تاثیر رفتار انسان در حمل و نقل فرصت طلب اجتماعی

The current Internet design is not capable of supporting communications in environments characterized by very long delays and frequent network partitions. To allow devices to communicate in such environments, delay-tolerant networking solutions have been proposed by exploiting opportunistic message forwarding, with limited expectations of end-to-end connectivity and node resources. Such solutions envision non-traditional communication scenarios, such as disaster areas and development regions. Several forwarding algorithms have been investigated, aiming to offer the best trade-off between cost (number of message replicas) and rate of successful delivered message. Among such proposals, there has been an effort to employ social similarity, inferred from users’ mobility patterns, to improve opportunistic forwarding solutions. However, such proposals present two major drawbacks: first, they focus on distribution of intercontact times over the complete network structure, ignoring the impact that human behavior has on the dynamics of the network; and second, most of the proposed solutions look at challenging networking environments where networks have low density, ignoring the potential use of delay-tolerant networking to support low cost communications in networks with higher density, such as urban scenarios. This paper presents a study of the impact that human behavior has on opportunistic forwarding. Our goal is twofold: i) to show that performance in low and high density networks can be improved by taking the dynamics of the network into account; and ii) to show that delay-tolerant networking can be used to reduce communication costs in networks with high density by considering the users’ behavior.

Wireless devices are becoming more portable and have increased capabilities (e.g., processing, storage), which is creating the foundations for the deployment of pervasive wireless networks based on a large set of personal devices (e.g., smartphones, embedded devices). Additionally, wireless technology has been extended to allow direct communication aiming to support safety information exchange (i.e., vehicle-to-vehicle), 3G offloading (i.e., device-to-device), and to overcome the lack of infrastructure entities (i.e., Wi-Fi direct). The combination of pervasive wireless devices and direct communication solutions can be used to support the deployment of two major type of applications: end-to-end communication in development regions, since today’s Internet routing protocols may operate poorly in environments characterized by very long delay paths and frequent network partitions; and low cost communications, namely data sharing, in urban scenarios, to bypass expensive data mobile communications and the unreliable presence of open Wi-Fi access points. These networking scenarios (from development regions to urban scenarios) are characterized by network graphs with different densities, which pose several challenges in terms of data forwarding. The goal of this paper is the investigation of the impact that human behavior may have on opportunistic forwarding, namely the awareness about users’ social and data similarities. Most of the prior art has been studying data transfer opportunities between wireless devices carried by humans, by looking at the distribution of intercontact times, which is the time gap separating two contacts between the same pair of devices [1]. In challenging networking environments, opportunistic contacts among mobile devices may improve communications among peers, as well as content dissemination, mitigating the effects of network disruption. This gave rise to the investigation of opportunistic networks, of which Delay-Tolerant Networks (DTN) are an example, encompassing different forwarding proposals to quickly transfer data between two peers even in the absence of an end-to-end path between them. Such proposals range from flooding data [2] in the network, up to solutions that take into account users’ social interactions [3], [4], [5], [6], [7] and [8]. In the latter case, wireless contacts are aggregated into a social graph, and a variety of metrics (e.g., centrality and similarity) or algorithms (e.g., community detection) have been proposed to assess the utility of a node to deliver data or bring it closer to the destination. It is worth highlighting that the structure of such graphs is rather dynamic, since users’ social behavior and interactions vary throughout their daily routines. This brings us to our first challenge: forwarding algorithms should be able to exploit social graphs that reflect people’s dynamic behavior. Prior art have studied forwarding algorithms that consider only the global network structure, without taking people’s behavior into account [9]. In this paper, we show that forwarding algorithms that exploit social graphs reflecting variations in people’s daily routines are able to improve the performance of opportunistic forwarding. Our second challenge is to analyze how to expand the deployment of DTN technology, which is normally seen only as useful to allow communications in challenging environments, such as development regions. For this study, we focus on data sharing since we assume that this should be the most popular application to take advantage of low cost communications in dense networks, such as in an urban scenario. In this case, we study two hypothesis to ensure good performance of opportunistic forwarding when the density of the network increases: i) forwarding based on social graphs, where aggregation is based only on social similarities; and ii) forwarding based on behavior graphs, where aggregation is done by combining different aspects of human behavior, such as social similarities and data similarities (derived from the interests that users demonstrate in specific types of data). Hence, in this paper we aim to investigate the possibility of developing an opportunistic forwarding system able to support low cost services in dense networking scenarios as well as basic services in extreme networking conditions, by exploiting users’ social and data similarities. Our work shows which type of opportunistic forwarding scheme is more suitable for delay-tolerant applications, based on the density of the network in scenarios spanning from developing regions to urban environments. Our findings lead to a new research challenge aiming to expand the impact of DTNs: the investigation of self-awareness mechanisms able to adapt their forwarding schemes based on users’ context, namely the density of the pervasive network where the user is. The remainder of the paper is structured as follows. Section 2 aims to motivate our work, namely in what concerns the goal to study methods to expand the deployment of DTNs, and the impact that a better understanding of human behavior can have in the development of efficient forwarding solutions. In Section 3 we present our definition of network density based on the deployment scenarios that we look at to pursuit our study and experiments. Section 4 presents a set of forwarding algorithms that are considered in our study, including our proposals. In Section 5, we analyze the performance of opportunistic forwarding over different network densities. Section 6 concludes our work, and identifies future research challenges.

Opportunistic forwarding proposals comprise social similarity (e.g., common social groups and communities, node popularity, levels of centrality, social relationships and interactions, user profiles) and/or data similarity (e.g., shared interests, interest of users in the content traversing network, content availability, type of content) aspects. Additionally, these proposals may (or not) take into account the network dynamics. In this paper, we highlight the possibility of having opportunistic forwarding that can provide support to i) basic services in extreme networking conditions (e.g., end-to-end communication in development regions); and ii) low-cost services in dense (e.g., urban) networking scenarios. Our results show that opportunistic forwarding, based on either social or data similarity metrics may present satisfactory performance over challenged (i.e., development region) and dense (i.e., urban) scenarios, even when forwarding proposals do not considering the dynamism observed in the users behavior. However, when social and data similarity metrics are considered alongside with the dynamics of user behavior, opportunistic forwarding presents interesting performance gains in both scenarios: average results over sparse and dense networks, shows that performance improvements go up to 54% regarding delivery capability while latency and cost can be reduced by 45% and 99% respectively, when compared to forwarding solely based on data similarity and completely agnostic to the dynamism found in user behavior. These findings point to a new research challenge regarding the impact of DTN application: the investigation of self-awareness mechanisms able to adapt their forwarding schemes based on the context of the user, namely the density of the network where he/she is currently. Based on our results, by i) being able to adapt to the characteristics of the scenario (sparse vs. dense), considering the dynamism of users’ behavior; and ii) also employing data similarity metrics, opportunistic forwarding allows the exchange of data in both development regions and urban settings.