Page 462 - Maxwell House
P. 462
442 Chapter 9
interfaces, and converting working codes to C++. … A computer can ingest an endless supply
of hours while generating nothing but PowerPoint presentations. … If your simulation brings
forth an amazing potentially-publishable result, look for a numerical instability.”
9.1 BASICS OF COMPUTER DESIGN
9.1.1 Design, Analyze, Build
Evidently, the first step in any simulation is the model development of some physical object
(antenna, filter, circulator, etc.)
that comprises several sub-
models: theoretical,
conceptual, mathematical,
computer imaging, statistical,
etc. The saddest moment that
any of such sub-models as well
as the final model are not the
real world entity but just a set
of computer images,
mathematical symbols and
linking all of them equations
that help us better understand
surrounding world. Evidently,
the surrounding world is so
diverse and complex that the
procedure of the model
construction that is entirely
adequate to a real object is
extremely challenging or just
Figure 9.1.2 Model development and test verification unattainable. As a result, the
thorny path to well-posed
models typically consists of several iterations, as the block diagram in Figure 9.1.2 illustrates.
An obvious point to start is the analysis of the system-level technical specification describing
RF characteristics and performance of the system to be developed. We need to translate it
directly into technical specs of different system sub-models. Figure 9.1.3a (see below) is the
5
exemplary system-level specification of a dual-frequency, dual-polarized, Doppler weather
radar. Suppose our engineering objective is to design a parabolic reflector for this system. We
confine ourselves to several main antenna features omitting the various and less critical on this
stage details.
5 Public Domain Image, source:
https://pmm.nasa.gov/sites/default/files/document_files/D3R_Technical_Specs.pdf