Research

Opportunistic networks (please see iClouds), formed by personal user devices, exploit private resources of users, mainly battery power and device memory, thus there arises the question, why should a user take part in the system at all?

An obvious benefit for the user is the potential fulfillment of his information needs. The device collects only information the user is interested in. In return, this information is shared with others.

In addition, the opportunistic network model can be extended with an incentive scheme to stimulate and thus increase user participation. The adPASS incentive scheme allows users to gain some kind of benefit by passing information on to other users. The core idea is presented next.

The incentive scheme rewards users who partly help to carry information from an information producer to an information consumer. Figure 1 illustrates this. A producer (blue node) passes information to nodes (green) in communication range. The information is passed on only if an entry on the iWish list matches the information (see iClouds project for iWish/iHave). In the example, the information is passed on from node to node until it reaches a consumer that uses the information for his benefit (red node). In return, all nodes framed in the yellow box should be rewarded by the producer or the consumer since they helped in bringing them together. The set of nodes framed in the yellow box is called bearer chain.

Figure 1: Incentive Scheme - Basic Idea

The incentive scheme distinguishes different roles of opportunistic network participants.

• An information producer is a node that creates information and initiates its dissemination
• An information bearer receives information from a producer or another bearer and passes it on to other bearers or consumers. A bearer is only interested in gaining a reward for transporting information
• An information consumer receives information from a producer or bearer. He takes some action on it that is beneficial for himself and the producer. This leads towards dispensing a reward to all bearers that took part in carrying information down the producer-to-consumer path. A node can act both as, a bearer and a consumer.

Since the incentive scheme is based on the opportunistic network model and nodes act and move autonomously, there must be a way to issue a reward to information bearers. A mediator keeps track of the users’ rewards. It works similar to a central database where the producer, the bearer, and the consumer have access to (for example via the Internet). Therefore, it guarantees accessibility to the inherently transient mobile nodes in the network. In addition, the mediator helps to keep nodes anonymous, if they wish so.

The incentive scheme conflicts with the privacy protection of a user. Consider an arbitrary information i. A user Alice who wants to claim a reward for being part of the bearer chain, thereby helping that i finds its way from the producer to the consumer, has to prove her participation in the bearer chain. This implies that Alice was interested in i and has shown this interest with a matching entry on her iWish-list. Therefore, the producer and the consumer might be able to learn, that a user Alice was interested in i. But, Alice should be able to prove her legitimate reward claim without revealing her identity. The adPASS incentive scheme makes use of public/private-key cryptography operations to support this in the following way:

• A node generates a public/private-key pair on the device as a user pseudonym for each bearer chain.
• The public-key is passed with the information i to flag the nodes participation.
• The private-key is used to sign the bearer chain in order to prevent fraud and prove the participation in the bearer chain.

A detailed elaboration on the usage of public/private-key pairs as part of the incentive scheme is given in An Anonymous Bonus Point System For Mobile Commerce Based On Word-Of-Mouth Recommendation, T. Straub, A. Heinemann. Proceedings of the 19th Annual ACM Symposium on Applied Computing (ACM SAC 2004). PDF file available here.

For more information, visit the project web site at iclouds.tk.informatik.tu-darmstadt.de

Scientific Advisor

• Prof. Dr. Max Mühlhäuser

Former Project Member

• Prof. Dr. Andreas Heinemann

If you are interested in our work, don't hesitate to  contact us.

Address
Darmstadt University of Technology
Department of Computer Science, Telecooperation
Hochschulstr. 10
D-64289 Darmstadt, Germany