Chiara Boldrini

HiBOp: a History Based Routing Protocol for Opportunistic Networks

In opportunistic networks the existence of a simultaneous path between a sender and a receiver is not assumed. This model (which fits well to pervasive networking environments) completely breaks the main assumptions on which MANET routing protocols are built. Routing in opportunistic networks is usually based on some form of controlled flooding. But often this results in very high resource consumption and network congestion. In this paper we advocate context-based routing for opportunistic networks. We provide a general framework for managing and using context for taking forwarding decisions. We propose a context-based protocol (HiBOp), and compare it with popular solutions, i.e., Epidemic Routing and PROPHET. Results show that HiBOp is able to drastically reduce resource consumption. At the same time, it significantly reduces the message loss rate, and preserves the performance in terms of message delay.

HCMM: Modelling spatial and temporal properties of human mobility driven by users' social relationships

In Mobile Ad Hoc Networks (MANETs), the mobility of the network users can heavily affect the performance of networking protocols because it causes sudden connectivity changes and topological variations. This is even more important in recent promising paradigms proposed in this field, such as opportunistic and delay tolerant networks. For this reason, it is important to understand the characteristics of the user movements in order to properly handle mobility when designing networking protocols for mobile ad hoc networks. In addition, it is highly desirable to have a mobility model that accurately reproduces the user mobility, thus enabling researchers to evaluate, either analytically or by means of simulations, their protocols under realistic mobility conditions. Recently, there have been many studies aimed to uncover the nature of human movements. In this paper, based on recent literature, we identify three main properties that are fundamental to characterize human mobility. Then, we propose a mobility model (HCMM) that integrates all these three features. To the best of our knowledge, the model proposed is the first one that combines notions about the sociality of users with spatial properties observed in real users movement patterns, i.e., their preference to spend time in a limited number of popular locations and to preferentially select short distances over longer ones. We study the HCMM both through simulation and analysis. Based on this study, we highlight some of its important temporal and spatial features, and we show that they are correctly reproduced in terms of key indicators such as jump size and inter-contact time distribution.

Human mobility models for opportunistic networks

Mobile ad hoc networks enable communications between clouds of mobile devices without the need for a preexisting infrastructure. One of their most interesting evolutions are opportunistic networks, whose goal is to also enable communication in disconnected environments, where the general absence of an end-to-end path between the sender and the receiver impairs communication when legacy MANET networking protocols are used. The key idea of OppNets is that the mobility of nodes helps the delivery of messages, because it may connect, asynchronously in time, otherwise disconnected subnetworks. This is especially true for networks whose nodes are mobile devices (e.g., smartphones and tablets) carried by human users, which is the typical OppNets scenario. In such a network where the movements of the communicating devices mirror those of their owners, finding a route between two disconnected devices implies uncovering habits in human movements and patterns in their connectivity (frequencies of meetings, average duration of a contact, etc.), and exploiting them to predict future encounters. Therefore, there is a challenge in studying human mobility, specifically in its application to OppNets research. In this article we review the state of the art in the field of human mobility analysis and present a survey of mobility models. We start by reviewing the most considerable findings regarding the nature of human movements, which we classify along the spatial, temporal, and social dimensions of mobility. We discuss the shortcomings of the existing knowledge about human movements and extend it with the notion of predictability and patterns. We then survey existing approaches to mobility modeling and fit them into a taxonomy that provides the basis for a discussion on open problems and further directions for research on modeling human mobility.

Design and performance evaluation of ContentPlace, a social-aware data dissemination system for opportunistic networks

In this paper we present and evaluate ContentPlace, a data dissemination system for opportunistic networks, i.e., mobile networks in which stable simultaneous multi-hop paths between communication endpoints cannot be provided. We consider a scenario in which users both produce and consume data objects. ContentPlace takes care of moving and replicating data objects in the network such that interested users receive them despite possible long disconnections, partitions, etc. Thanks to ContentPlace, data producers and consumers are completely decoupled, and might be never connected to the network at the same point in time. The key feature of ContentPlace is learning and exploiting information about the social behaviour of the users to drive the data dissemination process. This allows ContentPlace to be more efficient both in terms of data delivery and in terms of resource usage with respect to reference alternative solutions. The performance of ContentPlace is thoroughly investigated both through simulation and analytical models.

Modelling data dissemination in opportunistic networks

In opportunistic networks data dissemination is an important, although not widely explored, topic. Since opportunistic networks topologies are very challenged and unstable, data-centric approaches are an interesting direction to pursue. Data should be proactively and cooperatively disseminated from sources towards possibly interested receivers, as sources and receivers might not be aware of each other, and never get in touch directly. In this paper we consider a utility-based cooperative data dissemination system in which the utility of data is defined based on the social relationships between users. Specifically, we study the performance of this system through an analytical model. Our model allows us to completely characterise the data dissemination process, as it describes both its stationary and transient regimes. After validating the model, we study the systems behaviour with respect to key parameters such as the definition of the data utility function, the initial data allocation on nodes, the number of users in the system, and the data popularity.

From MANET to people-centric networking: Milestones and open research challenges

In this paper, we discuss the state of the art of (mobile) multi-hop ad hoc networking with the aim to present the current status of the research activities and identify the consolidated research areas, with limited research opportunities, and the hot and emerging research areas for which further research is required. We start by briefly discussing the MANET paradigm, and why the research on MANET protocols is now a cold research topic. Then we analyze the active research areas. Specifically, after discussing the wireless-network technologies, we analyze four successful ad hoc networking paradigms, mesh networks, opportunistic networks, vehicular networks, and sensor networks that emerged from the MANET world. We also present an emerging research direction in the multi-hop ad hoc networking field: people centric networking, triggered by the increasing penetration of the smartphones in everyday life, which is generating a people-centric revolution in computing and communications.

Context and resource awareness in opportunistic network data dissemination

Opportunistic networks are challenging mobile ad hoc networks characterised by frequent disconnections and partitioning. In this paper we focus on data dissemination services, i.e. cases in which data should be disseminated in the network without a priori knowledge about the set of intended destinations. We propose a general autonomic data dissemination framework that exploits information about the userspsila context and social behaviour, to decide how to replicate and replace data on nodespsila buffers. Furthermore, our data dissemination scheme explicitly takes into account resource constraints, by jointly considering the expected utility of data replication and the associated costs. The results we present show that our solution is able to improve data availability, provide fairness among nodes, and reduce the network load, with respect to reference proposals available in the literature.

Performance modelling of opportunistic forwarding under heterogenous mobility

Abstract The delay tolerant networking paradigm aims to enable communications in disconnected environments where traditional protocols would fail. Opportunistic networks are delay tolerant networks whose nodes are typically the users’ personal mobile devices. Communications in an opportunistic network rely on the mobility of users: each message is forwarded from node to node, according to a hop-by-hop decision process that selects the node that is better suited for bringing the message closer to its destination. Despite the variety of forwarding protocols that have been proposed in the recent years, there is no reference framework for the performance modelling of opportunistic forwarding. In this paper we start to fill this gap by proposing an analytical model for the first two moments of the delay and the number of hops experienced by messages when delivered in an opportunistic fashion. This model seamlessly integrates both social-aware and social-oblivious single-copy forwarding protocols, as well as different hypotheses for user contact dynamics. More specifically, the parameters of model can be solved in a closed form in the case of exponential and Pareto inter-meeting times, two popular cases emerged from the literature on human mobility analysis. In order to exemplify how the proposed framework can be used, we discuss its application to two case studies with different mobility settings. Then, we discuss how the framework can be also extended to accommodate inter-meeting times following a hyper-exponential distribution. This case is particularly relevant as hyper-exponential distributions are able to approximate the large class of high-variance distributions (distributions with coefficient of variation greater than one), which are those more challenging, e.g., from the delay standpoint. Finally, we provide a validation for the framework with both ideal contacts (i.e., exactly following a given distribution) and contacts extracted from a real mobility trace. This evaluation highlights the strength of the framework in terms of its ability both to provide very accurate predictions under ideal mobility and to effectively approximate the behaviour of the delay moments under real mobility.

Autonomic behaviour of opportunistic network routing

In opportunistic networks, end-to-end communication among users does not require a continuous end-to-end path between source and destination. Network protocols are designed to be extremely resilient to events such as long partitions, node disconnections, etc. which are very features of this type of self-organising ad hoc networks. This is achieved by temporarily storing messages at intermediate nodes, waiting for future opportunities to forward them towards the destination. Clearly, designing routing and forwarding schemes is one of the main challenges in this environment. In this article, we provide a survey of the main approaches to routing in purely infrastructure-less opportunistic networks, by classifying protocols based on the amount of context information they exploit. Then, we provide an extensive quantitative comparison between representatives of protocols that do not use any context information, and protocols that manage and exploit a rich set of context information. Mainly, we focus on the suitability of protocols to adapt to the dynamically changing network features, as resulting from the user movement patterns that are driven by their social behaviour. Our results show that context aware routing is extremely adaptive to dynamic networking scenarios, and, with respect to protocols that do not use any context information, is able to provide similar performance in terms of delay and loss rate, by using just a small fraction of the network resources.

Modelling inter-contact times in social pervasive networks

Thanks to the diffusion of mobile user devices (e.g. smartphones) with rich computing and networking capabilities, we are witnessing an increasing integration between the cyber world of devices and the physical world of users. In this perspective, a possible evolution of pervasive networking (hereafter referred to as social pervasive networks, SPNs) consists in closely mapping human social structures in the network of the devices. Links between devices would correspond to social relationships between users, and communication events between devices would correspond to communications between users. It can be shown that fundamental convergence properties of SPN forwarding protocols are determined by the distributions of inter-contact times between the individual nodes (i.e. the time elapsed between two successive communication events between the nodes). Individual pairs inter-contact times are hard to completely charaterise, while the distribution of the aggregate inter-contact times is often a much more convenient figure. However, the aggregate distribution is not always representative of the individual pairs distributions. Therefore using it to characterise the properties of SPN forwarding protocols might not be correct. In this paper we provide an analytical model based on fundamental models of human social networks from the anthropology literature, which shows the exact dependence between the two in heterogeneous SPNs. Moreover, we use the model to i) study cases in which analysing the aggregate distribution is not enough, and ii) find sufficient conditions that guarantee that studying the aggregate distribution is enough to characterise the properties of SPN forwarding protocols.