Binary Classification of Human-body Communication Channels Based on Gain and Phase Information (Toward Applications to Smart Identification Systems)

IR Web Site

Bibliographic Information

Other Title
  • 利得・位相情報を利用した人体通信チャネルの2クラス分類～スマート認証システムへの応用に向けて～
  • リトク ・ イソウ ジョウホウ オ リヨウ シタ ジンタイ ツウシン チャネル ノ 2 クラス ブンルイ : スマート ニンショウ システム エ ノ オウヨウ ニ ムケテ

Search this article

Description

Human body communication (HBC) is expected as a technology that can realize smart personal identification systems where users do not have to touch with IC cards. The most severe problem of the HBC technology is that an outsider walking in the vicinity of the identification systems is often misidentified. A feasible approach to this problem is to classify signals sent from the user and outsider by a receiver located in the system. In our previous study, it was demonstrated that the transfer function of the HBC channel H(f) can be utilized as feature vectors for classifying the received signals. However, the previous study was based only on its gain information |H(f)|. Therefore, it is expected that the classification accuracy will be further improved by using both the gain and phase information, i.e., |H(f)| and arg{H(f)}. To validate this expectation, we investigated the classification accuracy obtained with feature vectors composed of the gain and phase information. As expected, it was demonstrated that the classification accuracy is certainly improved by utilizing both |H(f)| and arg{H(f)}.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top