Researches Of Channel Estimation In IEEE802.11n MIMO-OFDM Systems

With the fast progress of wireless LAN (Local Area Network), IEEE802.11 has become the main flow standard in WLAN. Recently, 802.11n work group has been established to constitute a new high-speed wireless LAN standard, namely 802.11n, which has great break through in throughput and will be the standard of next generation WLAN without doubt. 802.11n could provide higher throughput, larger range and more reliability for application that is the most sensitive to bandwidth. Combined with many kinds of advanced telecommunication technologies, including spatial multiplexing MIMO (Multi-In Multi-Out), 20/40M bandwidth and dual frequency point (2.4 GHz and 5 GHz), 802.11n could support very high throughput and is compatible with previous IEEE 802.11b/g.Due to the indefinite standardization of channel estimation algorithm in 802.11n, therefore, how to design a perfect training sequence to identify different antennas and choose a proper channel estimation method to estimation the channel information accurately is the main research orientation of this thesis. This thesis will introduce the technique characteristics of MIMO-OFDM and the whole system model firstly. Then the wireless fading channel model and its characteristics will be analyzed. After these works, the channel estimation algorithms based on OFDM system are reviewed, including LS (Least Square) channel estimation algorithm and MMSE (Minimum Mean Square Error) channel estimation algorithm. This thesis analyses the channel estimation implementation scheme in IEEE 802.11a in detail, such as design of test sequence, pilots, and algorithms. This thesis focuses on the research of channel estimation in MIMO-OFDM system, including current research status, algorithm systematic architecture and variance definition. The derivation of uniform time-frequency domain result using LS channel estimation and design of interleaving preamble and superposition preamble is emphasized. At last, an improved interleaving preamble using PS code is proposed, which is based on the channel model mentioned before. A traditional LS channel estimation algorithm is simulated on 802.11n system. The simulation result is compared with the algorithm proposed by this thesis, which shows that.The improved algorithm has much better performance. Besides, a novel channel estimation algorithm for 802.11n is proposed, which improves the performance of channel estimation greatly. It does not estimate the length of CIR, but the length of minimum MSE in situation of real time noise. More simulation and comparison using different channel estimation schemes are given at the end of this thesis.All in all, the thesis focuses on channel estimation algorithms based on interleaving preamble and frequency LS method. In practical application, 802.11n employs training sequence to make channel estimation, therefore, we mainly research design of preamble by considering physical channel characteristics and frame structure of realistic system, and look for the optimum channel estimation scheme for 802.11n.