Memory effects in microwave devices

This research proposal aims at the realization of a new measurement method for the quantification and manipulation of memory effects in microwave components. Furthermore, we want to prove the validity of the new method by measurements. Memory effects in microwave amplifiers lead to intersymbol interference in modern communication systems and, therefore, significantly deteriorate their performance. They also affect the effectiveness of common amplifier linearisation schemes. Usually memory effects are characterized indirectly, i.e. in the frequency domain by imbalances in the upper and lower intermodulation products. This method, however, is not suitable for the direct quantification of the adverse effects induced in information transmission systems. This research project will create a new solution to this problem. The evaluation of linear and nonlinear memory effects in the time domain will open the possibility to measure and to quantify their influence on time discrete information transmission systems. Furthermore, the dependence of the memory effects on the externally applied terminations and bias conditions will be investigated. For this purpose, we want to integrate an active baseband tuner into an existing RF source- and load-pull system and combine it with a broadband sampling system operating in the baseband. The extended source- and load-pull system provides the possibility to create any source and load impedance at the RF- and envelope frequency range seen by the nonlinear, memory-loaded device under test. The second system allows the creation and the recording of RF signals by their equivalent baseband representation. In this project we want first to establish the mathematical tools for the identification and description of the memory effects. In parallel to this activity we will implement the complete measurement setup including the necessary control programs. Then we will measure and quantify the memory effects of known microwave transistors in order to verify the validity of our method. In the following we want to investigate the influence of source- and load impedance on the memory effects. If we can observe a significant influence, this will lead to a new approach in microwave power amplifier design and consequently to the possibility of a commercial exploitation of the new measurement method. We apply for funding of 36 person months within 24 months duration of the project. The emphasis of this project rests on the development of the mathematical tools to describe the memory effects, on the implementation of the measurement system, and on the verification of the new method.