Abstract
Abstract Wireless local area networks (wireless LANs) are flexible data communication systems, which make us access the internet more unconstraint. Using radio frequency (RF) waves, wireless LANs transmit and receive data over the open space, minimizing the need for arranging wired cabling. With wireless LANs, users can acquire shared information without looking for a place to plug in, and network managers can set up or arrange networks without installing or moving wires. Up to present, many institutes have created their standards of wireless LANs and one of the most popular is standardized by the IEEE. IEEE Societies and the Standards Coordinating Committees of the IEEE standards Association (IEEE-SA) Standards Board had developed a standard 802.11a for one of the implementations of wireless LANs. The 802.11a system is an OFDM-based wireless LAN, which can provide a data rate up to 54 Mbps in 5 GHz bands. When implementing the 802.11a wireless LAN, one of the important components is the estimation of wireless channels. There exist many effects (multipath fading, time delay spread, and Doppler spread) in a wireless transmission environment where the channel is worse than a wired transmission environment. In order to achieve a higher data transmission rate within an acceptable bit error rate, a good channel estimator is necessary. In this thesis, we propose two practically achievable channel estimation algorithms for the 802.11a wireless LAN. One is called the average-in-frequency-domain (AIF) estimator, which uses the average concept in the frequency domain. The other is called the decision-aided-average-in-time-domain (DA-AIT) estimator, which uses the detected signals as “virtual” pilots and averages more individual pilot-estimated channel responses in the time domain. As compared to a widely used channel estimation method for the 802.11a wireless LAN, the proposed approaches have better performance in both additive white Gaussian noise channels and slowly time-variant channels with no significant frequency selective effects. However, each of these two proposed methods has its own benefits and limitations. The AIF estimator has less complexity but is not suitable for frequency selective channels. In contrast, the DA-AIT estimator is applicable to frequency selective channels, but it needs more computational complexity.