This work is concerned with the application of digital filtering methods to the problem of simulation of communications and signal processing systems. The requirements of a generalised simulation scheme are discussed in the light of the extensive literature on digital simulation. Digital filtering methods are compared with numerical integration methods at a conceptual level. The hypothesis is then developed that, for simulation of the type of system under consideration, model design should be based upon accurate simulation of frequency response. The main body of the work then concentrates upon the design of digital filters which accurately model the frequency responses of continuous filters. To this end, digital filter transformation design methods are investigated and methods are developed for simulation-filter design. Digital filtering methods are also shown to be useful for the simulation of function generation elements by providing alias error correction. Non-linear and modulation system elements are discussed, but unfortunately the scope of the work is not sufficient to cover these aspects in great detail. In addition to the mainstream of the work an interesting side issue arose during the investigation of a filter design method, resulting in the development of a practical computer algorithm for system identification via the Weiner-Lee transform. This type of work requires extensive use of digital computer aids. Software development is time-consuming and the written work cannot reflect the effort so involved. For this reason an appendix has been included which describes the more important elements of the developed software.
|Date of Award||1976|