Context-aware Communication

Overview

Project A2 (Context-based Communications) deals with innovative context-aware communication paradigms that can be realized based on a detailed world model. The goal of this project is to utilize the world model in order to distribute messages selectively among users within a certain context.

Our current research is based on the results of the first funding period in the field of geographic communication protocols (geocast). Geocast protocols can be used to send messages to the users located in certain geographic areas. In the first funding period, we developed geocast protocols for infrastructure-based systems; in the second funding period we focus on geocast protocols for infrastructure-less networks and the integration of infrastructure-less and infrastructure-based geocast approaches. We consider infrastructure-less networks consisting of a set of nodes using wireless communication technologies to communicate directly with their neighboring nodes. These nodes form an ad-hoc network in order to send messages to distant nodes that cannot be reached directly. In particular, we consider so-called wireless mesh networks, which may span larger geographic areas like a whole city where both stationary and mobile nodes are involved in forwarding messages. For such infrastructure-less networks, we shall develop geocast routing protocols based on suitable location models. Moreover, we investigate how such infrastructure-less networks can be integrated into the geocast infrastructure developed in the first funding period in order to achieve a universal geocast service.

Besides geographic communication our research is focused on an enhanced context-aware communication mechanism called contextcast. Rather than just considering geographic information, contextcast may also utilize further context information from the Nexus world model to define the recipients of a message. Typical examples for such extended context information are user type, further user attributes, and the state or situation of a user. Contextcast communication has many interesting use cases. For instance, a message could be sent to a worker in the "Smart Factory" (Project D1) who has special skills and is located near a failed machine, or a warning message regarding high ozone concentration could be sent to older and ozone-sensitive people in a certain region. Besides information about the current user context, historic context information managed by the Nexus system can also be used for addressing. In order to realize an efficient contextcast service for large, potentially global systems, we shall develop suitable overlay network structures on top the IP infrastructure. In particular, we investigate how such an overlay network can be structured and adapted efficiently according to the dynamic context of recipients. Moreover, we also consider security and privacy aspects, in particular the protection of private user context information or the protection from receiving unsolicited messages ("spam"). These aspects are investigated in close cooperation with Project A3.

Institute

	Institute of Parallel and Distributed Systems
	Distributed Systems Department

People

	Project Head
		Prof. Dr. rer. nat. Dr. h. c. Kurt Rothermel
	Research Assistants
		Dipl.-Inf. Frank Dürr
		Dipl.-Inf. Lars Geiger