Indoor White Space Identification

Recent years have witnessed a surge in the number of mobile and wireless devices.To catch up with the demand for available radio frequency spectrtum,a number of approaches have been explored.Utilizing TV white spaces(e.g.,unused VHF and UHF spectrum bands)is one of the promising ways to fulfill the growing demand for available spectrum resources.The first and most important step of utilizing white spaces is to accurately recognize occupied and vacant channels.While detecting outdoor white spaces has been extensively explored in recent years,the indoor case is still open for investigation.Moreover,most existing works rely on an experimental propagation model.However,white spaces in buildings have different features from outdoor ones due to the complicated indoor environments.The propagation model does not take count for walls and obstacles that exist in buildings.Hence,simply applying the outdoor approaches to indoor white space identification is clearly not satisfactory.In this paper,we introduce WhiFind,which is an indoor white space identification system.It provides accurate indoor white space availability information with only a small number of spectrum scanners deployed.We first carry out intensive measurements and perform comprehensive analyses on indoor white space features,and reveal that indoor white spaces,inherently different from outdoor ones,have several interesting and important characteristics,such as low-rank structure,temporal stability,spatial and spectral correlation.Above observations facilitate us to design a more accurate and efficient white space identification system,namely WhiFind(for White Space Identification for Indoor Environments),which is a matrix completion based approach.We deploy our prototype on the third floor of our school building and conduct real-world experiments.Experiment results show that our design can identify 26.6% more white spaces with considerably high accuracy and low sensor cost compared with the best known existing approach.