COINS, IRAC WORKING GROUP REPORT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82M00531R000400010019-4
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 61
• Document Creation Date: December 16, 2016
• Document Release Date: April 21, 2005
• Sequence Number: 19
• Publication Date: April 18, 1973
• Content Type: MF

IC
ARApproved For Tease 2005/05/??CBIATRDP82M00531004000100
F~kri/
18 April 1973
SUBJECT: COINS, IRAC Working Group Report
Attached is a copy of subject study which was forwarded to the
IRAC members by the Director of Central Intelligence (DCI) on 16 April 1973.
There is also included copies of correspondence on this material by the
Assistant Secretary of Defense (Intelligence) and the DCI.
25X1
Executive Secretary (Acting)
25X1
Approved For Release 2005/05/23 : CIA-P
MEMORANDUM FOR: IRAC Working Group
 Approved Fort ,@Jease 2005/05/23 : CIA-RDP82M00534i&000400010019-4
DR/wi
COINS
REPORT
1 February 1973
DRA1 f Approved For Release 2005/05/23 : CIA-RDP82M00531R000400010019-4
 Approved For RWease 2005/05/23 : CIA-RDP82M00531WO0400010019-4
SUMMARY
REPORT
I. INTRODUCTION
II. OBJECTIVES
III. METHODOLOGY
IV. RESULTS
V. CONCLUSIONS
VI. RECOMMENDATIONS
ATTACHMENTS:
A. STUDY RESULTS
B. FILES
APPENDICES:
I. TERMS OF REFERENCE
II. COST BREAKDOWNS
III. COINS II PROPOSAL
Approved For Release 2005/05/ 3 IA-RDP82M00531 R000400010019-4
 SECRI T
Approved For lJease 2005/05/23 : CIA-RDP82M00531p00400010019-4
This report covers the activities of COINS Review Group,, from
October. 1972 to February 1973. Tasked by the ASD (I) under the auspices
of the IRAC, the Group examined the broeed issue of automated intelligence
information exchange, and evaluated the COINS experiment as a viable
means toward promoting such exchange.
The major study conclusions are generally positive:
1. There is a, present need for automated information exchange
among members of the intelligence commaunity, and this need will rapidly
increase in. the future with the advent of operational systems capable of
delivering data in near real-time.
2. The COINS system has demonstrated the feasibility of on-
line data base exchange..
3. As presently configured, COINS has enhanced the effective-
ness of intelligence analysts, resulting in. overall manpower savings,
and relieving burdens on certain communication facilities.
4. The benefits of COINS have been achieved in a highly cost-
effective manner, utilizing existing hardware and software.
Several notable weaknesses in the COINS configuration have been
revealed.
1. COINS executive support is not presently at a high enough
level to deal effectively with the problem of coordinating diverse intelli-
gence agencies.
2. The files accessible through COINS are not an optimum
set of existing files.
3. The system operates at the SI security level and suffers
from the exclusion of vital TK data.
5. The present COINS configuration is growth-limited with
respect to files which may be accessed and speed with which access
may be achieved.
Approved For Release 200/1J3) : RDP82M00531 R000400010019-4
 Approved For F. tease 2005/05/23 : CIA-RDP82M00531QD00400010019-4
Based on the above conclusions and problem highlights, the
following recommendations are made:
1. The ASD(I) be appointed COINS executive agent, under
the auspices of the DCI.
2. The COINS Project Office be placed under the direction
of the executive agent, and be staffed by representatives of the entire
community.
3. COINS funding be continued, under the following stipula-
a. The present system be upgraded to the TK level.
b. COINS be purged of unused files and these files re-
placed by the files of interest to the community, as identified in this
study.
c. COINS management reporting be strengthened to
allow evaluation of usage, timeliness, and effectiveness.
d. The COINS PMD submit a detailed plan for an up-
graded system which will address the wealuiesses previously identi-
fied in a cost-effective manner.
Approved For Release 261,T/M?7CIA-RDP82M00531 R000400010019-4
 . btui t I
Approved For Flease 2005/05/23 : CIA-RDP82M00534,$000400010019-4
FINAL REPORT
of the
COINS REVIEW GROUP
1 February 1973
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
QLAEI
 SECRET
Approved For 1ease 2005/05/23 : CIA-RDP82M005318000400010019-4
I-i
1. INTRODUCTION
The COINS Review Group was established under the general
auspices of the Intelligence Resources Advisory Committee to review
and evaluate the COINS system as a means for automated information
exchange among intelligence agencies in the Washington area. This
report is the result of the review, performed between November 6,
1972 and February 1, 1973.
Members of the Review Group were:
Department of State Mr. Brewer J. Merriam
Director, Office of Current Intelligence
Bureau. of Intelligence & Research
Mr. William P. Deary
Deputy Director, Office of Current
Intelligence
Bureau of Intelligence & Research
Office, DCI
M/Gen Robert Taylor III
Chairman, Information Handling
Committee
Chief of IHC Support Committee
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
.25X1
25X1
25X1
 Approved For lease 2005/05/23 : CIA-RDP82M00531W00400010019-4
25X1
OASD(I), Chairman Dr. Myron S. Malkin
Deputy Assistant Secretary of Defense
(Technical Evaluation)
Members of the Working Group were:
Department of State Mr. Don Stigers
Mr. Julian Adler
OASD (I), Chairman Mr. Norman Solat
Assistant for Science & Technology
ODASD(SE)
25X1
Approved For Release 200 1+Q5/23'-:,13WRDP82M00531 R000400010019-4
 Approved For Release 2005/05/23 CIA-RDP82M00531QP00400010019-4
II. OBJECTIVES
The objectives of the Review Group are best stated in the context
of the objectives of the COINS experiment itself, with particular reference
to the background and the history which led to the experiment. Briefly,
the origins of COINS stem from a review of the operation of the
Washington area intelligence community performed in 1965 by the
President's Foreign Intelligence Advisory Board. The PFIAB in-
formcd the President that community progress in applying informa-
tion science and technology was unsatisfactory and stated: "Unless
strong and immediate actions are undertaken in this area, there is
danger that the efficiency of the production and dissemination of intelli-
gence within the intelligence community will decline progressively, and
that the already high costs involved will climb so steeply as to jeopardize
rational support of the broad intelligence effort".
Three actions: were recommended, the second of which led even-
tually to what is now called COINS. This was that the NSA Technical
Information Processing System (TIPS) project be expanded to include
participation by other member agencies of the intelligence comrr:'.;nity
in an experiment as a first step toward interagency information handling.
The President approved the recommendations and directed the DCI to
undertake the responsibility for implementing them.
In July 1967 the PFIAB reported to the President that although
the community had-made noteworthy strides in information handli: ;,
the problem was not being addressed adequately on a concerted com-
munity-wide basis, and made several recommendations for a phased
implementation of a community-wide information handling system. By
May 1968 the COINS experiment was more clearly defined and an
Evaluation Panel was to be established by the IHC to evaluate the experi-
ment as soon after July 1969 as was practical. Due to the continuously
fluctuating nature of the COINS system, that evaluation was not performed.
In September 1972, in response to an NSA proposal to proceed with
the development of a COINS II system, the ASD (I) suggested that a group
not presently charged with implementation of the system review the
broad question of information exchange under the general auspices of
IRAC. Consequently, the COINS Review Group was formed. The
Group was charged in its Terms of Reference (Appendix I) with the
following objectives:
Approved For Release 2005/05/23 : CIA_ -RDP82M00531 R000400010019-4
 Approved For lease 20053+:DP82M00534W00400010019-4
1. To review and evaluate the COINS system, determining; its
effectiveness in meeting the present and anticipated requirements of
the user community.
In this context, due to difficulties in arriving at a precise
operational definition of COINS, the Review Group found it advisable
to expand the objective to allow
a. Evaluation of the broad requirement for automated informa-
tion exchange.
b., Evaluation of the ability of COINS, as presently consti-
tuted, to satisfy that requirement.
2. To identify those areas where the effectiveness of COINS
in improving the intelligence product may be enhanced.
3. To recommend courses of action and consider their impact
on present and future budget allocations.
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
.SECRET
 L (LL
Approved For P.Ipjease 2005/05/23 : CIA-RDP82M0053 .000400010019-4
III-1
III. METHODOLOGY
The Review Group, formed in late October 1972, was comprised
of representatives of the DCI, the national agencies involved in automated
intelligence information handling (NSA, CIA, and the Department of State),
DIA as the DOD intelligence producer, a major COINS participant, and
representative of the services and U. S. Commands, and chaired by the
ASD (1) representative.
Sitting in regular sessions at approximate intervals of two weeks,
the Review Group established detailed Terms of Reference, and outlined
a phased program of activities which would permit the fulfillment of the
objectives within the allotted time. A continuing Working Group was
formed with representation from each of the above organizations, and
chaired by an ASD (I) representative. This latter group was augmented,
as required, with additional personnel possessing needed substantive
expertise. The Working Group was responsible for the performance of
the study tasks and the maintenance of the schedule.
The Review Group was briefed at various times on subjects of
interest (e. g. , description of COINS, evaluation efforts, and the problems
of computer security). As constituted, the Group was able to foster a
significant interagency dialogue covering each of the major issues.
The severe time constraints for the study effort imposed certain
limitations on the ability to generate new information. Therefore, the
Working Group made maximum use of existing data, reports, and
statistics, and whatever on-going work was available. However, for
analysis of the total community file structure, it proved necessary to
develop information which did not exist previously. The Working Group
solicited and obtained descriptive material from each agency regarding
all pertinent files, whether or not presently classified as a COINS file.
The compilation of file information, together with agency expressions of
interest and data regarding use, represents a significant contribution of
previously unavailable information which was used to derive some of the
major conclusions and .recommendations of this report. This compilation
is included in Attachment A.
Approved For Release 200~5/5~2 ~:4-RDP82M00531 R000400010019-4
 ~/i,.RV 4f.v i
Approved For Rg ease 2005/05/23: CIA-RDP82M00531 Q00400010QV 4
IV. RESULTS
A summary of the information gathered during the course of the
study is presented as Attachment A, and briefly commented on in this
section.
The results cover two general topics:
A. System Descrinntions - current system and. alternatives,
including advantages, disadvantages, limitations, and costs.
B. Files - summary of the data on Washington area files,
including COINS and non-COINS files, with comments regarding use
and interest.
A. 5 stem Descri bons
1. Current System
COINS is a tenant system (i. e. , COINS hardware arid software
reside in other systems) comprised of three principal elements:
The principal advantages of the configuration are that it has
proved to be cost effective, and is now operational, with analysts
being familiar with its use.
The configuration is limited, however, by the fact it operates
at a single security level (TS/SI) thereby limiting the information
available to users. Additionally, massive files, such as AEGIS,
(a CIA textual. file) cannot be accommodated.
A summary of the pro-rata and unique cost (FY 73) for all
agencies for COINS is contained in the following table. Complete cost
breakdowns are contained in Appendix II.
25X1
25X1
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 Approved For Release 2005/05/23 : CIA-RDP82M005314M00400010019-4
2. Alternative Configurations - Four broad categories were con-
sidered.
a. Maintain the present configuration and upgrade the
system security level.
b. Eliminate COINS and satisfy existing and future inter-
agency exchange requirements either through bilateral,
direct terminal access arrangements, -or by reverting
to off-line only exchange.
c. Upgrade COINS to eliminate major weakrnesses.
d. Replace COINS with central information storage
and retrieval system.
Alternative 1 - Change COINS Security "Umbrella"
There are three sub-alternatives presented in Attachment A:
(a) upgrade to TK, (b) upgrade to TK and Gamma, and (c) upgrade to
complete multi-level security. This summary discusses only the
first. The system itself would remain as previously, but the
"umbrella" under which COINS operates wild be changed from. SI
to SI/TK. All terminal areas would be physically secured. to meet
TK requirements.
The principal advantage of this modification lies in the immediate
availability of TK data through a purely administrative mechanism. A
disadvantage is in the additional requirement for people cleared to the
TK level.
The cost for the conversion has been estimated to be between
Alternative 2-- Eliminate COINS
(a) Revert to Off-Line Exchange.
The only justification for this alternative would be an over-
riding need to realize whatever resource savings might be achieved
from outright abolishment of COINS. The maximum theoretical savings
achievable would amount tol That saving would be reduced
by the amount of resources necessarily rededicated to the production
of hard copy reports and the reinstitution of electrical summaries.
Approved For Release 2005 5E/Dpi El -RDP82M00531 R000400010019-4
 Approved For W ease 2005/05/23 : CIA-RDP82M005310040001001I9_ 3
The disadvantages of this option include the loss of a proven
asset and the retrograde re-substitution of less efficient methods of
information exchange.
(b) Exchange information through direct terminal/system
connections.
The advantages of this option lie in the simplicity of the
access. The major disadvantage lies in the fact that costs will grow
geometrically with the addition of new agencies or processors. The
total annual cost for institution of this option is estimated
per year.
Alternative 3_- Upgrade COINS to eliminate major weaknesses.
The principal advantages of concepts are that each agency would
maintain autonomy, the switch would be eliminated, and all current
hardware and software remain useable. The major disadvantage is
that multiple retrieval languages remain necessary. The cost of im-
nlementinm an ARPA type system is estimated to be approximately
Alternative 4-.- Central Community Information Storage and Retrieval
System.
In this configuration, a single centralized processor is used
to access all community files. Each user accesses the files through
an encrypted link.
Advantages of the approach include commonality of retrieval
and file maintenance languages. Data standards become mandatory,
easing interagency liaison problems, and facilitiating growth capa-
bilities.
25X1
25X1
Approved For Release 2005/05/2 C1A~;Rtj82M00531R000400010019-4
 ,EGRET
Approved For Wease 2005/05/23 : CIA-RDP82M00531W00400010019-4
The principal disadvantage is that the size and the design of
the system would render it difficult to change.
25X1
One of the principal tasks of the Review Group was the collection,
identification, and cataloging of file information, and the analysis of
the Washington area files for indications as to their content and. the
reactions of the community to their interchange. These reactions were
expressed through statistics of the community use of files presently
available through COINS, and through user agency expressions of
interest in on-line access to other agency files.
A description of all files was presented for analysis by each
agency. Information contained in these disclosures included: host
agency, security classification, geographic coverage, subject
category and means of access, as well as descriptive narrative
about the file itself. A major result of this effort has been in the
compilation of this information, contained in Attaclun,?nt A.
SECRET
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 SECRET
Approved For Wease 2005/05/23 : CIA-RDP82M00531W00400010019-4
V. CONCLUSIONS
A. Requirement for Automated Information Exchange
The general conclusion regarding the exchange of information on-
line is that a requirement presently exists and the requirement will
significantly increase in the future. This conclusion is supported by
a number of observations.
1. There is an increasing dissatisfaction within the community
with manual files generally. There is a correspondingly growing
awareness within agencies of the existence of files on-line in other
agencies for which access would be desirable and beneficial.
2. There are large bodies of information presently collected
which can only be disseminated most efficiently by electrical means.
Special processing centers are increasingly maintaining such infor-
mation on-line, and the prognosis is for rapid expansion. Examples
are air movements, naval movements, photography readouts, etc.
3. When one agency maintains information on-line, the most,
and in some instances the only efficient means of access, is on-
line. For examtnle, the entire air movements base was put on-
line by NSA under USIE direction.
4. The user agencies in the Washington area are expending
resources to participate in the COINS experiment and additional
organizations both inside and outside of the Washington area
have expressed a requirement to participate.
5. The user agencies represented on this Review Group have
expressed high interest in on-line access to a total of 54 existing
files of which 12 are presently available through COINS. There
was a variable degree of interest among the agencies in particular
files.
6. Available data show that more than one third'" of the COINS
interrogations represent one organization interrogating the files of
another.
The exact percentage cannot be determined at this time due to statistical
inadequacies. The DIAOLS system files may be accessed from terminals
in either a batch mode or interactively, at the option of the interrogator. In-
terrogations through the network switch may only access the files in a batch
mode. DIA does not consider interactive access as part of the COINS system
and has not provided t}, a means to count such interrogations. P Therefore, the
per cer~trovaud cFordF efease Ra1~~1~Abd8ii 6G46( ' 4f( V' mode.
pp
S~' P
 SECR~j
A
pproved For gjease 2005/05/23 : CIA82M00531WO0400010019-4
B. COINS Effectiveness
COINS has demonstrated that on-line data base exchange among
community users is feasible and achievable. This conclusion is sup-
ported by the following observations:
1. COINS now provide:, operational support to community users
and has a potential for even greater operational utility, provided
present limitations are overcome. Additionally, it has developed
experience for the community in dealing with such problems as:
a. Netting together of diverse computer systems.
b. Management of diverse networks.
c. Security
d. Data standards
e. Retrieval languages
f. User education and support
2. An organizational structure providing the mechanism for
the handling of present and future problems of interagency e:x-
change has been evolved.
3. There is general satisfaction with the ability of COINS to
provide timely responses to interrogations. Median response
time of less than 20 minutes, with an average response time of
less than one hour is now the rule. This is considered timely
for batch mode interrogations.
4. Analysts have been provided a manipulative capability
which has helped to enhance their effectiveness, and has resulted
in an overall manpower savings.
5. A number of daily electrical end product reports have been
replaced by on-line COINS access tailored to consumer require-
ments, at the same time relieving a burden on communication
facilities.
Approved For Release 2005/0 0 3,F DP82M00531 R000400010019-4
 SECRET
Approved For IJease 2005/05/23 : CIA-RDP82M00531$p'00400010019-4
V-3
6. The above benefits have been achieved at a minimal cost
through the use of existing hardware and software. Very little of
the FY 73 operating cost is unique to COINS
C. COINS Weaknesses
The study has revealed notable shortcomings in the present COINS ap-
proach to information exchange. Some of these are management-oriented,
and others purely technical.
1. The assignment of a single agency as the executive agency for
the COINS experiment has not provided sufficient management in
authority to deal effectively with relevant problems of conflict be-
tween agencies regarding roles and missions with regard to intelli-
gence collection, processing, dissemination and production.
2. Security and compartmentation is a critical issue. Signifi-
cant information is now excluded because of security restrictions.
Without a solution to this problem, COINS or any other general in-
formation exchange network will be of limited value. Solutions to this
problem have been proposed and the Computer Security Subcommittee
has them under advisement.
3. The community is not satisfied with the selection of files
available through COINS. In this connection it must be stressed,
however, that the decision to place any file on COINS rests wholly
with the sponsoring agency. Efforts being made to change CO:[NS
files at present must deal on a relatively low priority basis with
other agency problems. In the last analysis, COINS is only as
good as the files which may be accessed through it.
4. File maintenance and updating needs improvement. Agencies
have maintained only those files which ate of internal operational
value, and have also resisted external pressures to alter data for-
mats and query systems, or to include more detailed data than they
themselves need.
5. Education, documentation, and training of users and potential
users are problems of continuing concern because of' the changes in
the user population, the addition or modification of files, and changes
in service.
25X1
Approved For Release 2005SLIRDP82M00531R000400010019-4
 Approved For Wease 2005/ 5/23-DP82M00531$Q00400010019-4
6. The present COINS configuration limits its capability for
expansion to satisfy future interagency access requirements.
a. The system is not interactive, which limits its useful-
ness for certain types of data.
b. Reliance on the switch is inconvenient. Closure at night, or
over week-ends and holidays can prevent access to certain needed files.
Switch failure can also be critical.
Approved For Releas 0 i~: CIA-RDP82M00531 8000400010019-4
 8` ECG 'ET
Approved For Release 20055/23 : CIA-RDP82M00531QP00400010019-4
VI-1
VI. RECOMMENDATIONS
Based on the results of this study, the Review Group recommends
the following:
1. The DCI direct that an aggressive program of automated
information exchange be undertaken.
2. To implement the above, a majority of the Review Group pro-
poses that the ASD (I) be appointed the COINS executive agent under
the auspices of the DCI with the COINS Project Management Office
(PMO) made responsible to the executive agency.
3. COINS funding be continued, subject to stipulations included
below.
4. The USIB Security Committee and its Computer Security
Subcommittee address the multi-level security problem with an eye
toward immediate interim solution. This Review Group suggests
that pending the attainment of a. true multi-level security capability,
the most practical course of action is:
a. to upgrade the entire COINS system to TK;
b. to handle GAMMA and. other cornpartm-ented data on a
case-by-case basis with dedicated lines and terminals;
c. to handle extremely time--sensitive event data on a
case-by-case basis with dedicated lines and terminals.
5. The COINS data base be purged of files for which little or
no interest has been demonstrated, and other available non-COINS
machine files for which interest has been expressed should be
included in COINS as quickly as possible. A complete listing of
file suggestions, including additions, modifications, and new files
to be added is included as Attachment B.
6. -Present COINS reporting procedures be modified to allow
the collection of statistics more amenable to permitting evaluation
of system usage, timeliness, and effectiveness. The COINS project
manager should be directed to submit a plan for statistical reporting
.
Approved For Release 2005/0 ty y R p P82M00531 R000400010019-4
 Approved For Fgjease 2005/05/23 CIA-RDP82M00531QP00400010019-4
which would (a) identify the objectives of such reporting (i. e. what
must be learned about the system), (b) the items of data to be col-
lected to satisfy the objectives, and (c) the analysis to be performed
on the data to provide the desired information.
7. The COINS Project Manager and each participating organi-
zation take appropriate steps to accelerate and expand the program
to train, inform, and motivate both users and potential users.
8. The COINS PMO submit a detailed plan for an upgraded
system which will address the weaknesses noted in Section V in
a cost-effective manner. Specifically to be addressed are means
to eliminate the switch, provide interactive capability to all users,
and the inclusion of world-wide user hosts and participants.
Approved For Release 200 3 x ADP82M00531 R000400010019-4
 PRUNE [
Approved For RjWease 2005/05/23: CIA-RDP82M00531PA$00400010019-4
ATTACHMENT A - STUDY RESULTS
This attachment provides a summary of the information gathered
during the course of the study. No attempt is made here to interpret
these results or to draw conclusions from them.
The results presented cover two general topics:
A. System Descriptions - current system and alternatives,
including advantages, disadvantages, limitations, and costs.
B. Files - summary of the data on Washington area files,,
including COINS and non-COINS files, with comments regarding use and
interest.
A. ystem Descriptions
1. Current stem
.25X1
It should be noted that there is some disagreement, even at this
date, as to a proper definition of the COINS system. The disagreement
stems mainly from the fact that COINS was implemented using existing
hardware, software, and communications facilities. In fact, most of
the hardwart,, software, and the files were already available for users
prior to the existence of COINS (i. e. , some degree of dissemination of
file data, on a direct terminal access basis, was occurring, and this
dissemination mode was absorbed into COINS).
The advantages of the configuration of Fig A-1 are:
a. The system is operational and provides modest operational
support.
Approved For Release 2005/ ejTDP82MOOS3l R000400010019-4
 25X1 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 SECRET
Approved For Rlease 2005/05/23 : CIA-RDP82M00531WO0400010019-4
b. Sizable fixed cost investment, particularly in software,
and procedures continue to be capitalized.
c. Analysts have been given initial training of the system.
d. Systems analysts and technical personnel are thoroughly
familiar with the system and operational performance is still improving.
The disadvantages of the configuration are:
a. Technology used in implementing the present COINS is,
in some cases, obsolete.
b. Two query languages must be learned to use all files.
c. Only formatted files may be handled.
d. Operates at a single security level (TS/SI) and as a
result, not all information is available to users.
The configuration limit.: the potential of interagency exchange,
a. Massive filesi all source
textual file at CIA, cannot be accommodated should sucn on-line access
be required.
25X1
b. Interactive and time-sharing operations are not possible. .
Present procedures for recording or reporting rra nagement
information do not readily lend themselves toward the pro-rata
assignment of costs. In order to derive these costs, it proved necessary
to review the program elements relating to COINS, and estimate the
percentage bf these costs which are attributable to COINS operations.
A summary of the pro-rata and unique cost (FY 73) for all agencies
for COINS is contained in the fQllpwing table. Complete cost breakdowns are
contained in Appendix II.
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
SECRET
 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
2. Alternative Configurations - Four broad categories are
a. Maintain the present configuration and upgrade the
system security level.
b. Eliminate COINS and satisfy existing and future
interagency exchange requirements through bilateral, direct terminal
access arrangements.
c. Upgrade COINS to eliminate major weaknesses.
d. Replace COLTS with central information storage
and retrieval system.
The alternatives and the cost implications are summarized in what
follows.
Alternative I_-- Change COINS Security "Umbrella"
There are three subalternatives here: (a) upgrade to
TK, (b) upgrade to TK and D/G, and (c) upgrade to complete multi-
level security.
(a) Upgrade to TK
Description: The system itself would remain as previously
The "umbrella" under which COINS operates would be changed from SI
to SI/TK. All terminal areas would be physically secured to meet TK
requirements. All personnel having access would be cleared for TK
access as well as SI. System control of information would be to the terminal
level. Control beyond that would be by conventional procedures.
Approved For Release 2005/05/23 4 CIA-RDP82M00531 R000400010019-4
25X1
R
 Approved For# Jease 2005/05/23 : CIA-RDP82M00531 000400010019-4
Advanta. es:
(1) Immediate addition of general interest files from a
special processing center would be possible.
(2) Action to accomplish is totally administrative. Impact
on present users would be minimal.
Disadvantages:
(1) Some present users might require TK clearances even
though they do not use TK information, simply to access.
(2) Non-COINS members (e. g. , PACOM, CONARD) now
accessing COINS files would be required to secure similarly, or IDHSC
and COINS could not share facilities.
Growth Potential: Additional files and users can use the system.
(b) Upgrade to TK and Gamma.
Description: The system itself would remain as above,
except that the umbrella under which COINS functions would-be changed
to include TK and Gamma as well as SI.
Advantages: Generally as indicated above.
Disadvantages: The potential users, probably. constitute a
relatively small group yet all users would require clearance, even
though they do not use Gamma.
(1) In order to raise the security level of COINS to TK and
Gamma it would be necessary to upgrade all terminal facilities to that level.
25X1
Approved For Release 2005/05/23.: .Ci. DP82M00531 R000400010019-4
4
 %)4VIIL I
Approved For lease 2005/05/23 : CIA-RDP82M00534000400010019-4
This is somewhat in conflict with the philosophy of limited access which
lies behind such compart.rnentation. Effectively, Gamma would be
eliminated as a compartment.
(c) Upgrade to a multi-level security system
Description: A data system in which hardware and software
can protect and segregate information of various security levels to a
degree of reliability required by security regulations. Operationally,
the system would provide data only to the level of ''clearance'' of the
terminal and/or user.
Advantages: The system would be available for use universally,
without security compromise.
Disadvantages: The research and development for such a
system would require anumb;.r of years, variously estimated at three
to seven, depending upon resource level invested.
Limitations: No improvement in the present situation would
be available for some time.
25X1
25X1
Approved For Release 2005/05(O) EDP82M00531 8000400010019-4
5
 Approved For Rgiease 2005/05/23 : CIA-RDP82M00531Q00400010019-4
Alternative 2 - Eliminate COINS
Under this alternative, it is possible to envisage at least two
basic sub-alternatives. The first would encompass not only the
abolishment of COINS but the cessation of all on-line exchange of
information hitherto exchanges through COINS and a return to off-line
methods (e. g. , hard-copy exchange, electrical dissemination of
summary types reports or indirect access to data bases via "opscom"
links). This first sub-alternative presupposes a prior community
judgment that there does not exist any credible community requirement
for on-line interagency file access. The second sub-alternative would
involve the substitution, for the present COINS, of a series of
bilateral interagency terminal links characterized, for purposes of
convenience, a terminal swap. This second alternative clearly
presupposes a need for some continued on-line interagency file access
of a type and magnitude at least as great as that which COINS now
accommodates.
j2 ion,A - Abolish COINS/Revert to Off-Line Only Exchange
Advanta fle s
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
25X1
o .VREt
 Approved For lease 2005/05/23 : CIA-RDP82M00534&000400010019-4
Disadvantages
By contrast with very small, cost savings that might be gained
from abolishing COINS and ceasing further on-line exchange of
information within the Community, the disadvantages associated with
such a course are considerable,
1. The loss of a substantial body of experience in the on-line
exchange of data. and the loss of an'asset cf proven worth. for accomplishing
such exchange.
2. A retrograde resub:stitutiori of inherently less efficient
methods of information exchange, especially with respect to those bodies
of intelligence which, in terms of their volume and character (discrete
elements of forxnattable (lata), most readily lend themselves to exchange
in digital automated form and which, in terms of time sensitivity, are
most effectively provided through on-line systems. Air movements
are, An this respect, only the most conspicuous examples to date of this
genre of information.
3. The loss, for the indefinite future, of the only instrumentality
available to whole community for further experimentation with file
access on a community as distinct from a purely agency-to-agency basis.
25X1
") cOEI
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
7
 Approved For Release 2005/05/23 : CIA-RDP82M00531GO00400010019-4
Option L' - Eliminate COINS. Exchan e information through direct
terminal/system connections.
Definition: Diagrams for this alternative are shown in Figure A-2.
Commu icaLtion links, terminals, and crypto equipment are shown in the
diagram. The numbers of terminals, lines and other equipment are
estimated from information supplied by each of the agencies.
O rating Procedures and Assumptions:
(1) A system which services only present COINS files is required.
(2) Each agency would make bilateral agreements with another
agency to share data. Generally, the agency would provide a compatible
terminal, a communications line and the suitable crypto gear, (2 units)
necessary to interface with the host system as if it were an internal
station. There would be little interaction with the community as a whole.
The agency interfacing with the host computer would be required to follow
the same operating procedures as any other station connected to that
system.
(4) The capabilities provided at the terminals will be constrained
by the capabilities of the host system and any other agreements reached
by the two agencies.
(5) Funding for this approach would be provided out of each of
'the agencies'' operating budgets.
(6) Security procedures would be worked out between the two
participating agencies.
(7) Couriers would deliver long listings in order to conserve
system resources.
25X1
82M00531 R000400010019-4
Approved For Release 2005/05/2S W
- ~-
8
 25X1 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 VJq VI 4L'w il.
Approved For'`PROease 2005/05/23: CIA-RDP82M00531&600400010019-4
Advantages
(1) A relatively simple communications system.
(2)
Eases interagency liaison problems.
(3) Reduces community investment in developing file processors.
(4) Provides direct access to the desired system.
(5) Provides a better means of coping with the multi-level
security problem.
(6) Provides a basis for resource sharing if interagency
arrangements are satisfactory.
Disadvantages:
(1) Provides no impetus to.=,wards commonality (data standards,
languages, etc.
(2) Although initial costs are low, they grow geometrically
with the addition of new agencies processors.
(3)
Investment in working software is lost.
(4) Mix of files may dictate major changes in cost.
(5) Unable to take advantage of remote terminals already in place.
(6) Does not anticipate legitimate needs for computer-to-
computer communications.
(7) Does not provide a positive atmosphere for interagency
sharing of information.
Limitations /Growth Potential.
This approach to providing access to agencies' files is severely
limited in its growth potential. The addition of new systems or agencies
into the community of users would cause system costs to expand rapidly.
Approved For Release 2005/05/23 CIA-RDP82M00531 R000400010019-4
 25X1 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Next 1 Page(s) In Document Exempt
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 JLt'I I.
Approved For RgIease 2005/05/23 : CIA-RDP82M0053i&D00400010019-4
Alternative 3 - Upgrade COINS to eliminate major weaknesses.
There are several concepts which may be proposed that offer
maximum utilization of existing COINS hardware and software, while
at the same time significantly upgrading the COINS system. The concepts
may be differentiated on the basis of whether they present near-or far-
term configurations, and the extent to which they require additional
hardware and software development. The one presented in this section
utilizes a concept developed by ARPA. Another proposal is found in
Appendix III.
(1) Each participating aagency maintains absolute and full.
autonomy.
(2) The COINS Switch i eliminated.
(3) All current hardware and software (with exception of Switch
support) is useable.
(5) Does not require highly sophisticated or complicated crypto-
equipment (no multi-variable storage and control).
(6) Little impact on current network operation.
:('7) Provides relative easy key variable changes on a circuit-
by-circuit basis with essentially no system ''downtime. "
(8) Where deemed necessary each end of the line (Iiost/IMP
cable) between the Host computer and secure IMP could have crypto-
units. Thus, the message content would be secured on a Host-to-Host
basis while the header of the message would be secured on an IMP-TO-
IMP bads.
25X1
Approved For Release 2005/OgFDP82M00531R000400010019-4
12
 Approved Fort lease 2005/05/23: CIA-RDP82M0053 0004000100t$g4lre A-3
COINS II PI-LASE I
DISTRIJ,UTED NIETAVORK
(Utilizing ARPA Concept)
DIA/ IDHS
Computer System
CONAD
PACOM Network
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 25X1 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Next 1 Page(s) In Document Exempt
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 Approved For W ease 2005/05/23: CIA-RDP82M00531WO0400010019-4
(6) Modification of the Host's executive software (again, cost depends
upon the particular I-lost, typically, 4 to 12 man-months of system
programming required).
Recurring Costs
25X1
Alternative 4 - Central Community Information Storage and
RetrievalSyste:-r
Definition: A diagram of the network is shown it Figure A-4.
The informatics, Inc. , study of 1970 set forth an estimate of 50
terminals (initially) growing to 200 over a period of 5 years. Estimates
are based on ZOO terminals located in five agencies.
Oe.ratiProcedures and Assumptions:
(1) One agency (independent or participant) would be designated
to install, manage, maintain, and operate the CentralCommunity
Information Storage and Retrieval (CCIS&R) System.
(2) The CCIS&R System will provide the following capabilities:
(a) Remote batch (query and maintenance),
(b) On-line, interactive query capability.
(c) On-line, maintenance.
SECRET
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
15
 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Next 1 Page(s) In Document Exempt
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 Approved For Tease 2005/05/23 : CIA-RDP82M00534 00400010019-4
(8) Training courses will be set up to info'-m and indoctrinate
Advantages:
(1) A common retrieval language; a common file maintenance
language.
(2) Data standards are mandatory.
(3) Common communication and communications security
equipment.
(4) A new user could join the system at little cost.
(5) A relatively simple communications system.
(6) Eases interagency liaison problems.
(7) Reduces community investment in developing several
file processors.
Disadvantages:
(1) Does not anticipate or allow for legitimate needs for
computer-to-computer commun-dcations.
(2)
(3)
May be at odds with DLA-IDHS plans.
One agency imposes many technical decisions on others.
(4) Service under crisis conditions may not be acceptable.
(5) Size and design of system would probably render it very
difficult- -technically and politically-to change.
(6) Services of common concern have (historically) received
.,.their budgets through line office planning rather than through the
relatively unwieldy actions of USIB/IRAC committees.
(7) Enforcement of file maintenance may be impossible, i. e. ,
value of files may deteriorate. .
Approved For Release 2005/0512- .3 GI A~RDP82M00531 R000400010019-4
 Approved For F,jease 2005/05/23 : CIA-RDP82M00531QA00400010019-4
(8) Establishment of an agency's file at the central facility would
probably not eliminate the need to also keep the file locally. Therefore,
in the worst case, every file on the central computer would have its
counterpart on some other computer system--with all that implies.
(9) Specifications for the software system might be extremely
difficult to establish since the system would have to satisfy all user
requirements. Alternatively, it would be necessary to compromise,
and the result would be disaffection and dissatisfaction on the part of
many users.
Limitations /Growth Potential: If the software system for a
central complex is properly designed, growth--presumedly to accommodate
more users--is simply a matter of adding hardware, storage devices,
and retrofitting the executive program. There is nothing to prevent
(i. e. , the state-of-the-art permits) this central complex from becoming
a node of a larger network.
Estimated Cost of A Central Comm-lun Informz!,taon Storage
and Retrieval System-
(Assume s 5 year design and development schedule)
18
Approved For Release 2005/05/S3E EY82M00531R000400010019-4
25X1
 Approved For Release 2005/05/23 : CIA-RDP82M005314WO400010019-4
B. Files
For the purpose of providing a consistent set of around rules
for the analysis of file data, the following definitions were established:
(1) Files - All pertinent ~'~ ashin,ton area automated files,
whether or not presently cart of COI tiS and whether or not presently
accessible on line. The files considered are contained in Tables C-1
and C-2, and NC-1 throwh -NC-o. A list of those files not considered
is also provided (Tables F-l, 2 and 3).
(2) Community Use - Defined as present on line usage by one
or more agencies, other than the file host. For non-COINS files, off-
line dissemination is therefore classified as no community use. For
those nor.-COINS files where some present on-line usage has been noted,
the stated degree of this use (low, moderate, or high) represents an
educated guess on the part of the file host. The degree of use of the
present COINS files comes from data provided by the COINS program
i_:anage r.
(3) Community- interest - The present levels of available
information is insufficient. to Ticrlnit accurate definition of interagency
requirements for on-line exchange on a file-by-tile basis. Accordingly,
for this r ev ie%v each user a ency expressed an opinior as to the derce
of agency interest in on-line access to each file. This expression of
interest is intended to serve as a TDoint-of-deparutre for an eventual
definitive statement of need.
25X1
(5 Redundancy-- The positive identification o.f redundant or
overlapping files will require a detailed analysis of all file data elements.
This attachment identifies and flags those files which give outward
appearances of containing information possibly redundant with that of
other files. These files are shown in Tables R-1 and R-2.
Approved For Release 2005/05/25-.CiA-RDP82M00531 R000400010019-4
 Approved Forlease 2005/05/23 : CIA-RDP82M00534W00400010019-4
(6) Priority - To attempt some correlation between use, interest
and importance of community access to a particular file, the range
of priorities for information contained in each file from the applicable
geographic area was taken fror-i the intelligence objectives and priorities
as stated in DCID 1/2 and JSOF FY 75-82, Annex A. An asterisk
is used to indicate a special category, where the file coverage crosses
many intelligence objectives and many geographical areas, and a single
priority cannot be determined.
The notation for community use and interest is:
N: None
L: Low
M: Moderate
II: High
? : Unknown or unable to determine
The e pressi.ons of interest represent estimates, its discussed
previously. For files which are presently accessed through COINS,
statistical data are available regarding file: use. These numerical
data have been converted to the above format by dividing the October
1972 use counts into quartiles.
The notation for priority ratings is:
1-8: High through Low DCID/JSOP ratings.
--: None stated (no priority)
Special (see paragraph 6)
? Unknown or unable to.determine.
Use, Interest and Priority selections for the table have been
chosen as the highest of the range presented by the different agencies.
With regard to the file redundancy indications of Tables R.-1 and
R-2, two of the areas in the table were selected for further examination in
greater detail - Ground Forces files and Installations files... The..
evaluation was performed by the file sponsors on the basis of a review
of the following data on each file:
Approved For Release 2005/05/23 : OA-RDP82M00531 R000400010019-4
SECRE
 25X1 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 25X1
Approved Forlease 2005/05/23 : CIA-RDP82M00500400010019-4
Review of the installations files led to the following notes:
1. Unwarranted duplication among location /installation files
within the corrnmunity was found to be slight:. ;' hile geo;ranhic areas
of interest overlap (4 out of 5 files are world.*iide), the files are
significantly different based on purpose and use, structure, and
classification level. The periods of coverage can be segregated by
those files (or portions) that date back to the l9-1O s, and those whose
date level begins in 1968 or 1969. A file u?date trend shows the
movement toward daily or as required after an event or mission.
recommended.
It is felt that the location/installation file differences outweigh the
similarities; therefore, a data element by element comparison is not
~e-r1 r
Approved For Release 2005/05f23~ : CIA Lk-bP82M00531 8000400010019-4
25X1
 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Next 21 Page(s) In Document Exempt
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 Approved Forl ease 2005/05/23 : CIA-RDP82M00534W00400010019-4
A:RPE`:DI>
Approved For Release 4~ ffrcItPP82MOOS3I R000400010019-4
 f b11urlL'
Approved For Rase 2005/05/23 : CIA-RDP82M00531400400010019-4
FOR STUDY ON AUTOMATED INFORMATION EXCHANGE
AMONG INTELLIGENCE AGENCIES IN TIIE WASIIINUiTON AREA
Objectives:
1. To review and evaluate, by 1 February 1973, the COINS system,
determining its effectiveness in meeting the present and anticipated require-
ments of the user community.
2. To identify those areas where the effectiveness of COINS in
improving the intelligence product may be enhanced.
3. To recommend courses of action and consider their impact on
present and future budget allocations.
Phase 1. Collect, identify and catalog the present and anLicipa.tecl auto-
mated files of each agency.
Describe the general nature and content of each file
with reference to-
(1) Countries covered
(2) Data element:s
(3) Security levels
b. Determine which of the files are presently part of COINS.
To be completed by the meeting of 22 November 1972.
Phase 2. Identify those files that are required by users in more than one
agency. Catalog which of the required files are presently part of COINS,
and evaluate the ability of COINS to satisfy user requirements through
access to available files. Identify new files that might be of value in
meeting common user requirements. Determine the extent and importance
of the use on a file-by-file basis of inter-agency access to an agency's
files.
To be completed by the meeting of 7 December 1972.
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
rMIN 11
C U141
 Approved For Release 2005/05/23: CIA-RDP82M00531iiD00400010019-4
7.
Phase 3. Determine the effectiveness of COINS in satisfying the require-
ments of the user community by examining the utilization of the present
prototype system. Identify the utilization problem areas through studies
of:
a. Frequencies of queries by file and by user.
b. Adequacy of responses to queries.
c. Utility of responses in format and timeliness.
d. Ultimate disposition of responses.
e. Multi-level security.
To be completed by the meeting of 28 December 1972.
Phase 4. Determine the usefulness of the COINS approach in satisfying
multi-agency information exchange requirements and identify the costs
in both manpower and hardware resources currently committed to COINS.
Recommend improvements or alternatives which appear beneficial from
the points of view of
a. Satisfaction of user requirements.
b. Enhancement of rrul.ti-agency exchange ca pabiliti.es.
c. Cost savings.
To be completed for review by the meeting of 18 January 1973.
Final. report to be presented 1 February 1973.
Approved For Releas " Eft-RDP82M00531 R000400010019-4
 Approved For F3g1ease 2005/05/23 : CIA-RDP82M00531W00400010019-4
APP?E'\ DD< II
Approved For Release 2
6113 EQIA-RDP82M00531 R000400010019-4
 25X1 Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
Next 2 Page(s) In Document Exempt
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 Approved For Please 2005/05/23 : CIA-RDP82M00531W00400010019-4
Approved For Release 2005/05/23 : CIA-RDP82M00531 R000400010019-4
 SECRET
Approved For Ralease 2005/05/23 : CIA-RDP82M00531QR00400010019-4
COINS II PROPOSAL
Approved For Release 2009rUX CIA
-RDP82M00531 R000400010019-4
 Approved For f*&Jease 2005/05/23: CIA-RDP82M005314000400010019-4
APPENDIX III
C TN5 _T
PROBLEM
1. respond to crisis or time sensitive situations.
2. correlate and analyz o information from a number of different
sources in a timely fashion.
B. This situation is made even more acute today by the fact that each
agency must strip: e or sel:: suf.:c:enc; ay ... lcai. and ma,:7tainir ; its o:.~i
intelli