home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.umcs.maine.edu
/
2015-02-07.ftp.umcs.maine.edu.tar
/
ftp.umcs.maine.edu
/
pub
/
WISR
/
wisr6
/
proceedings
/
ascii
/
smith.ascii
< prev
next >
Wrap
Text File
|
1993-10-19
|
12KB
|
318 lines
Reuse Insertion in Small Embedded Systems
Larry Smith
United States AirForce
Software Technology Support Center
OO-ALC/TISEB
7278 4th Street
Hill Air Force Base, UT 84056-5205
Tel: (801) 777-9730
FAX: (801) 777-8069
Email: smithl@stsc.hill.af.mil
Vern Phipps
SofTech, Inc
Layton, UT 84056
Tel: (801) 544-9200
FAX: (801) 777-8069
Email: phipps@stsc.hill.af.mil
Abstract
The Department of Defense has many large and well-funded projects investigating and in-
serting software reuse. There is a tendency among those organizations to neglect reuse in small
or embedded defense software projects. The STSC has researched a typicalsmall, emb edded
system and has results of that study. The STSC has also defined a reuse methodology that can
be applied to these systems with limited resources.
Keywords: Reuse, process, process improvement, embedded systems
Workshop Goals: Learn of actual reuse insertion projects, all lessons learned, reuse tools
being used, and start-up reusemetho dologies. Contribute information on embedded system
reuse efforts and small software project reuse.
Working Groups: Reuse process models; reuse maturity models; reuse management, organiza-
tion and economics; design guidelines for reuse; reuse and formal methods; useful and collectible
metrics; tools and environments; reuse handbook.
Smith- 1
1 Background
The Software Technology Support Center (STSC) is part of the Air Force Embedded Comput-
ers Software Improvement Program (ESIP). The overall objective of the STSC is to improve the
software process in the Department of Defense (DoD). The STSC's specific ob jective is to trans-
fer technologies and exchange information to satisfy its customers needs to continually improve
software quality and life cycle productivity. Sp ecial emphasis is given to tools and tool support.
The STSC has established an organization to investigate software reuse in the Air Force and
Department of Defense. The STSC has contracted with SofTech, Inc., to provide support in this
effort. Efforts include researching the state-of-the-art in software reuse, determining STSC customer
needs in software reuse, defining software reuse tools, cataloging and evaluating software reuse tools,
and developing a software reuse prototype.
2 Position
2.1 Limitations of DoD Efforts
The Department of Defense (DoD) has invested large quantities of money and personnel over many
years to investigate software reuse with the intent of reducing it to a viable approach [1]. Software
reuse has been envisioned as a method of controlling software expenses that take up such a large
portion of current and future budgets. Programs such as STARS, RAPID and AdaNET have spent
large amounts of resources directed towards necessary research and pilot implementations. The
results have been important and useful but their objectives have not always been in harmony with
the software needs of the Department ofDefense.
The emphasis in DoD reuse efforts hasb een towards large Management Information Systems (MIS).
This emphasis has left much of the software efforts in the DoD beyond the scope of reuse insertion.
This represents a significant neglect since thereis a significant amount of software in the DoD that
can not meet the parameters required by large reuse efforts.
2.2 Embedded Systems
A large amount of software in the DoD is embedded into weapon systems. Missiles, aircraft,
tanks, and command and control systems all rely nearly completely upon software for correct
mission fulfillment. Often the software process for these types of systems is poorly defined and
is implemented by a small group of people. Also, embedded systems are commonly written in
assembly language and are comparatively small. The systems have existed for many years yet may
be poorly documented and engineered. In many circumstances, this software may be most critical
to the national defense and requires quick and accurate upgrades to b emost effective.
2.3 Smaller Pro jects
DoD Reuse efforts do not provide suggestions and methods that areapplicable to smaller software
projects. There is a tendency to imply thata software project must be large in order to support
reuse. We are often told that successful reuse can occur only if a high level language is used, there
Smith- 2
is a defined process, a domain analysis is performed, and a formal repository is supported. This
necessitates a degree of formality and sophistication to software development that may not exist in
practice in small or embedded existing systems. The expected approach also necessitates a large
and well funded effort with strong management support and understanding.
2.4 Experiences
Reuse may occur without all of these expectations being met, although not in as large ascale. An
example is the F-16 software [2, 3 , 4 ]. There is a highdegree of functional reuse between various
versions of the Operational Flight Program (OFP), yet there is no formal attempt at reuse. It
occurs as a result of a small group of software engineers trying to do the most effective job.
Significant savings can be realized within the DoDby introducing reuse to small, embedded systems.
The savings may not be as measurable or as dramatic as otherlarger and more visible systems.
Yet more people can accomplish more work in less time, and this is a necessary goal in the era of
reduced defense budgets.
We can learn from the experiences of organizationswho have been forced into reuse by constraints
such as time or resource availability: They simply had to getthe job done with less resources and
implemented reuse methods as their approach. Examples of this are the Navy's RNTDS project
and ICBM software.
2.5 RNDTS
The RNTDS project [5 ] has a very high degree of successful reuse and was implemented before
methods were developed by such organizations as the Software Engineering Institute (SEI) or the
Software Technology for Adaptable, Reliable Systems (STARS)Project. The project engineers were
faced with the reality of not enough resources to do their job and developed a reusable software
system as a result.
2.6 ICBM Software
There were two significant cases of reuse in the ICBM: The development of Minuteman III from
Minuteman II technology and the development of Peacekeep er from Minuteman III technology.
This reuse was the result of critical national need where theluxury of re-developing software was
not available. No formal analysis of ICBM software reuse has occurred, but it is a fertile field for
determining methods and successes.
2.7 STSC's Contribution
The STSC has conducted a pilot program into a small, embedded system [6, 7]. The F-16 A/BHead-
Up Display (HUD) software (and the supporting process) was analyzed from a reuse perspective. It
was found that some reuse occurred as a natural result of engineers doing the b est job they could.
Many opportunities for reuse were identified thatwere not currently being utilized, especially in
non-code assets. There is a process in place that generally supports reuse, but a higher degree of
emphasis would result in measurable savings.
Smith- 3
2.8 Conclusions
Reuse will occur if it must be done to meet external forces. A formal, repeatable process as defined
by the major DoD reuse efforts does not have to be present in order to have success. The reuse-
oriented organizations need to realize this and offer help to those that are currently not being
noticed. The types of software reuse described in this section has often been dismissed as ad hoc
and thus not interesting. Yet this type of reuse has had and can continue to have a significant
benefits on software cost in the DoD.
2.9 Recommended modifications
DoD reuse efforts should take smaller and embedded systems more seriously in their research into
software process improvement. Much ground can be gained by simply educating organizations
involved in small and embedded systemsin the principles of reuse. There needs to be an attitude
change that these systems are not important or interesting enough to be considered when it comes
to software reuse.
2.10 STSC Efforts
Much was learned from the STSC's review of the F-16 A/B HUD software. These experiences can
be extrapolated into other F-16 OFPs and other embedded systems. The experience demonstrates
that there are viable possibilities for improving the software process with a reuse emphasis in small,
embedded systems.
The STSC has also produced a reuse methodology that can be applicable to any sort of system,
whether small, large, embedded or MIS. This methodology takes into accounts the funding and
resource limitations that may exist in small projects and deals with concerns of existing embedded
system projects.
3 Comparison
The DoD efforts in software reuse have a tendency to support the concept that reuse can only be
successful in large projects with well-defined domains and software processes. This implies also
that the projects must be well-financed and visible. The STSC contends that this unnecessarily
leaves a significant portion of the DoD software community out of the reuse technology.
References
[1] "DOD Software Reuse Initiative: Vision and Strategy."
[2] "Domain Reviewof the F-16 HUD OFP," tech. rep., SofTech, Inc, April 1992.
[3] "Review of theF-16 A/B HUD OFP C5 Document," tech. rep., SofTech, Inc, April 1992.
[4] "Review of theF-16 HUD OFP B5 Document," tech. rep., SofTech, Inc, April 1992.
Smith- 4
[5] J. Aridis, "Deriving and managing functional commonality for software reuse: An rntds expe-
rience," CrossTalk, February 1992.
[6] "STSC Reuse State-of-the-Art Report," tech. rep., STSC, August 1992.
[7] "STSC Reuse Metho dologyRep ort," tech. rep., STSC, July 1993.
4 Biography
Larry W. Smith is a Software Engineer at the AirForce Software Technology Support Center
(STSC). He is the Project Lead for Software Reuse. Additionally, he works closely with other
software technology domains to assist Air Force and DoD Software Development and Maintenance
Activities (SDSA) with software process improvement. He received a Bachelors of Science in Elec-
trical Engineering from the University of Utah, and is currently working towards a M.S. degree
in Computer Science from Utah State University. Vern Phipps is a Software Consultant with the
Ogden, Utah SofTech Office. He is currently the Ogden Office Reuse Project Lead Engineer. Mr.
Phipps previously worked for TRW, Inc. where he supported the analysis and MODE maintenance
of Minuteman and Peacekeeper software and performed nuclear safety analysis. Previous to TRW,
Mr. Phipps was employed by the National Computer Security Center and was assigned as the
NSA Liaison Officer for Computer Security for the I-S/A AMPEprogram at Gunter AFB, AL.
He received a Bachelors of Science degree in Computer Science from Brigham Young University in
1983.
Smith- 5