of Things (IoT) is a major component of the connected world. With billions of battery-powered devices connected to the internet, energy and bandwidth consumption become significant issues. Embedding intelligence/cognition in the apparatus is recognized as one of the solutions to mitigate these issues. Global Positioning System (GPS) is recognized as one of the most energy-consuming mobile sensors in smart vehicles/systems. This paper proposes a smart adaptive sampling method for GPS sensors using the accelerometer data. Our approach adapts the sampling frequency of the GPS sensor according to the data stream of the accelerometer, without causing significant distortions to the data. In our experiment, we could reduce the GPS sensing by 78% while preserving an accuracy of 91.4%.

Connected Autonomous Vehicles (CAV) are expected to revolutionize the transportation sector. However, given that CAV are connected to internet, they face a principal challenge to ensure security, safety and confidentiality. It is highly valuable to provide a real-time and proactive anomaly detection approach for Vehicular Ad hoc Network (VANET) exchanged data since such an approach helps to trigger prompt countermeasures to be undertaken allowing the damage avoidance. Recent machine learning methods show great efficiency, especially due to their capacity to handle nonlinear problems. However, an accurate anomaly detection in a space–time series is a challenging problem because of the heterogeneity of space–time data and the spatio-temporal correlations. An anomalous behavior can be seen as normal in different context. Thus, using one deep learning model to classify the observations into normal and abnormal or to identify the type of the anomaly is usually not efficient for large high-dimensional multi-variate time-series datasets. In this paper, we propose a stepwise method in which the time-series data are clustered on spatio-temporal clusters using Long Short Term Memory (LSTM) auto-encoder for dimension reduction and Grey Wolf Optimizer based clustering. Then, the anomaly detection is performed on each cluster apart using a hybrid method consisting of Auto-Encoder for feature extraction and Convolution Neural Network for classification. The results shows an increase in the accuracy by \(2\%\) in average and in the precision by approximately \(1.5\%\).

In its native version, the Controller Area Network (CAN) bus protocol used in most personal vehicles does not use any encryption nor message authentication mechanism. In order to test solutions dedicated to signing messages and protecting CAN infrastructures from external attacks, we built CAn enCryption simulAtion mOdule (Cacao). It is a CAN bus simulation platform dedicated to simulate a real CAN network. The following work presents this tool and the signature solution we did integrate in it to implement various vulnerabilities protection among the CAN bus.