백색 가우시안 잡음 통신채널에서 실시간 신호대잡음비 추정 방법의 오차해석 연구

Abstract

Underwater acoustic communication is an essential way for exchanging information among underwater vehicles and/or a control center at surface. As underwater vehicles move, the channel capacity varies continuously according to their positions. In this case a fixed modulation method gives low communication efficiency because it cannot use maximum channel capacity for all channel conditions. For this reason adaptive modulation was introduced to increase communication efficiency. Especially, power consumption is one of the important factors for long lifetime of battery-operated underwater systems. Therefore, proper selection of source level as well as the choice of modulation method is critical for operation of such underwater systems. The key parameter for this selection of a modulation method and source level is signal-to-noise ratio (SNR). Consequently, accurate SNR estimation is needed. This paper reviews real-time SNR estimation methods and estimates bias and random errors for various configuration so that a criterion of parameter range may be suggested for a given environment. As an example we found that 3dB estimation error bound for the ML estimator requires at least 128 symbols for SNR range of 0-30dB.