In this diploma thesis, I developed algorithms to improve the throughput of data networks based on single-input single-output (SISO) and multiple-input multiple-output (MIMO) wireless links. The aim is to use the spectrum more efficiently and flexible, while providing the required reliability of the transmission and adaptation of network behavior. These techniques are based on theoretically optimal solutions maximizing the throughput and are too demanding to be used directly in the practice. Therefore, some simplifications are needed and I evaluated their impacts on the network performance.
The optimal solutions are originally designed for uncoded systems. To overcome this shortcoming in more practical applications, I created a framework for coded systems with bit error ratio and packet error ratio adaptation. In this framework, the type of coding can be easily replaced once the theoretical formulas for the new modulation and coding scheme are known. Similarly, the error distribution modeling used in the packet error ratio evaluation can be further improved. The performance of this approach is evaluated in a WiMAX simulator and compared to the achievable channel capacity as a reference.