Hwang Jun Gyu, Kyungpook National University; Cui Shuyu, Kyungpook National University; Lee Chang-ho, Mobile Testing & Compliance Certification Lab.; Park Joon Goo, Kyungpook National University

View Abstract Sign in for premium content

Abstract:

Although LBS(Location Based Service) is provided in various ways, higher positioning performance is required in a complex environment. Therefore, research on various positioning techniques is being conducted. This paper proposes an indoor positioning technique using Super Resolution Generative Adversarial Networks (SRGAN). We make image based RSSI(Received Signal Strength Indicator) and AP(Access Point) information for SRGAN. And then learning was executed to induce more accurate positioning performance. Through this, efficient indoor positioning can be performed by generating a Fingerprint-Map even with insufficient sample data.