An important aspect of wireless networks is their dynamic behavior. The conventional protocol stack is inflexible as various protocol layers communicate in a strict manner. In such a case the layers are designed to operate under the worst conditions as opposed to adapting to changing conditions. This leads to inefficient use of spectrum and energy. Adaptation represents the ability of network protocols and applications to observe and respond to channel conditions.

Traditional simulation methodologies independently model the physical and higher layers. When multiple layer simulations are required, an abstraction of one layer is inserted into the other to provide the multiple layer simulation. However, recent advances in wireless communication technologies, such as adaptive modulation and adaptive antenna algorithms, demand a cross layer perspective to this problem in order to provide a sufficient level of fidelity. However, a full simulation of both layers often results in excessively burdensome simulation run-times. The benefits and possible parametric characterization issues arising due to the cross-layer integration of lower physical and higher network layers are investigated in this thesis. The primary objective of investigating cross-layer simulation techniques is to increase the fidelity of cross-layer network simulations while minimizing the simulation runtime penalties.

As a study of cross-layer system design a medium access control (MAC) scheme is studied for a MANET wherein the nodes are equipped with smart antennas. Traditional MAC protocols assume the use of omnidirectional antennas. Nodes with directional antennas are capable of transmitting in certain directions only and significantly reduce the chances of collision and increase the effective network capacity. MANETs using omni-directional antennas severely limit system performance as the entire space around a node up to its radio range is seen as a single logical channel. In this research a MAC protocol is studied that exploits space division multiple access at the physical layer. This is a strong example where physical and MAC design must be carried out simultaneously for adequate system performance.

Power control is a very important in the design of cellular CDMA systems which suffer from the near-far problem. Finally, the interaction between successive interference cancellation (SIC) receivers at the physical layer and power control, which is a layer 2 radio resource management issue, is studied. Traffic for future wireless networks is expected to be a mix of real-time traffic such as voice, multimedia teleconferencing, and games and data traffic such as web browsing, messaging, etc. All these applications will require very diverse quality of service guarantees. A power control algorithm is studied, which drives the average received powers to those required, based on the QoS requirements of the individual users for a cellular CDMA system using SIC receivers.