Today we would like to present you another use case of drasyl: A colleague of ours developed a distributed data structure based on the SkipNet presented by Microsoft. In addition to that, he added capabilities to store multi-attributed data on a SkipNet.

Abstract

In distributed sensor networks, finding data sources is a particular challenge. Users and their provided services depend on the different collected data and on an efficient discovery of existing data sources.

Since sensor data is very ephemeral and privacy is also of great importance in a citizen-operated network, data cannot be stored multiple times in the network on different nodes as in conventional approaches. Rather, it must be possible to locate the exact data sources in the distributed network and query their data. For this purpose, we have developed an attribute-based approach on the distributed data structure “SkipNet”. This allows to describe data sources with different attribute-value pairs, which can be used to find them in an efficient time.

With the help of these attribute-value pairs and the search we have adapted, all sensors of a certain type, for example, are now grouped at a defined location and can thus be found efficiently.

However, the SkABNet approach is not limited to our use case. It is designed in such a way that any other information in distributed networks can be located. For this, only the attribute-value pairs have to be defined differently.

Further optimizations are planned for the future: For example, it should be possible for searchers to be continuously informed about new data sources to prevent repeated searches. The current development status can be found in his public Git repository.