Section: New Results

Massive mobile dense wireless networks

Participants : Aline Carneiro Viana, Ana Cristina B. Kochem Vendramin, Kanchana Thilakarathna, Eduardo Mucceli.

routing protocols, analytical models, content distribution.

Scientific achievements

Social Relationship Classified

Understanding human mobility is of fundamental importance when designing new communication protocols that exploit opportunistic encounters among users. In particular, human behavior is characterized by an elevated rate of regularity, but random events are always possible in the routines of individuals as hardly predictable situations that deviate from the regular pattern and are unlikely to arise repeatedly in the future. These random events veil the ordinary patterns by introducing a significant amount of noise, thus making the process of knowledge discovery in social dataset a complex task. However, the ability to accurately identify random and social events in large datasets is essential to social analysis as well as to applications that rely on a precise description of human routines, such as recommendation systems, forwarding strategies and ad-hoc message dissemination schemes focusing on coverage efficiency with a limited number of redundant messages. In such a context, we have proposed a strategy to analyze wireless network scenarios where mobile users interact in a rational manner, reflecting their interests and activity dynamics. Our strategy, named Random rElationship ClASsifier sTrategy (RECAST), allows to classify user relationships, separating random interactions from different kinds of social ties. The goal is achieved by observing how the real system differs from an equivalent one where entities decisions are completely random. We have evaluate the effectiveness of RECAST classification on datasets of real-world user contacts in diverse networking contexts. Our analysis unveils significant differences in the relationship dynamics of the datasets, proving that the evaluation of network protocols on a single dataset cannot lead to conclusions of general validity.

Social-aware Forwarding Protocol

Pervasiveness of computing devices, ubiquitous wireless communication, emergence of new applications, and cloud services are examples of current new emerging factors that emphasize the increasing need for adaptive networking solutions. The adaptation, most of the time, requires the design of more interdisciplinary approaches as those inspired by techniques coming from biology, social structures, games, and control systems. The approach we consider brings together solutions from different but complementary domains - i.e., networking, biology, and complex networks - aiming to deal with the problem of efficient data delivery in mobile and intermittently connected networks. For this, we have designed the Cultural Greedy Ant (CGrAnt) protocol to solve the problem of data delivery in mobile and intermittently connected networks referred as Delay Tolerant Networks (DTNs). CGrAnt is a hybrid Swarm Intelligence-based forwarding protocol designed to deal with the dynamic and complex environment of DTNs resulting from users mobility or varying conditions of wireless communications. CGrAnt is based on (1) Cultural Algorithms (CA) and Ant Colony Optimization (ACO) and (2) metrics which characterize opportunistic social connectivity between wireless users. CA and ACO are used to direct the network traffic, taking into account a set of social-aware metrics that may infer relevant structures in meeting regularities and mobility patterns of users. The most promising message forwarders are selected through a greedy transition rule based on local and global information captured from the DTN environment. Through simulation, we have analyzed the influence of ACO operators and CA's knowledge on CGrAnt performance. We have then compared the performance of CGrAnt with PROPHET and Epidemic protocols under varying networking parameters. Results have shown that CGrAnt achieves the highest delivery ratio and lowest byte redundancy.

Opportunistic Content Dissemination

Here, we focus on dissemination of content for delay tolerant applications/services, (i.e. content sharing, advertisement propagation, etc.) where users are geographically clustered into communities. Due to emerging security and privacy related issues, majority of users are becoming more reluctant to interact with strangers and are only willing to share information/content with the users who are previously identified as friends. In this environment, opportunistic communication will not be effective due to the lack of known friends within the communication range. Thus, we have proposed a novel architecture that addresses the issues of lack of trust, timeliness of delivery, loss of user control, and privacy-aware distributed mobile social networking by combining the advantages of distributed decentralized storage and opportunistic communications. We have formally defined a content replication problem in mobile social networks and show that it is computationally hard to solve optimally. Then, we have proposed a community based greedy heuristic algorithm with novel dynamic centrality metrics to replicate content in well-selected users, to maximize the content dissemination with limited number of replication. Using both real world and synthetic traces, we have shown that content replication can attain a large coverage gain and reduce the content delivery latency.

Data Offloading-aware Hotspot Deployment

With the steady growth of sales of smart-phones, the demand for services that generate mobile data traffic has grown tremendously. The growing use of traffic data generated from mobile devices overloads the network infrastructure, which is not always prepared to receive such demand. To tackle this problem, we are studying the mobile behavior and resource consumptions of people on a metropolitan area in a major city and turn it into a set of well located WiFi hotspots. For this, we have proposed a data offloading-aware hotspot deployment. It is methodologically divided as (i) creation of a time dependent weighted graph to represent people's mobility, traffic and its relation with places/locations able to receive a hotspot, (ii) measurement of location's importance and selection of the best-ranked ones. Better positioned hotspots are likely to provide better coverage, and therefore, be able to offload more data.

Collaborations

• Professors Anelise Munaretto and Myriam Regattieri Delgado from Federal Technological University of Parana (UTFPR), Brazil,

• Professors Aruna Seneviratne and Henrik Petander from NICTA and School of EE&T, UNSW, Sydney, Australia,

• Pedro O.S. Vaz de Melo and Antonio A. F. Loureiro, Federal University of Minas Gerais, Brazil,

• Marco Fiore and Frederic Le Mouel from INSA Lyon, France,

• Katia Jaffrès-Runser, University of Toulouse, IRIT/ENSEEIHT, France.