Research On ICI Elimination And Channel Estimation Of OFDM System In High - Speed Mobile Environment

With the increase of the number of mobile terminals and the intelligence of cellphones, the requirement to communication and network grows in such a high speed which has never been seen before, which requires more and more efficiency and reliability in a communication system. Among all kinds of technology, OFDM has been selected as a core in several standards because of its robusty to frequency selective channel and high spectrum utilization. However, in high speed environment, the orthogonality of subcarriers will be broken due to the severe Doppler effect, which generates the intercarrier interference (ICI) and degrades the system performance significantly.This paper mainly discusses the pilot aided channel estimation used in high speed environment in OFDM. We focus on two channel model, Linear Time Varying (LTV) model and Basis Expansion Model (BEM), which are widely used in time-varying channel estimation. These models are analyzed and simulated using different pilot construction and estimation algorithm.We studied ICI cancellation algorithm in LTV model. Then dual ICI cancellation algorithm is introduced to deal with ICI on pilot subcarriers. The influence of ICI to pilot can be mitigated by using pre-cancellation method. However, pre-cancellation can not exactly eliminate ICI on pilot subcarriers. So we proposed an ICI cancellation algorithm based on improved group pilot scheme which can achieve better performance. The influence of ICI to pilot can be obviously reduced by comb pilot construction based on group pilot.We studied several BEM which based on different basis function including CE-BEM, P-BEM, DKL-BEM, DPS-BEM. Two estimation algorithms, LS and LMMSE, are adopted in BEM analysis and simulation. The result shows DPS-BEM outperform other mentioned BEM in high SNR environment while all BEM achieve similar performance in low SNR environment. So we took advantage of several basis functions and proposed mix BEM which can consider both performance and calculation complexity. The simulation result shows mix BEM can be used in such environment where channel condition changes frequently. It keeps high performance close to DPS-BEM while greatly reduces computation complexity.