Research On Countermeasures Against Multi-Path Fading In WIMAX

With the development of internet and multimedia applications, there is an increasing need for large scale deployed broadband wireless access system which can support high-speed data transmission. To satisfy this need, IEEE developed and published 802.16 standard, often referred to as WiMAX. In the IEEE802.16-2004 standard, four physical layer interfaces have been specified: WirelessMAN-SC, WirelessMAN-SCa, WirelessMAN-OFDM and WirelessMAN-OFDMA. WirelessMAN-OFDM is a physical layer scheme for 2GHz-11GHz non-line-of-sight transmission. In the non-line-of-sight environment, the main challenge is how to counteract the degeneration caused by multi-path propagation. Diversity is an effective method to solve this problem. In a WirelessMAN-OFDM system, frequency domain diversity can be exploited by combining the OFDM multi-carrier technology, interleaving and channel coding. Moreover, one might use the optional Alamouti space-time coding in the downlink, by which the space domain transmit diversity can also be exploited. When using space-time coding, our concern is on how to obtain the multiple channels'state information. For this reason, This thesis is mainly focused on channel estimation algorithm in Alamouti space-time coded WirelessMAN-OFDM system. Firstly, according to the specifications of STC preamble and STC comb-type pilot in the standard, By utilizing the frequency orthogonality of STC preamble and the code orthogonality of STC comb-type pilot, multi-channel estimation in space-time coded WirelessMAN-OFDM system can be decoupled into several channel estimations just as in a SISO(single input single output) enviromnet.After carefully analyzing the algorithms in SISO, a near-complete channel estimation scheme based on WirelessMAN-OFDM specification is introduced. Both STC preamble assisted initial channel estimation and STC comb-type pilot assisted channel tracking are included in this scheme. 1) initial estimation: The frequency-domain LS algorithm, LMMSE-SVD algorithm and time-domain LS algorithm are compared in this thesis. Simulation results show that the frequency-domain LS algorithm is simple but the performance is poor, LMMSE-SVD algorithm is modestly complex but has better performance, time-domain LS algorithm has the best performance but is more complicated and requires high accuracy of timing recovery. 2)channel tracking: When the maximum time delay of channel impulse response is low, eight STC comb-type pilots are enough to be used to estimate the channel. The comb-type pilot estimation is composed of pilot-tone channel estimation and interpolation. Two interpolation methods—time domain interpolation and transform domain interpolation, both of which can avoid the approximation error of direct frequency interpolation are discussed with emphasis. Analysis of ideal conditions shows that time domain interpolation is more suitable for practical systems. In addition, The performance of time domain interpolation in strong white noise background can be further enhanced by combining the adaptive line enhancer tech.