AN OVERVIEW OF INTELLIGENCE COMMUNITY AUTOMATIC DATA PROCESSING

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP83M00171R002100170001-6
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 43
• Document Creation Date: December 12, 2016
• Document Release Date: May 9, 2002
• Sequence Number: 1
• Publication Date: August 15, 1977
• Content Type: REPORT

Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PREBEARCH
PRESEARCH INCORPORATED
2361 S. JEFFERSON DAVIS HIGHWAY, ARLINGTON, VA. 22202 (703) 920--5740
DIA review(s) comcr ed For Rel~2'19791~3eq~gggpc~$t~F?~pyl~~,se.
 Approved For Rele %250pAq/83H: CIA-RDP83MOOl71 R002100170001-6
INCORPORATED
P.I. Log No. 77-0596; Copy
Preliminary Draft
Technical Report No. 349
AN OVERVIEW OF INTELLIGENCE
COMMUNITY AUTOMATIC DATA
PROCESSING
by
Leonard P. Gollobin and Ken R. Powell
15 August 1977
Prepared for
Director of Central Intelligence
Under
Contract 77-M-100400
2361 South Jefferson Davis Highway
Arlington, Virginia 22202
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For R94gf 6 fOZf1j3 :i P SRP)?I?aYgg171 R002100170001-6
This report summarizes the results of a brief assessment of
Automatic Data Processing (ADP) activities within the Intelli-
gence Community (IC), to identify promising avenues for future
improvements.
Users needs and commonality between hardware and software sys-
tems are reviewed, and some alternative concepts are presented
for data base management systems. Finally, policy and techni-
cal issues are identified together with means for achieving a
Community-wide information system.
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For RelegWE R L2/Q7Q3H CJSQORPP8d3A1AgJZ1 R002100170001-6
analyses should be part of an organized effort which address?s
the feasibility and value of adopting or developing a Comm=-nzry
standard data base management system. This effort also exterAs
beyond the SAFE/ADISS question, because with time the existir-z
"home grown" DBMS will be outdated technically and will
ret;uire
updating and replacement.
approaches and associated advantages/disadvantages of each.
tailed cost-benefit analyses are needed to home in on syst nE
that smoothly transition from the maze that exists today, -o
more efficient, centralized solutions for the future. 'fee
in elements of the IC have promoted decentralized planning ar1
management of software and hardware systems and data bases.
Many of these systems have common characteristics, and in_thE`
sense are duplicative. It is believed that these represen
portunities for improved efficiency and cost saving.
For example, the prevalent IC ADP systems perform commtni-
cation, collection, processing and production functions. oe
of these systems possess the characteristics of data base mar-
agement systems (DBMS), and are therefore considered aPtnaoi
to some consolidation, commonality and Community-wide access-;,g.
The CIA SAFE and DIA ADISS Systems fall into this general cate-
gory; in addition, 6 other major systems of this type were q ici-
ly identified in this brief overview, and we understand that -here
may be 30 such IC systems alitold. Hence the current Ccr.gt-es-
sional interest in SAFE/ADISS may in fact only be focused on the
tip of the iceberg!
This paper presents a few alternative concepts for dev?.-
oping broader based, Community-wide DBMS, merely to illustrate
The rapid growth of the Intelligence Community and the
orders of magnitude expansion of Automatic Data Processing wl
Fundamental to achieving Community-wide systems is t__e
resolution of a family of policy and technical issues whic:
complex, and have mostly been treated piecemeal in the past.
These relate to sharing data bases; agency access; technic-l
data standards and languages; interfaces and security. :p
ics are discussed in Section IV.
ii
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCOFPORATED
An area of immediate concern to the DCI is a timely, is
well thought out response to the Seaate Select Committee on
telligence. However, this response is only a subset of broil r
questions of Community-wide structuring, integration, plantain-,
and management. The SSCI presentation can be viewed as a
frogging opportunity to generate an agreed-upon roadmap for
short-, medium-, and long-range IC \DP planning. For exam?-Ae
creating a centralized management structure would be an ea-eat
of the DCI's intent to provide positive control over future
planning and utilization of Community ADP resources.
Regretably, resources on hand at the IC Staff level Opp:ar
woefully inadequate for the SSCI effort and certainly icr .r
broad Community planning function. Some specific steps nee to
be undertaken, perhaps providing soine help for handling the r:)st
immediate problems, followed by creation of a new, higher leP.el
management mechanism to address longer term ADP integration,
technical, budgeting and planning functions.
Approved For Release 2002/07/03 : CIA.RDP83M00171R002100170001-6
 Approved For Releases(0g/I7d(CgAil~eg&;l6RVE1~2002100170001-6
TABLE OF CONTENTS
ABSTRACT . . . . . . . . . . . . . . . . . . . . .
SUMMARY . . . . . . . . . . . . . . . . . . . . . -ii
LIST OF FIGURES . . . . . . . . . . . . . . . . . v
LIST OF TABLES . . . . . . . . . . . . . . . . . . vi
I. INTRODUCTION . . . . . . . . . . . . . . . . . . . 1
Background; Objectives and Scope;
Contents.
II. USER NEEDS AND COMMONALITY . . . . . . . . . . . . 5
Market; ADP Software System
Commonality; DBMS User
Requirements.
III. SOME ALTERNATIVE DATA BASE MANAGEMENT
SYSTEM CONCEPTS . . . . . . . . . . . . . . . . . 14
Standard DBMS; Alternative
System Concepts.
IV. COMMUNITY-WIDE INFORMATION SYSTEMS . . . . . . . . 28
Major Policy Questions; Planning
for a Community-Wide System;
Implementation.
1v
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH IN, ORPORATED
Page
2.1. Projected Daily Defense Communication Support
for Data Traffic . . . . . . . . . . . . . . . . . . 6
3.1. Community On-Line Intelligence System (COINS II)
Distributed ARPA-Type Network . . . . . . . . . . . 19
3.2 Use of Multiple Languages with COINS II . . . . . , 21
3.3. Alternative 1: Standard DBMS . . . . . . . . . . . 22
3.4. Alternative 2: Centralize Community Files . . . 23
3.5. Alternative 3: Multiple Language Translator . 25
4.1. Planning for a Community-Wide Information
System . . . . . . . . . . . . . . . . . . . . . . . 35
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For Relep*g 3W&Wf@3-C%4RDY M 7-l R602100170001-6
LIST OF TABLES
Page
2.1. Categories and Systems Reviewed . . . . . . . . . . 8
2.2. Collection and Communications Systems . . . . . . . 9
2.3. Processing and Production Systems with Data 10
Base Management System Characteristics . . . . . .
2.4. User Functional Requirements . . . . . . . . . . . . 13
4.1. Overview of Major Policy Questions . . . . . . . . . 30
vi
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
1.1 This report summarizes the results of a brief assessmeni
of Automatic Data Processing (ADP) activities within the Inte:-
ligence Community (IC). The work was performed for the Director
of Central Intelligence under Task No. 4 of Contract 100400.
1.2 The Intelligence Community utilizes automatic data proces-
sing systems extensively to collect, process and produce National
Intelligence.- In the past, each organization in the Community
has independently planned, managed, and budgeted for data pr(.ces-
sing related activities and computer hardware generally with~)ut
regard to the needs or existing capabilities of the Commun_Ltv as
a whole. The evolution of ADP and telecommunications within the
Community has resulted in the development of many different com-
puter software systems, some of which have duplicative charac-
teristics and most of which are nonstandardized. This histcrical
lack of centralized ADP management leadership and planning teas
resulted in the following:
? Duplication of technical efforts, which have
partially resulted from the rapid growth of
ADP in the Community
? Limited and questionable accountability for
resource expenditures (no effective ADP cost
or utilization/analysis)
-r Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
 Approved FopF~e~~/Q(/03 : CIA-RDP83M00171R002100170001-6
INCORPORATED
? Limited Community planning for system inter-
faces with only some interoperability and
interconnectivity between systems
? Lack of effective Community standards for
security, data elements and files, and sys-
tem performance (despite efforts of the In-
formation Handling Committee formed to pro-
mulgate standards)
? Proliferation of diverse computer hardware,
line protocols, and user terminals.
1.3 Recognizing these deficiencies, the Director of Central
Intelligence in the "National Foreign Intelligence Program
(NFIP) and Resource Guidance: FY 79-83" has expressed his deter-
mination that Automatic Data Processing and Telecommunications
are major issues that must be dealt with from a total Community
point of view. Other government organizations have made similar
observations: the Senate Select Committee on Intelligence re-
port of 19 May 1977 noted ADP resource activities need more
careful coordination, direction, and interagency planning; the
House Appropriations Committee report of 21 June 1977 was criti-
cal of the planned expenditures of fiscal rezources for agency
systems (CIA SAFE and DIA ADISS); and the Director of the Office
of Manpower and Budget stated that the Intelligence Community
should develop a coordinated approach to the development of com-
puterized data bases that maximizes the rapid and free flow of
information vital to the quality and timeliness of the intelli-
gence product and minimizes the duplication in both data files
and ADP equipment.
2
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For Relea R ~qOJ 7400 HCIA RDP83 l001E711 R002100170001-6
1.4 More specifically, the SSCI requested that DCI prepare a
comprehensive report and long-range plan for coordinated acqui-
sition and utilization of IC ADP resources by November 1977.
Types of detailed information requested include trend data and
analysis of ADP costs from FY 70 - 79 by agency; identification
and description of computer assets and use; near-term replace-
ments or upgrades; identification of data files and interagercy
exchange; projected major problems, priorities and new initia-
tives, and a broad organizational plan directed at improving
coordination of ADP hardware acquisition, software and data base
development, and interagency data base access. Efforts are cur-
rently underway in IC Staff and DoD to generate inputs for ti=e
DCI response. (As an independent observation, it is believed
that the Information Handling Division of the IC Staff which is
responsible for the DCI response, including development of a
Community-wide plan, is totally understaffed for responsibilities
of this magnitude at only four persons.)
1.5 The objectives of this study are to identify areas of com-
monality in present or contemplated Intelligence Community ALP
utilization and potential avenues for avoiding or eliminating
duplication, as inputs for future planning.
1.6 The scope of the project was intentionally broad to incor-
porate the Community as a whole with the depth limited to what
might be accomplished with about 1 man-month of research effort.
Hence the emphasis was on developing an overview of Community
ADP usage, rather than in-depth analyses of specific problem
areas. The approach used to gather information for this report
involved reviewing written material and interviewing individuals
within the Community.
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCORPORATED
1.7 The next section of the report reviews user needs and com-
monality between ADP systems. Section III discusses some alter-
native concepts for data base management systems, and the last
section examines some relevant issues for achieving a Community-
wide information system.
4
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
 Approved For Rp1W@e.-6Wf EI943 : R002100170001-6
II. USER NEEDS AND COMMONALITY
2.1 The rapid growth of the Intelligence Community has contrib-
uted to the duplicative and nonstandardized nature of ADP acStA lNTL
ties within the Community. 1/ The total fiscal funding for AJP
search and experimentation. -
2.2 The volume of available information within the Communit~j
is rapidly increasing. Advanced collection systems have expo-
nentially increased the availability of information. To effec.;-
tively handle this information, technology must be exploited o
ensure the analyst can access, extract, manipulate and output
data requested in a timely manner. The Defense Communication
Agency has developed a projection for the level of transmissijn
of data on AUTODIN II. This projection, displayed in Figure ::.1,
indicates the magnitude of the increasing exchange of Community
information--a fifty-fold increase in the period '72 to '76, .nd
The number of intelligence organizations has increased from
3 (FBI, Army's G-2, and Office of Naval Intelligence) in 11_941
to over 20 today.
STATINTL 2/ The Information Handling Committee's budget used for Com-
munity planning, research, and experimentation is less rharn
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
"RESEARCH IN,ORPORATED
3 x 1012
3 x 1011
ST
TINTL
3x 101
3x108
3x10'
3x 106 1 1 1 1 l __ A
1965 1970 1975 1980 1985 1.90
Year
FIGURE 2 . 1
PROJECTED DAILY DEFENSE COMMUNICATION SUPPORT
FOR DATA TRAFFIC*
25X-1 A
I IAUTODIN Ii," Proceedings of Department
formation System Managers Conference,
2.5-30 September 1976.
Approved For Release 2002/07/03 : CIA-RDP83MOO171R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCORPORATED
another twenty-fold jump by 1984, with accompanying growth in the
numbers of terminals and computers.
ADP SOFTWARE SYSTEM COMMONALITY
2.3 Categories of software systems were examined to identify a
select group of ADP software systems with commonality among Com-
munity organizations. This cursory look at systems with duplica-
tive characteristics centered around systems with characteristics
of data-base-management-systems (DBMS). Table 2.1 lists systems
that were identified and categorized according to: collection
systems (of which only imagery systems were reviewed); commun.ca-
tions systems; processing systems; and production systems. The
processing and production categories contained the systems with
data-base-management-like characteristics.
2.4 Because of satellite costs and the relatively short time
frame during which computerized imagery has existed, the imagery
collection systems have evolved, for the most part, to fulfill
specific Community needs. Communication systems (networks) a]-
ready involve a large number of Community users. Table 2.2 pro-
vides a brief description of imagery collection and communications
systems.
2.5 A more suitable target for this analysis is the prolifera-
tion of similar data base management systems within the proces-
sing and production categories. Table 2.3 provides a brief de-
scription of systems that are in the processing and production
categories that potentially possess duplicative characteristics.
While this list may include all types of systems, it probably
represents only a fraction of the total number of systems.
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Next 3 Page(s) In Document Exempt
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH 11',[ COR,10 RATED
DBMS USER REQUIREMENTS
2.6 By reviewing system literature and documentation, a list
of user requirements partly common to these DBMS systems was
constructed. This list, with refinement, could be used to iden-
tify the characteristics of a standard Community DBMS. If stan-
dard DBMS's could be developed or adapted to fulfill user re-
quirements, potential savings would result by not proliferating
new systems. Table 2.4 contains a description of user functicnal
requirements, summarizing the capabilities provided by all of the
systems described in Table 2.3.
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
 Approved For F3qje@W~~A0KI/8974P3 - 4' FAO IVfOIO Z1 R002100170001-6
TABLE 2.4
USER FUNCTIONAL REQUIREMENTS
DATA MANIPULATION
- Search (indexed or free text)
- Retrieve (browse, keyword, text, content analysis)
- Hold
- Edit
- Output
- Print
- Alert
FILE SUPPORT FUNCTIONS
- Maintain (create, change, delete, replace, add)
- Compose (page, scroll, insert, change, delete, move,
print)
- Storage, retrieval, and execution of past search
strategies
COMPUTATIONAL FUNCTIONS
- Unique processing functions, simulation languages,
APL, LISP, FORTRAN
- Statistical functions
- Curve fitting
- Mensuration and isometrics
- Models (heuristic, econometric, simulation)
COMMUNICATIONS FUNCTIONS
- Route (files and messages)
- Alert
USER AIDS
- Computer aided instruction
- Help function
- Diagnostics and error messages
13
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For RplW@e32LQ0A2WZ41( : ,CIA-oRDP8o3M001E p1 R002100170001-6
III. SOME ALTERNATIVE DATA BASE MANAGEMENT
SYSTEM CONCEPTS
STANDARD DBMS
3.1 SAFE of the CIA and ADISS of the DIA are two systems be:ng
planned that are currently receiving major attention and discirs-
sion. From the relevant literature prepared for these two sy!.-
tems (although ADISS does not have user requirements completely
defined), it appears that the two systems have a number of st?_n-
dard DBMS characteristics. This is alarming since several Other
systems, identified in Table 2.3, also have standard DBMS char-
acteristics, and the table contains only a small number of highly
visible DBMS systems. The COINS Project Office estimates that
there are more than 30 DBMS-like systems operational in the Com-
munity today. Many of these systems are "home-grown" and will
eventually come up for replacement as the state-of-the-art
advances.
3.2 The issue then becomes obvious. First, does the Community
need all of these DBMS systems, and second, can a Community stan-
dard be developed or adapted and implemented? It is believed
that the Community does not need all of these different DBMS sys-
tems. Organizations typically argue that their applications are
unique--when in fact they are not. The feasibility of a Community
standard for a DBMS or a set of DBMS systems is worthwhile inves-
tigating.
3.3 The Department of Commerce has on several occasions spon-
sored a Conference of Data Systems Language (CODASYL) and this
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
14
 Approved For Re. r e~ VA/R77Q3A : CIA-RDP83M00171 R002100170001-6
INCORPORATED
organization has been active in the development of a common Data
Description Language. Before additional retrieval languages are
developed within the Community and the magnitude of the ADP com-
monality/multiple language problem increases, the adaptation of
CODASYL-like standards should be investigated.
3.4 Ten years ago there were only a handful of specialized
retrieval languages, usually only suited for one type of hard-
ware. Today there are many "good" data base management systems
and retrieval languages. Before another DBMS system like SAF
or ADISS is developed, "off-the-shelf" DBMS systems should be
investigated. There maybe a significant opportunity to reduce
costs by building on these languages and modifying only part of
the existing larger DBMS systems. In developing or adapting new
DBMS systems, it is clear that the Community does not need the
following:
? Unique, inflexible and application-oriented
data base systems that are limited to nar-
row functional use
? Procedures that are oriented toward ADP-
inclined personnel
? Language not optimally designed for non-
ADP-trained analysts
? Unique DBMS applicable only to one or two
types of hardware.
ALTERNATIVE SYSTEM CONCEPTS
3.5 Some alternative conceptual designs for both the networking
of the hardware and the software interface are examined brief"y
below. Axiomatic to this investigation is the belief that a
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
15
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCORPORATED
"single united communications network" and a "common language to
access information" has the potential of contributing signifi-
cantly to Community efficiency and effectiveness.
Network Alternatives
3.6 There are at least three network alternatives that should
be considered, AUTODIN II, COINS, and IDHSC. A comprehensive
evaluation of these three network alternatives is beyond the
scope of this discussion, but the following summarizes status
and capabilities.
? AUTODIN II. Time estimates place the opera-
tion of this network in the CY 82-84 time
frame. The network will continue to be sup-
ported by Defense Communication Agency (DCA)
when it is operational. Since this network
is 6 to 8 yr from estimated completion, we
can only consider this alternative for the
future. However, current design compatibility
should be established to ensure that AUTODIN II
can be used when it becomes operational.
? COINS. The COINS I network is presently
operational, while COINS II is currently
being tested. COINS II will use the com-
munication technology developed in the ARPA
network, which could provide for secure com-
munication on common carrier lines. COINS
has made tremendous strides in providing in-
telligence analysts with a means to transfer
and communicate information on a world-wide
basis by linking to IDHSC. A wealth of ex-
perience exists both in the knowlege of
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Approved For ~e era en0#p7,T3 : CIA-RDP83M00171R002100170001-6
INCORPORATED
technical staff and the management staff.
The COINS Project Management Office (PMO)
is impressive in its forethought, experi-
mentation, and thorough investigation of
alternatives.
? IDHSC. The Intelligence Data Handling Sys-
tem Communication or the Worldwide Intelli-
gence Communications System (WICS) provides
DODIIS and COINS users with worldwide com-
munications. This DIA sponsored network
currently supports DIAOLS (DIA's data base
management system). COINS and WICS II have
been coordinated on a technical basis, but
there are still major incompatibilities in
the communication technology used by each.
One problem claimed by IDHSC to have been
resolved is the massive interfacing required
with CIA, NSA, State, etc.
3.7 The two alternative network technologies for near-term use
appear to be COINS and IDHSC, since the time estimate for AUTO.-
DIN II operation is CY 82-84. From this limited analysis,
COINS II technology appears the most promising candidate for a
near-term solution to a single communication network, if the
COINS II tests are successful. The duplicative characteristics
of COINS and IDHSC require that both be carefully evaluated to
ascertain whether both networks are justified.
Multiple Language Problem
3.8 Alternative conceptual system (hardware and software) de-
signs can be identified regardless of the decision to use COINS,
IDHSC, or some other third undetermined network alternative.
Approved For Release 2002/07/03 : cAAA-RDP83M00171 R002100170001-6
 Approved Forg D-1
a eE2~ ?/8T/031 NCj_~-WRgVYPQ'71 R002100170001-6
But the experience and "lessons learned" in COINS and IDHSC
should be used as a basis for long-term improvements in informa-
tion handling and processing.
3.9 The changing economics of hardware, the outlook for data
transmission, and the large volume of data contained within t1-e
worldwide intelligence community combine to suggest distributed
system architecture. Distributed data base systems are attrac-
tive where data are generated at several locations and are needed
there for further processing, yet some applications require theta
stored at other locations as well. Distributed data base sys}ems
provide the ability for parts of Community data to be managed by
several processors and provide the analyst with the ability to
view data from a single location and not be concerned about tie
geographical location of the data. COINS II and IDHSC are being
designed using the distributed data base concept. Figure 3.1
conceptually displays the COINS hardware network.
3.10 There are technical problems associated with distributed
systems that may be significant ("lockout" and "deadlock") an.
these issues will require a detailed knowledge of the hardware
at each system node. Using the distributed data base concept,
a few alternatives are identified below as means for establish-
ing a common language to access information.
3.11 The multiple language problems currently facing COINS 11
or the DODIIS/IDHSC network are displayed in Figure 3.2. A user
enters his information requests/commands via a computer terminal.
That terminal is connected to computer hardware at location P..
If the data requested is resident at location "A," then DBMS [
is used to access the data files. But, if the user wants to ac-
cess files at location "B,", then DBMS II must be used to acL-ess
data files. This multiple language problem is compounded ea'..h
18
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 1 II [ 1 1. l 1 I 1 1. 1 1 1 1 t 1 1 1
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
Agency A I
I 1 MY IMP
Agen
Agency C
IMP IMP
I
TAS CPU Terminals
TAS = Terminal Access System
IMP = Interface Message Processor
CPU = Central Processing Unit
F1~titar. z ~
COMMUNITY ON-LINE INTELLIGENCE SYSTEM (COINS II)
DISTRIBUTED ARPA-TYPE NETWORK
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
Agency B I
Agency D j
I
I
CPU I
 Approved For Release 2002JQ 3 : CIA-RDP83M00171 R002100170001-6
PRESEA INCORPORATED
time a new node is added with a different language (some agen
cies have multiple languages).
Alternative 1: Standard DBMS
3.12 One alternative for solving this problem is to replace all
data base management languages with a Community standard DBMS
language. Figure 3.3 displays a conceptual view of this alterna-
tive.
3.13 The advantages and disadvantages of this approach are
listed below.
Advantages Disadvantages
1. Standard DBMS 1. Costly solution
2. Single language 2. Long conversion time likely
3. File sharing and processing 3. Drain on personnel resources
possible 4. Bound to one DBMS architecture
S. Major impact on users
Alternative 2: Centralize Community Files
3.14 Another alternative for solving this problem is to central-
ize all Community files at one node of the network (probably us-
ing current hardware). Figure 3.4 displays a conceptual view of
this alternative, whose advantages and disadvantages are described
below.
Advantages Disadvantages
1. Community standard DBMS 1. Provide single failure po:nt
2. Maximum of two languages 2. Duplicate files
3. Centralized management 3. Added file conversion and
control maintenance
4. Large file size
Approved For Release 2002/07/0: CIASRDPWpJJ' P,O91 79 01f 1 e s
 I I I t I I I I I I
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
USER
TERMINAL
NODE
HARDWARE
A
IMP*
NODE
HARDWARE
B
CURRENT
DBMS I
NODE A
DATA FILES
NETWORK
* IMP = Interface Message Processor
CURRENT
DBMS II
I I I
NODE B
DATA FILES J
FIGURE 3.2
USE OF MULTI PLE LAN a vw h i o 11
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
 i 1 t L [ I l l [ 1 I I [ I
Approved For Release 2002/07/03 : CIA-RDP83M00171 R002100170001-6
USER
N
N
NODE
HARDWARE
A
STANDARD
DBMS
NETWORK
NODE A NODE B
DATA FILES DATA FILES
* IMP = Interface Message Processor
FIGURE 3.3
ALTERNATIVE 1: STANDARD DBMS
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
NODE
HARDWARE
B
STANDARD
DBMS
 C i 1 1 1 I I 1 1 1. 1 1 I I i
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
USER L
TERMINAL
NODE
HARDWARE
A
NODE
DBMS
NODE A
DATA FILES
NETWORK
HARDWARE
SERVING
COMMUNITY
COMMUNITY
DATA FILES
IMP *
* IMP = Interface Message Processor
FIGURE 3.4
ALTERNATIVE 2; CENTRALIZE COMMUNITY FILES
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCORPORATED
Alternative 3: Multiple Language Translator
3.15 A third alternative for solving the multiple language prob-
lem is to develop a multiple language translator. This alterna-
tive, displayed in Figure 3.5, would allow the user to use cur-
rent node DBMS language as the standard Community language. Ad-
vantages and disadvantages are described below.
Advantages
1. Minimal impact on users
2. Single Community user
language
3. Minimal conversion costs,
if no existing files need
to be converted
Disadvantages
1. Needs multiple language transla-
tion packages (potential risk?)
2. Difficult to maintain individ-
ual translation packages fcr
each vendor
3. Additional personnel required
to maintain translation pack-
ages
Alternative 4: Multiple Language Translator and Standard DBMS
3.16 This is the last alternative that is presented, although a
continuum of alternatives is possible. 11 Essentially, there is
a combination of alternatives 1 and 3, which seems particularly
relevant in light of the recent SAFE/ADISS interest and discus;-
sion. As part of the development of SAFE/ADISS, a user language
must be defined. The characteristics and attributes of the defi-
nition of this language could be used to create a multiple lan-
For a more detailed description of alternatives to the
multiple language problem (without regard to SAFE/ADISS?,
see a "Study of Multi-Language Problems in COINS," May
1975,
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 i k t t i 1 i kt t t t
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
USER
NODE
HARDWARE
A
TERMINAL LANGUAGE
TRANSLATOR
CURRENT
DBMS
NODE A
DATA FILE
NODE B
DATA FILES
* IMP = Interface Message Processor
FIGURE 3.5
1LTL ylyTI Y`L 3 . 4ULTIFLI LA GUi i re v>-, +,~-~
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
NODE
HARDWARE
B
LANGUAGE
TRANSLATOR
CURRENT
DBMS
 Approved For RpjVEc 2LOOA2, Q Q : C4q gPP A42Q lL71R002100170001-6
guage translation package. The actual development of SAFE/AMISS
could be directed to the needs of the entire Community and wc,uld
result in the development of a standard DBMS.
3.17 The advantages and disadvantages of this approach are
described below.
Advantages
Disadvantages
1. Single Community language
2. Community standard DBMS
that could be phased in
over time and serve as
DBMS for SAFE/ADISS
3. Minimal conversion costs
4. Minimal impact on users
5. File sharing and proces-
sing possible
Summary
1. Multiple language translation
packages must be developer
2. Maintenance would be difficult
for translation packages, but
would phase out over time
3. Additional personnel may be
required to maintain transla-
tion packages
4. Bound to one DBMS architecture
5. Requires strong, central man-
agement control
3.18 The selection of an approach to solve the multiple language
and DBMS problem presents a challenge in achieving a reasonab=le
balance between the expected user benefits and associated costs.
Therefore, to perform a comprehensive evaluation of the alter-
natives presented in this section or any others, the costs ard
implementation schedule associated with each alternative must be
developed in detail. However, certain relevant observations can
be made. There are a large number of DBMS-like systems within
the Commuinity, and these systems will eventually come up for
replacement as the state-of-the-art advances. SAFE and ADISS
may well be part of what will be a continuing trend to upgrace
and replace obsolete "home grown" DBMS systems. As a result,
alternative 4 (Multiple Language Translator and a Standard DBMS)
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
26
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCORPORATED
appears to be an attractive compromise because this alternative
provides a means to move toward a Community standard while min-
imizing the impact of change, and eventually evolving toward a
standard. However, to refine this approach, in-depth cost/ben-
efit analyses should be performed for these and other possible
alternatives. This effort should identify and describe the spe-
cific characteristics of the DBMS system and resident hardware
as well as describe the costs and benefits associated with each
alternative.
27
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
 Approved For Relee 2QVRK/ CIA F P8o31K0 1 J1 R002100170001-6
IV. COMMUNITY-WIDE INFORMATION SYSTEMS
4.1 Community data, that is, intelligence data needed by more
than one organization, is a reality. The set of complex policies
and technical problems accompanying this reality must be deter-
mined to establish an efficient and effective manner by which to
communicate and transfer data within the Community. The concept
of Community data has been demonstrated in the Community On-line
Intelligence System (COINS), a Community test-bed system that
links NSA, CIA, DIA, NPIC, and State, and which can be accessed
by world-wide commands. This system currently provides service
for 9,000 queries per month on over 65 files. Also, the Depart-
ment of Defense Intelligence Information System (DODIIS), an
evolving DoD Community system, has demonstrated the concept of
Community data by providing for the exchange of information on
part of over 175 files contained within DODIIS. In fact, 86 bil-
lion characters of automated intelligence information (enough for
300 complete sets of Encyclopedia Britannica) are exchanged an-
nually within DODIIS alone. l/
4.2 Potential cost justification for Community information sys-
tems has been demonstrated by COINS. Over 100 million communica-
tion transmissions were saved annually in COINS by eliminating 13
NSA reports, most of which were daily or weekly publications.
However, this is only a small step in the cost savings that could
be realized by the Community. The fact that there are software
systems within the Community that have duplicative capabilities
High, Paul L., Jr., "DoD Automated Intelligence Flow," Pro-
ceedings of Department of Defense Intelligence Information
System Managers Conference, 26-30 September 1976.
Approved For Release 2002/07/03 : CMRDP83M00171 R002100170001-6
 Approved For Pp~tsg J0R220I 3 :, CIAa K 3M,.Qpjp1 8002100170001-6
and that fewer systems could potentially serve the Community
provides a basis for significant future cost savings.
MAJOR POLICY QUESTIONS
4.3 To lend perspective to the larger problem of coordina~:.ing
and integrating the ADP planning elements within the Commun:it-ti'
for the development of a more accessible Community informat:io:-i
system, it is important to understand the areas in which poli_-y
must be defined for Community ADP activities. For the purpos:.
of the overview, we have identified the following seven areas
? Shared data bases
? Multiple retrieval languages and data base
management systems
Data standards
? Communication network interfaces
? Training and user aids
? Research and experimentation
? Security
Table 4.1 describes specific issues and some alternative ap-
proaches associated with each area of major policy. Each of
these policy questions is complex and has been addressed to some
extent over the past several years by IHC and special subcommit-
tees and study groups. It is believed that to achieve signifi-
cant benefits of Community-wide efficiency and effectiveness via
commonality will require "reasonable" resolution of all of these
thorny policy (and related technical) questions.
Approved For Release 2002/07/03 : C11RDP83M00171 R002100170001-6
 t t
I C l I It 1 t I i i
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
TABLE 4.1
OVERVIEW OF MAJOR POLICY QUESTIONS
Policy Areas
Specific Issues
Alternative Approaches
et Responsibilities: What organi-
Bud
? Host computer facility
g
zation is responsible for budgeting
ta files and
f d
? Proponent of data base
? Shared with users on a proportionate
a
for Community use o
validating the cost/benefit of Com-
use basis
? Information Handling Committee (IHC)
munity data file?
? Community Council (new organization)
Shared
data bases
Access $ Control Responsibilities:
? IC staff
What organization is response e
authorized access?
nsurin
f
? IHC
? Proponent of data base
g
or e
? Other group
Quality $ Timeliness of Data: What
? Proponent of data base
ensures the accuracy, quality and
timeliness of the data?
? IHC
? Community Council
? Other group
Community Standards: Should there be
? Single standard
community standard for retrieval
? Community guidelines
Multiple
language or DBMS's?
? Multiple standards
No standard
Retrieval
?
Languages G
Maintenance Update $ Modification:
? Host computer facility
Data Base
at organization is responsible
? Community Council
Management
for the maintenance required for
? Contractor(s)
Systems
operational use of these languages/
? DODIIS/COINS
(DBMS)
DBMS's?
Multi le Language Translator: Can a
? No
multiple anguage trans ator be.
? For some languages
developed?
? All languages
Necessit : Does the Community need
dards?
t
? None
For selected areas
Data Element
an
data element s
For historical and new data files
Standards
? For new data files
haw mucii
Policy:
? Individual agencies
--n n?
standardization?
? IHC
? Community Council
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6
 Y 1 t 1 1 1 1 1 1 1 1 1 t
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
TABLE 4.1 (Cont)
Policy Areas
Specific Issues
Alternative Approaches
Approach: Given that data standards
? Solve technically by developing a
are required, what approach should
cross-walk file
be used to create standards?
? Use a committee to identify data
Data
element standards
Standards
? Use a contractor to propose data
(cont)
element standards
a Create a new Community group (in a
full-time capacity) to address
data standards
Budget Responsibilities: What organi-
? IC staff
zation is responsible for budgeting
? IHC
for Community data standards?
? Others (e.g., Community Council)
Management Control: What organiza-
? DCA
Communica-
tion is responsible for defining
a IHC
tions
line protocols and gateway inter-
? IC staff
Network
faces, and ensuring exchange of
? Committee of Host Agencies
Interfaces
technology?
Maintenance: Ghat organization is
? Host agencies
responsible for maintaining inter-
? Contractors
face software?
? Interagency working group
? IHC
Training
Responsibilities: What organization
? IC staff
and
has the responsibility for training
? DODIIS/COINS
User Aids
Community users?
? IHC
? Others (e.g., Info Science Center, CIA)
? File sponsors
Management Responsibilities: What
? IHC *.Community Council
organization(s) has responsibility
a COINS ? IRF,D Council
Ex erimenta-
for coordinating $ directing over-
? ARPA
Lion and
all experimentation $ research?
? IC staff
Research Budgetary Responsibilities: What or-
a All agencies
h
ganization s as responsibility
d
? Selecte agencies
fur budgetiug fox Cuwuwiity experi-
? ? IC staff ? Community Louncii
1
F
Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
 I I C I I I l I I I i
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
TABLE 4.1 (Cont)
Policy Areas
Specific Issues
Alternative Approaches
Authoritative Source: Current refer-
? IHC
ence material on security for ADP
? IC staff
is vague and inadequate. What
? DCA
organization is responsible for
? Individual agencies
developing an authoritative refer-
? Other Community group
ence that defines security require-
ments for a transmission over a
network and data stored in computer?
Inspection F Enforcement: What orga-
? IHC
nizational group is responsible for
? IC staff
investigating ADP hardware and soft-
? Individual agencies
ware (i.e., data base protection,
? Technical accreditation group
Security
valid "need-to-know" operating
system security, etc.)?
Research $ Experimentation: What
? IHC
organization is responsible for
? IC staff
budgeting for research and experi-
? DCA
mentation and is tasked to resolve
? Individual agencies
new problems brought on by techno-
? Other Community groups
logical developments?
Sharing Community Data: What organi-
? IC staff
zational group is responsible for
? IHC
threat definition, risk clarifica-
? Other Community group
tion, and other security issues
? Individual agencies
associated with the sharing of
Community data?
Approved For Release 2002/07/03 : CIA-RDP83M00l71 R002100170001-6
 Approved For RelepReL~D
Fb0P q 1F 91~$ q1 i 1 R002100170001-6
4.4 The list of policy issues compresses to the following basic
questions in each area:
Policy Area Issues
? Shared Data Bases-- What organization(s) has the
responsibility for Budgeting,
ensuring proper access and
control, and verifying the
quality and timeliness of in-
formation contained within
data bases; what is to be
shared and with whom?
? Multiple Retrieval Should there be a Community-
Languages and Data standard language, since this
Base Management is a major interface problem
Systems-- to the analysts? Is it fees-
ible to develop and maintain
a language translator? What
organization is responsible
for maintenance required for
operational use of these re-
trieval systems?
? Data Standards-- Does the Community need stand-
ards and is the Community will-
ing to allocate the resources
required? If so, what approach
should be used (solve techni-
cally, use Committee, use
contractors, use in-house
personnel)?
? Communications What organization has manage-
Network Inter- ment responsibility for de-
faces-- fining communication standards
(line protocols; gateway in-
terfaces, etc.) and evaluating
competing network technologies?
? Training and User What organization is responsi-
Aids-- ble for training Community
users and what is the cost to
the Community, and is present
investment adequate?
? Experimentation and What organization has responsi-
Research-- bility for coordinating overall
Approved For Release 2002/07/03 : C#$-RDP83M00171R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCORPORATED
experimentation and research
and avoiding duplication? By
what means are the results
passed to the Community?
? Security-- What organization is responsi-
ble for developing an authori-
tative reference on ADP secur-
ity? How can an individual
organization's security proced-
ures be effectively inspected
and accredited? What are the
multi-level compartmentation
problems associated with ADI'
security?
4.5 Development of a Community-wide system is a complex process
involving centralized direction and interagency coordination and
planning. Some of the steps which converge on a master plan for
developing such a system are listed below. What is contained in
these steps, and how they relate, is illustrated in the suggeEted
flow diagram, or "plan for the plan," shown in Figure 4.1.
? Identify current information system capa-
bilities and existing hardware configura-
tions
? Evaluate current information handling
capabilities
? Study present organization of Community
ADP elements
? Improve management control of ADP elements
in the Community
? Perform top-down information requirements
study
Approved For Release 2002/07/03 : (N4-RDP83M00171 R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
(.RESEARCH INCORPORATED
IDENTIFY CURRENT INFORMATION
SYSTEM CAPABILITIES #r EXISTING
HARDWARE CONFIGURATIONS
-Identify cost and workload
trends
Identify inventory of com-
puter systems user agencies,
purpose, degree of interface
and life span
- Identify projected replace-
ments
- Identify current major soft-
ware systems
-Describe major data files
and degree of information
exchange
Develop dictionary
describing hardware
and software
capabilities
EVALUATE CURRENT INFORMATION
IANDLING CAPABILITIES
- Evaluate systems with dupli-
cative characteristics
(including SAFE/ADISS $
COINS/WICSII)
- Identify and evaluate incon-
sistencies and information
gaps
-Evaluate systems on
-Timeliness
Relevance/need
-Completeness
-Security
-Performance
- Perform cost/benefit
analysis
IDENTIFY INFORMATIONAL
FLOW
IDENTIFY FUNCTIONAL
REQUIREMENTS OF
ANALYSTS
DC_' 8
PRC (I )
Decisions
STUDY PRESENT ORGANIZATION
OF COMMUNITY
ADP ELEMENTS
- Each organization in
relation to the Community
-Management control of
each organization
CONDUCT POLICIES F I
STANDARDS REVIEW
PERFORM TOP-DOWN
INFORMATION REQUIREMENTS
STUDY
IDENTIFY AND PROJECT
INFORMATION NEEDS
FOR COMMUNITY
C '7
IMPROVE MANAGEMENT CONTROL
OF ADP ELEMENTS IN COMMUNITY
- Define new authority and
responsibilities
- Identify new policies
- Provide fiscal resources and
staff
- Describe new fiscal reporting
procedures for Community agencies
DEFINE AND EVALUATE
HARDWARE AND SOFTWARE
ALTERNATIVES FOR COMMUNITY
1979-1985
- Community networks
- Standard data base management
software
-Security standards
-New hardware
-Define R&D activities
PREPARE AND UPDATE
COMMUNITY ADP ACTIVITY
MASTER PLAN
)COMMUNITY J
CONDUCT FISCAL
REVIEW
PLANNIN F-OR A C20 2/ 7T 3Y IDFD INFORMATION SYSTEM
pprove or a ease 0 0 IA-R P83M00171R002100170001-6
S
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESE.ARCH INCOPPORATED
Define and evaluate Community hardware and
software alternatives
0 Prepare and update Community ADP master plan.
IMPLEMENTATION
4.6 Currently, the DCI Intelligence Information Handling Com-
mittee is charged with the responsibility to serve as the cata-
lyst for bringing about Community-wide ADP planning and implemen-
tation. The available funds, manpower, and the expertise re-
quired to perform this crucial responsibility are not adequate
to allow the IHC to effectively execute this stated responsibil-
ity. Further, the Committee concept has demonstrated little
progress in addressing the most basic Community-wide ADP prob-
lems (i.e., data standards, security issues, etc.).
4.7 To provide for future integration and planning efforts, it
would be valuable to identify a central focal point to coordinate
the development of new hardware and software systems to reduce
redundancies and to maximize utility across the Community. At
the same time, reporting and budgeting policies need to be changed
to identify allocations of ADP resources, beyond the limited ac-
countability found today.
4.8 One approach would be to dissolve the IHC and to create a
new "Office" with line responsibilities. This office could re-
side within the IC Staff, or perhaps be answerable to the DCI
directly, since ADP integration efforts are so fundamental to
Community-wide integration as a whole. This office would be
charged with ADP Community-wide planning and implementation and
be the central focal point within the Community for the coordi-
nation of new major software systems, reducing the possibility
Approved For Release 2002/07/03 : MA-RDP83M00171 R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PF ESEA RCH INCORPORATED
of major system redundancy. As part of its responsibilities.
the office would have budgetary authority over ADP activities,
and related Community R&D programs. This new office would re-
solve the limited and questionable accountability for ADP re-
source expenditure by developing and implementing new report:_ng
and budgetary policies.
Approved For Release 2002/07/03 : CRDP83M00171R002100170001-6
 Approved For Release 2002/07/03 : CIA-RDP83M00171R002100170001-6
PRESEARCH INCORPORATED
2361 S. JEFFERSON DAVIS HIGHWAY, ARLINGTON, VA. 22202 (703) 42tt-5740
Approved For Release 2002/07/03 : CIA-RDP83MOOl71 R002100170001-6