저전력 수중음향통신을 위한 선형등화기의 최적화에 대한 연구

Abstract

A communication system with low power consumption is required in underwater sensor network because of environmental limitations. For the low-power system, it is necessary to reduce the transmit power, and to optimize the sparse coefficients of equalizer, or the tap length. In this thesis, the method that reduce a computation time by optimizing computation process is proposed to realize low-power underwater acoustic communication system. At first, dependency of decision delay on tap length of linear equalizer was investigated using simulation. According to the result, increasing the tap length resulted in an improvement of MSE performance within a limited range even though tap made system became complicated. Moreover, excessive tap length induced over-estimation effect, thus it make error increase, and system showed the best performance when decision delay of signal sequence had appropriate value. In addition to decide optimal tap length with decision delay, we extracted the MSE-decision delay graph. From the graph, we obtained variance value of the MSE-decision delay, and estimated the optimum decision delay range from the variance value. Also, using the extracted optimal parameters, we performed a simulation. In this simulation, we considered two cases of depth of receiver, 130 and 170m. According to the result, the simulation employing optimal tap length, which is only 33% of maximum tap length, showed a satisfactory performance comparable to simulation employing maximum tap length.