Software Engineering

175
10.1 
Motivation for the functional architecture
The functional analysis and allocation practice provides an approach for trans-
forming software requirements into the functional transactions that the software 
product must enable. The functional architecture represents a comprehensive, inte-
grated set of data processing transactions. Functional analysis and allocation is 
employed to achieve the following six fundamental design challenges confronting 
software product quality.
1. 
Elaborating a solution.
 The design solution involves many individual software 
elements that must work together to support the business or operational process. 
This encompasses numerous data processing transactions, services, or threads 
of behavior. Software data processing transactions involve many analytical com-
binations and permutations that are difficult to comprehend. Functional analy-
sis provides the means for identifying the myriad of possible data processing 
sequences the software must facilitate. The functional analysis and allocation 
practice systematically identifies and investigates discrete transactional threads 
to completely specify a functional solution.
2. 
Clarifying ambiguity.
 Language, which is used to communicate and express 
requirements, involves imprecise, vague, unclear, ill-defined, inexact, nebulous 
words and expressions. Functional analysis ensures that every function is clearly 
specified so that no misunderstanding exists in the description of the functional 
solution.
3. 
Resolving assumptions.
 All assumptions must be resolved with stakeholders 
before a solution can be finalized. Assumptions that are not challenged and 
resolved may result in a software product that does not effectively satisfy cus-
tomer needs and expectations. Functional analysis can be used to speculate 
about and evaluate assumptions, thereby eliminating judgment or opinions as 
the basis for product design. Functional analysis accentuates the existence of 
incomplete or inconclusive information to draw attention to suppositions that 
are not substantiated.
4. 
Achieving performance objectives.
 Specified performance requirements or objec-
tives must be comprehended and the software product designed to achieve these 
criteria. Software design techniques, drawings, and models must be utilized that 
exhibit software performance characteristics. Performance measures at the soft-
ware product level must be allocated to provide lower-level design aspirations. 
Initial performance budgets must be established against which design strategies 
can be assessed. Performance requirements can then be allocated and specified 
for the design elements once the solution has been appraised to be suitable.
5. 
Determining resource utilization.
 Data processing efficiency and effectiveness 
depends on the regulation of computing resource utilization. The functional 
architecture must model resource utilization to permit the design to be opti-
mized for efficient and effective execution. The software design must be sensi-
tive to the impact of resource utilization on performance objectives.
6. 
Simplifying the solution.
 Complexity is never a desirable characteristic of any 
software product. The user interface design and user interactions must not be 

---