Research On Equalizer In Underwater Acoustic Communication System

Because of enabling to provide high-rate and high-quality wireless for mobile communication service, OFDM system is becoming a very hot research topic recently. OFDM technology has been used in many fields, such as digital audio broadcasting, digital video broadcasting, wireless local network systems, underwater acoustic communication systems and so on. Since it is well known that underwater acoustic (UWA) channel is an extremely complex channel with the characteristic of time-frequency-space variant, narrow available bandwidth, great transmission loss and strong multi-path effect, it is especially one of the technical difficulties to estimate and equalize this dispersive channel. In this thesis, we are working on the topic based on some of the precedents' results.At first, the characteristics of the underwater acoustic (UWA) channel is introduced, secondly, the self-accommodate equalization technique is introduced, including the equalizers' types, structures and arithmetic, then introduced the equalization of the OFDM communication system, time field equalization and frequency field equalization. In frequency field equalization, two pilot signals' structures are introduced, including block pilots and comb pilots; two kinds of channel estimator based on pilot signals are discussed, including Least Square (LS) estimator, Maximum Minimum Square Error (MMSE) estimator, in addition, four inserting-values arithmetic such as linear inserting-values arithmetic are introduced.This paper modulates the UWA OFDM communication system which based on different pilot signals' structures, different channel estimators and different inserting-values arithmetic. This paper also implements the experiments based on above plans in the pool and in the sea. The experiments' results indicate that from pilots' interval, the system's error rate when pilots' interval equal to four is lower than pilots' interval equal to five; from channel estimators, these two estimators affects the system's error rate almost the same; from inserting-values arithmetic, the advanced arithmetic such as spline inserting-values arithmetic is more applicable to the case of higher signal noise ratio, and the more simple arithmetic such as linear inserting-values arithmetic is more applicable to the case of lower signal noise ratio.