Collaboration in Opportunistic Networks
Motivation. With the increasing integration of wireless short-range communication technologies (Bluetooth, 802.11b WiFi) into mobile devices, novel applications for spontaneous communication, interaction and collaboration are possible. We distinguish between active and passive collaboration. The dev...
Main Author: | |
---|---|
Format: | Others |
Language: | English en |
Published: |
2007
|
Online Access: | http://tuprints.ulb.tu-darmstadt.de/834/1/heinemann07-diss.pdf Heinemann, Andreas <http://tuprints.ulb.tu-darmstadt.de/view/person/Heinemann=3AAndreas=3A=3A.html> : Collaboration in Opportunistic Networks. [Online-Edition] Technische Universität, Darmstadt [Ph.D. Thesis], (2007) |
Summary: | Motivation. With the increasing integration of wireless short-range communication technologies (Bluetooth, 802.11b WiFi) into mobile devices, novel applications for spontaneous communication, interaction and collaboration are possible. We distinguish between active and passive collaboration. The devices help users become aware of each other and stimulate face-to-face conversation (active collaboration). Also, autonomous device communication for sharing information without user interaction is possible, i.e., devices pass information to other devices in their vicinity (passive collaboration). Both, active and passive collaboration requires a user to specify what kind of information he offers and what kind of information he is interested in. Object of Research: Opportunistic Networks. Spontaneous communication of mobile devices leads to so-called opportunistic networks, a new and promising evolution in mobile ad-hoc networking. They are formed by mobile devices which communicate with each other while users are in close proximity. There are two prominent characteristics present in opportunistic networks: 1) A user provides his personal device as a network node. 2) Users are a priori unknown to each other. Objectives. Due to the fact that a user dedicates his personal device as a node to the opportunistic network and interacts with other users unknown to him, collaboration raises questions concerning two important human aspects: user privacy and incentives. The users’ privacy is at risk, since passive collaboration applications may expose personal information about a user. Furthermore, some form of incentive is needed to encourage a user to share his personal device resources with others. Both issues, user privacy and incentives, need to be taken into account in order to increase the user acceptability of opportunistic network applications. These aspects have not been addressed together with the technical tasks in prior opportunistic network research. Scientific Contribution and Evaluation. This thesis investigates opportunistic networks in their entirety, i.e., our technical design decisions are appropriate for user privacy preservation and incentive schemes. In summary, the proposed concepts comprise system components, a node architecture, a system model and a simple one-hop communication paradigm for opportunistic network applications. One focus of this work is a profile-based data dissemination mechanism. A formal model for this mechanism will be presented. On top of that, we show how to preserve the privacy of a user by avoiding static and thus linkable data and an incentive scheme that is suitable for opportunistic network applications. The evaluation of this work is twofold. We implemented two prototypes on off-the-shelf hardware to show the technical feasibility of our opportunistic network concepts. Also, the prototypes were used to carry out a number of runtime measurements. Then, we developed a novel two-step simulation method for opportunistic data dissemination. The simulation combines real world user traces with artificial user mobility models, in order to model user movements more realistically. We investigate our opportunistic data dissemination process under various settings, including different communication ranges and user behavior patterns. Our results depict, within the limits of our model and assumptions, a good performance of the data dissemination process. |
---|