DVB-T2 頻道估算之比較與研究

Abstract

Channel prediction provides up-to-date information about the state of the channel. This information describes how a signal propagates from the transmitter to the receiver and it represents the combined effects of scattering, fading and power decay. Digital Video Broadcasting - Second Generation Terrestrial (DVB -T2) utilizes Orthogonal Frequency Division Multiplexing (OFDM) to achieve higher capacity and better performances. OFDM provides an effective and low complexity means of eliminating inter-symbol interference, and high rate transmission is achieved by using higher-order constellation. Channel prediction increases the OFDM systems complexity. To reduce the OFDM system complexity, channel prediction for Channel Frequency Response (CFR) is performed at particular subcarriers, called the pilot symbol subcarriers. The pilot symbol subcarrier’s CFR is based on the Least Square (LS) estimator. Incorporating interpolation methods, helps to find the CFR of the remaining subcarriers, data symbol subcarriers. The interpolation methods includes the Linear Interpolation (LI) algorithm and the Cubic Convolution Interpolation (CCI) function with variable coefficients algorithm. This thesis implements and discusses the LS channel prediction method for OFDM of DVB -T2 and present the interpolation channel prediction scheme: One-Dimensional frequency interpolation and Two-Dimensional time-frequency channel prediction scheme. Both schemes utilizes the Linear Interpolation algorithm and the Cubic Convolution Interpolation function with variable coefficients algorithm. The ‘Typical Urban 6’ (TU6) and the Long Delay channel types are used to evaluate the performance of the one-dimensional and two-dimensional scheme. Computer simulations shows that in terms of the Mean Square Error (MSE), under the TU6 channel type: One-dimensional cubic convolution interpolation function shows better performance when compared to one-dimensional linear interpolation. The two-dimensional cubic convolution interpolation function also out-performs the two-dimensional linear interpolation. When comparing the one-dimensional and two-dimensional linear interpolation, the two-dimensional performs better but the one-dimensional cubic convolution out-performs the two-dimensional cubic convolution interpolation function. Simulations with the TU6 shows little variations in terms of performance when comparing any two different interpolation methods. The long delay channel type: both with the one-dimensional and two-dimensional, the cubic convolution interpolation function out-performs the linear interpolation algorithm and also two-dimensional scheme performs much better than the one-dimensional scheme for both cubic convolution interpolation and linear interpolation function.