查看文章

In Wireless Sensor Networks (WSNs) the operating conditions and/or user requirements are often desired to be evolvable, whether driven by changes of the monitored parameters or WSN properties of configuration, structure, communication capacities, node density, and energy among many others. While considering evolvability, delivering the required information with the specified quality (accuracy, timeliness, reliability etc) defined by the user constitutes a key objective of WSNs. Most existing research efforts handle fluctuations of operation conditions in order to deliver information with the highest possible specified quality. In this paper, we take these aspects into consideration and survey existing work on Quality of Information (QoI). As a contribution, we categorize WSN information into a set of abstract classes for generality across varied application types. Our survey shows that currently QoI is usually addressed in isolation by focusing on discrete data processing operations/building blocks such as raw data collection, in-network processing (compression, aggregation), information transport and sink operations for decision making. This survey comprehensively explains the different views of QoI on attributes, metrics and WSN functional operations mapped with existing approaches. The survey also forms the basis for specifying needed QoI research issues.