Research On Optimal Pilot Sequence Design In OFDM System

Orthogonal frequency division multiplexing (OFDM) technology with high spectrum efficiency, and flexible bandwidth expansion can resist the channel multipath fading,to be a new generation of wireless communications research focus. Communication quality assurance of next-generation wireless communication system is mainly dependent on the insert of the density and location of the pilot signal in the data stream. Different types of business and system performance requirements on the different density and location of the pilot. The study of the pilot sequence design can effectively reduce system losses, improve the spectrum utilization, and therefore of great significance in the study of a new generation of mobile communications.This paper systematically analyzes the basic principle of OFDM system and the radio channel transmission characteristics, carried out a detailed introduction to the classic algorithm for OFDM system pilot sequence design. Including the distribution based on the uniformity of the minimum MSE criteria to minimize the pilot sequence, non-uniform distribution of the maximizing average SNR criterion, non-uniform distribution under the criteria of minimum SER criterion,and so on.This paper in-depth study of the classical algorithm for optimal pilot sequence design of a conventional OFDM system. Focus on the data utilization, channel state informationd and the computational complexity of the optimal pilot,we propose some Effectively and easily pilot sequence design algorithms.The traditional uniform distribution of the pilot sequence only in the ideal environment which consider only the Gaussian noise is the optimal pilot sequence. However, in the actual system transfer process, there are still the impact of multipath fading and other factors. So the fixed pilot sequence mode will cause the entire transfer process can not access to high channel estimation accuracy. In addition, the pilot sequence design, the classic algorithm is mostly for the pilot number or location of a single property of the pilot study, during the optimization process can not maximize the satisfaction of the BER or the MSE. To address the problem,we propose a kind of channel state information feedback of the pilot sequence design algorithm, this algorithm through the use of the previous time on a channel is estimated to obtain the channel state information, design a number of the next time pilot number and location. Then the optimal pilot sequence feed back to the sender, insert for the next time pilot, thus effectively reducing the system BER and MSE.Focus on the high computational complexity of the maximum average SNR criteria optimum pilot location, when high number of subcarriers not be easily achieved, this paper propose the pilot sequence design algorithm based on the judgment of the channel gain. This algorithm is in progress before the next time pilot position design of the channel gain judgment, get the value of each subcarrier channel gain. Then use a pilot distribution pattern which a lower channel gain distribution and uniform distribution combing. On the estimation accuracy, this proposed algorithm close to the best one channel state information feedback of the pilot sequence design algorithm. But this proposed avoid the brute-force method for solving linear programming problems and greatly reduce the computational complexity of the system and thus improve overall system performance.