Quality of Information for Semantically-Adaptive Networks (QoI-SAN)

The QoI-SAN thrust seeks to develop a foundational science to model, characterize, predict and control information delivered through multi-genre networks based on the semantics and context of information requests, and requisite composite quality-of-information measures (both intrinsic and contextual).  Our premise is that the overall information capacity of networks will be maximized by taking into account the context (semantics) in which the requested information is used, by using prior knowledge to drive information requirements or derive additional information, and by exploiting the tradeoffs between data representation, the desired information quality, and communication capabilities.  In the final year we will address semantic information theory (task Q1), distributed QoI-aware video analytics (task Q4) and anticipatory sensing for information acquisition (task Q5).

QoI-SAN has three tasks focusing on two main topics.

  • Topic 1: A Unified Framework for Semantically Aware Networks
    • Task Q1: Semantic Information Theory
  • Topic 2: Characterizing and Controlling the Trade-offs of QoI in Semantic Information Delivery
    • Task Q4: Quality Aware Video Analytics for Complex Tactical Activities
    • Task Q5: Workflow-assisted Anticipatory QoI Optimization