Table 6 Comparison between radiator and the surveyed systems
From: Radiator - efficient message propagation in context-aware systems
System | Main feature | How it compares to radiator |
|---|---|---|
Context Toolkit[2] | Inference leads to semantic aggregation | Radiator performs semantic aggregation both at the inference layer and at the distribution layer |
Syntactic aggregation at the distribution layer | Radiator is able to perform both syntactic and semantic aggregation at the distribution layer | |
Content-Based Pub-Sub | Matching is made at syntactic level or at limited semantic level (keywords with the same meaning) | In Radiator, the matching algorithm is more flexible because it can be defined as a function provided by the developer. |
Fulcrum[37] (Pub-Sub) | Allows developers to provide matching functions | Radiator also allows events propagation to be delayed. |
Berkovsky[38] (Pub-Sub) | Automatic subscription based on user context | Radiator adds to the automatic subscription the possibility of aggregating and delaying messages, improving scalability. |
Multi-cast trees | Message propagation using P2P overlays | Radiator does not incur the cost of rebuilding P2P overlays every time the context changes. |
DSPs | Efficient push-based propagation of messages | Radiator is able to extract semantic meaning from messages and infer the current context of its users, therefore adapting its propagation characteristics more efficiently. |