|dc.description.abstract||Wireless LAN products are gaining momentum in the market due to recent technological innovations in circuit devices and packaging, the availability of license-free frequency bands, and the new end-user needs for communication with portable computers.
In addition to providing secure communication, long battery life, and ease of management, the wireless LAN devices should be designed to operate in a robust
manner. The indoor environment is particularly hostile to radio frequency communication, because of mutipath fading, spatial loss (or distance loss) and shadowing of the receiver. A design of the RF transmitter and receiver should take the deleterious indoor channel into consideration to build a robust wireless LAN product .
The goal of this work was to design a robust transmitter and receiver for wireless
LAN products which operate in the indoor environment in the 2.4GHz ISM
frequency band. The thesis consisted of two phases. The first phase focused on
the study of the wireless LAN products to understand the application and then the
indoor environment and analyzed the performance of the optimal receiver. This sets
up the background for the design requirements of a robust transceiver design.
The second phase explored the spatial and vertical polarization techniques to
achieve diversity reception. The key contribution of the thesis is in the proposal of a
neural network based diversity combiner. Simulations studies were conducted which
showed that a multilayer feedforward neural network performs close to the optimal
receiver for a diversity combiner.||en_US