In this paper, we propose a dynamic resource optimization scheme for a heterogeneous wireless network combining with a radio frequency (RF) access point (AP) and multiple visible light communication (VLC) APs. For adapting to the stochastic media content arrival rates and dynamic wireless channel conditions, a two-timescale stochastic network resource optimization problem is formulated by employing the Lyapunov optimization technique. To this end, we propose an online two-timescale adaptive network resource optimization (ANRO) framework. At the ANRO, the network selections for user equipments (UE) are operated at a large timescale, while the resource allocations for both RF and VLC APs are operated at a small timescale. Simulation results exhibit that the ANRO not only stabilizes the network but also significantly reduces the energy consumption than other existing schemes.
Packet forwarding plays an important role in inter-vehicle communications. A predicted forwarding is a promising solution to address high mobility issues in a Vehicular Ad-hoc Network (VANET) as it increases the ratio of packet transmissions. The predicted forwarding scheme proposed, termed Vehicular-to-vehicular Urban Network (V2VUNet), is based on an angle measurement between a sender and a receiver and reduces the flooding mechanism time. Angle measurements can be used to determine the location of a relay, with the addition of a direction value to the relay's location at a current time, such that the sender can predict the future relay's location. This paper proposes and evaluates the predicted forwarding scheme which is indicated by the Vertical Relative Angle (VRA) parameter. It maintains a longer life time duration of communicating vehicles.
Participatory sensing has been emerging as an economical and practical way to collect and share information on the surrounding environment. The information includes both physical data produced by embedded sensors in the smart device and the observation and reasoning from human participants. However, most of the existing participatory sensing applications lack multi-functional capabilities; they are designed to collect one single or limited types of information. In this paper, we first identify the most important requirements of a multi-purpose participatory sensing application, based on a comprehensive literature review. We then propose a generic participatory sensing framework — Citizense — that emphasizes ease of use, and allows the creation, execution of context-aware, multi-purpose participatory sensing campaigns, and analysis of the results. We discuss the architecture, design and prototype implementation, and compare it against the earlier identified requirements.
We investigate the problem of minimising power consumption in dense Wireless Local Area Networks (WLANs), by optimally establishing the association of User Terminals (UTs) to Access Points (APs). This management allows to switch off some APs, granting important power savings, while at the same time guaranteeing to satisfy the data rate requirements of all UTs. The considered WLAN power minimization problem can be formulated as an Integer Linear Programming (ILP) model and can be in principle solved by any commercial optimization solver. However, the problem is NP-hard and, as we show through thorough computational tests, even a last generation state-of-the-art solver like IBM ILOG CPLEX can have difficulties in finding solutions of good quality in short amount of time, as required in real WLAN deployments. As a consequence, we propose two new fast heuristic algorithms for WLAN power minimization. Furthermore, we show that, in some cases, also a proper setting of the parameters of CPLEX can compute solutions associated with good power savings in a reasonable time.
The transit of cellular communications over WiFi networks is expected to account for 60% of mobile traffic in 2019. Consequently, mobile traffic management and advanced offloading have become hot topics that will enable operators to optimize resources, and improve QoS/QoE for high bandwidth mobile multimedia applications and services. On the other hand, concepts such as Network Function Virtualisation (NFV), Software Defined Network (SDN) and open protocols such as OpenFlow (OF) become the key enablers for the next generation of mobile networks. In this context, the objective of this paper is to investigate the opportunity of offloading the 3GPP Radio Access Network (RAN) traffic through WiFi access networks by the use of Wireless Mesh Networks (WMNs) that would be formed by the mobile phones themselves. This use case will demonstrate establishment and control of an IP wireless mesh network by the use of NFV controller (through the SDN control plane). We validated our work using NEON, an SDN solution developed by CEA LIST that supports fast devices configuration and services deployment in dynamic and unconfigured infrastructures contexts.
Typically, in low-power and lossy networks nodes communicate with servers over the Internet. Nodes collaborate wirelessly to relay their data to a gateway. A large-scale network can have multiple gateways. Selecting a proper gateway can have an immense impact on the network's performance. In this paper, we present joint routing and gateway selection protocols for low-power and lossy networks. To analyse the impact of per-packet and per-flow gateway selection protocols on a network's performance, we present two different protocols. Our protocols select the best gateway on per-packet and per-flow basis. Our protocols are based on a metric composed of the end-to-end available bandwidth and contention. Simulation and testbed results demonstrate that per-packet and per-flow gateway selection protocols perform similarly, and our protocols achieve up to 20% higher packet delivery ratio, 34% lower delay, and 40% lower retransmissions as compared to existing protocols.
Flying Ad-hoc Networks (FANETs) [1] are mobile ad-hoc networks formed by small and medium-sized UAVs. Nowadays, most UAVs are equipped with omnidirectional antennas. In addition, most of the existing routing protocols were designed assuming the use of omnidirectional antennas. Directional antennas have the potential to increase spatial reuse, save the battery's energy, and substantially increase the transmission range. However, these benefits come with a few challenges. Existing directional MAC protocols deal with these challenges, mostly in static ad-hoc networks. We define DA-FANETs as FANETs where directional antennas are used. In this paper, we investigate the performance of existing routing protocols in DA-FANETs. First we implement an 802.11b-based directional MAC protocol that embodies common directional MAC protocol features. Then we evaluate the following routing protocols in DA-FANET scenarios: AODV, OLSR, RGR, and GRP. The results show that each antenna beamwidth has an optimal network size that gives the highest amount of routing. In addition, RGR is the best option for DA-FANETs while GRP is the worst. Although RGR shows the best performance of all, it still leaves a lot of room for improvement with a PDR sitting at just 85% and a relatively high amount of overhead in terms of the number of transmissions performed.
Modular-based channel hopping (CH) rendezvous algorithms can provide guaranteed rendezvous for Cognitive Radio Networks (CRNs) without time synchronization or Common Control Channels (i.e., blind rendezvous). The Enhanced Jump-Stay (EJS) scheme [1] has now been recognized as arguably the best in terms of Maximum-Time-To-Rendezvous (MTTR) and bounds for the Expected-Time-To-Rendezvous (ETTR) for users with a different number of channels (asymmetric). In [2], we developed a probabilistic channel detecting jamming attacks that dramatically decreased the rendezvous success rates of EJS and developed the Random Enhanced Jump Stay (REJS) CH rendezvous algorithm that largely mitigated those jamming attack. Here we provide extensions of EJS and REJS and provide guidelines when they should be used. The focus in [2] was jamming mitigation but here we carefully analyze the performance of several new algorithms while still guaranteeing bounded MTTR and improved the ETTR over EJS. In fact, it appears EJS should seldom be used. We show our jump-stay extensions are better than EJS with significant decreases in the average TTR.
A number of applications of Intelligent Transportation Systems (ITS) rely on multi-hop broadcasting in order to disseminate useful traffic information to other vehicles located in a geographical area. However, the conventional broadcast mechanism may lead to the so-called broadcast storm problem, which causes a high level of contention, collision, and congestion due to an excessive number of broadcast packets. We propose a network layer congestion control algorithm based on Contention Based Forwarding mechanism, named CBF2C. The novelty here is that the proposed proactive strategy prevents the broadcast storm by reducing channel occupation before congestion happens, while exploiting transmission redundancy whenever the channel status allows. Our proposal is compared with the conventional solutions in a highway environment. Simulation results show a significantly increased packet delivery ratio and an acceptable levels of end to end delay for the proposed algorithm. Moreover, CBF2C ensures a trade off between efficient use of the radio resource and communication performance.
In this paper, we study the problem of jointly solving the contention and congestion distributed control problem in a bounded queue wireless ad-hoc network. The resulting flow rates satisfy fairness criteria according to a given Network Utility Maximization (NUM) function. In recent years a number of papers have presented solutions to this problem that are based on network utility maximization algorithms. However, this work typically necessitates either complex computations, heavy signaling/control overhead, and/or approximated sub-optimal results. In this paper, we combine a specific network utility maximization problem with a simple and efficient queue management mechanism that we believe is appropriate for wireless ad-hoc networks. We employ and adapt the IEEE 802.11 protocol to work with the utility maximization algorithm for contention optimization. Finally, we show via NS-3 simulations that the proposed Cross-Layer Design (CLD) significantly outperforms standard protocols such as TFRC.
Privacy is a main concern for mobile network users, and there are many proposed enhancements for the protection of the long-term subscription identifier. Some enhancements require asymmetric key operations, which increase both processing requirements and protocol message sizes. To the best of our knowledge, there has been no practical implementation feasibility study of these enhancements using commodity mobile devices. Neither is it clear whether the enhancements are sufficient. This paper highlights privacy weaknesses, when the long-term subscription identifier is used in Paging procedures, and proposes new ways to resolve these. Further, the paper evaluates an Android implementation of one of the enhancements, which includes the asymmetric scheme Elliptic Curve Integrated Encryption Scheme (ECIES). We conclude that it is feasible to implement asymmetric encryption methods for the long-term subscription identifier and that the highlighted privacy weaknesses can be efficiently countered. This removes another set of obstacles for realizing the protection in mobile network standards.
The number of communication networks' users using bandwidth-intensive application, such as video-streaming, video-conferencing, and multi-player gaming is on the rise. To handle a large number of such users, the bandwidth of cellular networks is becoming scarce. Nowadays, most user equipments (mobile phones, tablets, net-books, etc) support multiple radio access technologies (RATs), such as long-term evolution (LTE) and WiFi. Therefore, cellular network service providers are using multiple RATs to increase their networks' capacity, and such networks are called heterogeneous networks (HetNets). Usually, in a HetNet, a user equipment (UE) selects an access network based on the network service provider's priority list. An access network selection based on the priority list may not result in an efficient utilization of the available networks' resources. Therefore, in this paper we focus on continuously monitoring the state of a HetNet using received signal strength indicator (RSSI), throughput, and delay metrics. Based on these metrics, we present access network selection and switching algorithms. Our experimental results demonstrate that, if the aggregate data generation rate of UEs in a HetNet is higher than the achievable data rate on LTE uplink, RSSI is the best metric for the access network selection and switching. Otherwise, the metrics demonstrate similar performance.
New retail applications induce integration of many components as NFC (Near Field Communication), M2M (Machine-to-Machine), IoT (Internet of Things), Web applications, etc. in large business sectors, such as payment services, products manufacturing, supply chain management, etc. Sensors, integrated in everyday products may facilitate threats as users tracking and profiling. An increasing concern about privacy threats posed by data affluence and device ubiquity takes place. This paper presents a privacy scheme for retail applications, discusses challenges related to customer profiling, client consent, device and information security. To protect customers' data, we propose a Markovian game, with detailed states, actions, strategies and transitions available for data holder to reach a compromise between privacy concessions and incentive motivation proposed by data requester. Numerical results are used to analyze and evaluate the game theory-based model.
In this paper, we present an approach for the analysis of parking slots by exploiting urban surveillance networks. This work has been developed in a project which aims to facilitate to the car drivers the access to available car parking slots. This paper describes a framework that exploits video streams provided by urban networks of cameras. The objective is to remotely determine the occupancy status of parking slots. Besides, the paper presents a case study and a discussion about a broadcasting strategy that could be used to share the positions of identified vacant parking slots with drivers searching an available parking slot.
Video over vehicular networks continues to receive warranted attention, with envisioned applications having the potential to present entirely new opportunities and revolutionise existing services. Many video systems have been proposed, ranging from safety to advertising. We propose a novel system for VANETs, namely the TArgeted Remote Surveillance (TARS) module for the existing Greedy Perimeter Stateless Routing (GPSR) protocol which permits multiple mobile vehicles to request and receive live video feeds from vehicles within a select geographic region. The multi-hop, vehicle-to-vehicle system enables mobile units to surveil a target area in real time by leveraging the dashboard cameras of vehicles moving within the target region. We combine several proposed extensions to the core protocol to introduce a dynamic real time congestion aware clustering scheme to achieve this. Our proposed system is compared against existing routing protocols using mobility data from Nottingham. GPSR-TARS outperforms the protocols assessed in key criteria crucial for meeting the quality of service demands of live multimedia dissemination.
To provide better usage of spectrum, regulators, such as the FCC and Ofcom, have been considering how to share spectrum between operators and also between different types of network. Two important aspects of any mobile service that need to be considered are how to ensure the appropriate level of service to individual users and how to minimize power consumption. This paper addresses the issue by applying orthogonal spectrum sharing, with the objective being to serve as many as possible users that meet their QoS requirement, together with minimizing the overall power transmitted. Multi-objective Game Theory is split into sequential single-objective optimizations and the paper presents results of simulations that show that this approach can achieve significant savings in power while maintaining the number of qualified users.
In Vehicular Ad Hoc Networks (VANETs), a vehicle could be identified and tracked by eavesdropping its messages (e.g., beacons) by an adversary, since these messages contain personal information such as the location of the vehicle. This attack leads to threats on the user's location privacy. Frequent changing pseudonyms are accepted as a solution to protect the location privacy of users in VANETs. In this paper, we propose an efficient pseudonym changing strategy to provide location privacy in VANETs. In addition, we describe a novel tracking attack, that is, the cheating attack. To confront this attack, we develop a cheating detection mechanism. Finally, we show by simulations that the proposed scheme is effective.
Near Field Communication (NFC) is a prominent short-range, contact-less communication technology, which is rapidly getting popular in modern smart devices. For communication between two active devices via NFC, applications generally choose the peer-to-peer operation mode. In this paper, we exploit the possibility of using the NFC read/write mode, designed primarily for unidirectional data transfer from an active NFC reader to a passive NFC tag, for bi-directional half-duplex communication between two active NFC devices. The advantages of using the NFC read/write mode include low protocol overhead and permitting different data formats. However, the challenges are avoiding the reader collision problem, maintaining a secure session, and completing all transactions in an acceptable time frame. In this paper, we address the above challenges and propose a methodology for efficient communication between active NFC devices using NFC read/write mode. To evaluate the scheme, we design a secure Multi-Factor Authentication (MFA) system that requires bi-directional communication for mutually authenticating two NFC devices. The proposed methodology is experimentally verified using NFC-enabled Android smartphones and a Kerberos server as the third-party authenticator.
In emergency and crisis operations, group communication is essential to coordinate rescue efforts. Mobile Ad Hoc Networks (MANETs) can provide dynamic and resilient communication services in areas without a working communication infrastructure. The connectivity of MANETs can be improved using an elevated network node due to larger ground coverage, thus improving node reachability. Simplified Multicast Forwarding (SMF) is a scheme to efficiently distribute group communication packets in MANETs. In this paper, we explore methods to utilize SMF in a MANET supported by an elevated network node. We show that an elevated network node can improve the packet delivery ratio in a MANET with SMF-distribution of multicast packets. Further gains are made by unicasting the packets from the source to the elevated network node, moving the initial point of the SMF-distribution to the elevated network node. The consequences of a potential unicast packet loss are lessened by the source retransmitting the packet as multicast if it fails to receive the SMF-forwarded packet from the elevated network node. We achieve a high packet delivery ratio while maintaining a low cost for most topology sizes. Finally, the paper outlines future research directions for group communication in MANETs.
Cyber-Physical Systems (CPS) represent a fundamental link between information technology (IT) systems and the devices that control industrial production and maintain critical infrastructure services that support our modern world. Increasingly, the interconnections among CPS and IT systems have created exploitable security vulnerabilities due to a number of factors, including a legacy of weak information security applications on CPS and the tendency of CPS operators to prioritize operational availability at the expense of integrity and confidentiality. As a result, CPS are subject to a number of threats from cyber attackers and cyber-physical attackers, including denial of service and even attacks against the integrity of the data in the system. The effects of these attacks extend beyond mere loss of data or the inability to access information system services. Attacks against CPS can cause physical damage in the real world. This paper reviews the challenges of providing information assurance services for CPS that operate critical infrastructure systems and industrial control systems. These methods are thorough measures to close integrity and confidentiality gaps in CPS and processes to highlight the security risks that remain. This paper also outlines approaches to reduce the overhead and complexity for security methods, as well as examine novel approaches, including covert communications channels, to increase CPS security.
