Sound waves are highly conductive in the ocean; therefore, they are used in underwater detection and signal transmission. During these applications, we may receive some signals, such as radiated noise, self-noise, reverberation noise, target noise, and ambient noise. These signals are influential in the application of underwater acoustics, so many people study underwater noise and create numerous mathematical and physical models in order to improve the applications. According to the past researches, ambient noise is the most complicated one among all the other underwater noise. Until now, there are still some unknown factors in the ambient noise study, many of which are too intricate to be modulated. Accordingly, it is very important to know more about the ambient noise for the enhancing of the quality of underwater communication and detection, which is the reason why people keep researching on ambient noise. Among all kinds of ambient noise, wind-generated noise is not only loud in volume, but also wide in frequency scale, which makes it the most detectable noise during sound detection. In order to understand wind-generated noise, we have compiled past papers first to look for the correlation between physical mechanisms and models of mathematics. In addition, we have constructed a measuring system for underwater sound in the ocean and another measuring system for wind on land. Data from the two measuring system were analyzed by statistics. One of the correlations between wind and ambient noise is that the stronger wind appears, the louder ambient noise is generated. But wind does not always generate ambient noise; it has to be strong enough to generate ambient noise. However, even when the wind is strong enough, ambient noise is not generated immediately, but with a delay. By analyzing the data, we have not only proved the credibility of the data from the old researches, but also presented a method of statistics for analyzing the wind delay effect, and the minimum velocities of wind which generates ambient noise.

Owing to the need of high-directivity radiation in fix-point communication, this thesis designs the metamaterial with 4-fold rotational symmetric unitcells. Using their nearing-zero refractive index along two polarizations, we can get gain-enhanced antenna radomes and place them above dual- and circular-polarized antenna to verify our idea. It is known that the group delay variation in a communication system causes the degradation of symbol error rate, and positive group delay (GD) causes delay in signal propagation. Therefore, this research studies the negative group delay (NGD) behavior of metamaterial. Through analyzing the GD of metamaterial equivalent circuit caused by poles and zeros, we design a novel NGD bandpass circuit. Furthermore, this thesis develops graphic method for group delay investigatiion and study the relationship between NGD, loss and gain in order to simplify the design of GD compensation circuit.