K-Nearest Neighbors Gaussian Process Regression for Urban Radio Map Reconstruction
Yifang Zhang, Shaowei Wang
Abstract
Radio map is of great importance to interference control, network planning and resource allocation in wireless communications. In this letter, we develop an accurate radio map reconstruction method based on k-nearest neighbors Gaussian process regression, which can exploit the relationship of the received signal strengths at adjacent locations efficiently. Numerical experiments demonstrate that our proposed method outperforms the state-of-the-art ones significantly.
Topics & Concepts
Computer scienceInterference (communication)Gaussian processExploitKrigingGaussianRegressionk-nearest neighbors algorithmRadio resource managementWireless networkArtificial intelligenceAlgorithmWirelessData miningComputer networkTelecommunicationsMathematicsMachine learningStatisticsChannel (broadcasting)Computer securityPhysicsQuantum mechanicsIndoor and Outdoor Localization TechnologiesMillimeter-Wave Propagation and ModelingDistributed Sensor Networks and Detection Algorithms