Attachment
RECOMMENDATION
TO THE AIR STAFF
DEVELOPNiENT OF THE DISTRIBUTED
AD ]%lESSAGE-BLOCK NETWORK
August 30, 196
------------------------------<page break>-----------------------------
Attachments:
1. Rand recommendation.
2. Letter from the Air Force.
3. Chapter by Paul Baran from recent book on packet switching that
discusses rationale for each feature of packet switchln. and how it
: :'. -
came about.
4. P-1995.
5o Calendar pages.
6. IEEE Spectrum Aug. 1964, p. 114
7. List of RAND depository libraries.
52
------------------------------<page break>-----------------------------
R-11 rl D
30 August 1965
L-I7930
.DeputyChief of Staff, Research and Dvelopment
Headquarters, United States Air Force
%A'ashington, D.C. 20330
SUBJECT: DEVELOPfENT OF THE DISTRIBUTED ADAPTIVE MESSAGE-BLOCK
NETWORK
S If ?, HA RY
-e have reached an appropriate point in our work on "distributed netw'orks-
to recommend to the Air Fore a specific direction of development.
I wrote to you on this subject.on August 24, 1964 (L-17082) and General
Ourtin forwarded a request on December 23, 1964 for evaluation of the
eleven RAND Memoranda by AFSG (and RTD, ESD, Lincoln), AFLC, (and
GEEIA), and AFCS. Detailed comments were received and forv, arded to
RAND. A series of briefings and discussions ensued. Some technical
difficulties were revealed, but these appear to be correctable v,.th minor
changes in the proposed system. A RAND !,/[emorandum describing the:-;e
adjustments is in preparation.
In our view it is possible to build a new large common-user communicatzoa
network able to withstand heavy damage. Such a network can provide a
major step forward in Air Force military communication capability. Although
there can be no certainty of success because of the complex nature of the
required development, the high potential pay-off leads us to recommend
that the research and development portions of the following program be
undertaken now. 'e can now purchase an option on lead time at minimum
cost. If the research is successful, we vil! be in a position to move
rapidly Into implementation of the full net;:.ork. Only after the prelim'nary
engineering has been completed will there be sufficient information to
determine precisely where and how much of such a networ should be built.
------------------------------<page break>-----------------------------
Deputy Chief of Stmff,
Research and Development
-2- 30 August 1965
Although the system is a marked departure from the existing communication
system, we believe that after the R and D is proved out the new system can
be integrated with the present one on an incremental basis.
¾Ve will be happy to help in any way as the program evolves_. A further
discussion of the problem and the proposed solution is atta7ched.
Sincerely,
ORIGIN'At SlG.';KD
F. g. COLLC'.:M
F. R. Collbohm
Attachment: Limitations of Present Networks and the Improvements \¾e Seek
------------------------------<page break>-----------------------------
ATTACHMENT
LI'vIITATIQNS OF PRESENT NETWORKS AND THE IMPROVEMENTS WE SEEK
Most communication networks are highly vulnerable
to overt and covert attack. _A... he.w, la. rge, highl..-
Interconnected distributed network, shared am0[g:_--
all military users would offer a major improvement -
in survivability. (See OD I.)*
Most military communication is transmitted with
negligible protection against loss of critical infor-
mation to an eavesdropper. Only a small portion
of the total millta W communication load is crypro-
graphically protected. The mihtary is often highly
constrained in subjects that it can discuss over the
telephone for fear of eavesdropping. The universal
availability of secure communlcations will be help-
ful to those having need to transmit classified infor-
mation. (See ODC IX.)
Today, those who connect computers to other comput-
ers and to the remote entry devices uœed. in command
and control systems are not satisfied vith the high
error rate of a communication neW;ork which was not
originally designed for digital transmission. Com-
puters can be intolerant of errors caused by the com-
munication links. The use of automatic error correc-
tion features could help achieve the desired, almost
error-free user-to-user performance--less than one
error in 100 million transmitted bits. (See ¸DC VI
through VIII. )
Partially because of high cost, military "hard copy"
(paper) traffic is primarily limited to center-to-center
operation. This produces long delay times between end
users. A system that will reduce the cost of digital
transmission will help make hard copy facilities more
widespread at the action officer level. (See ¸DC V
and VIII. )
ODC Is an abbreviation for "On Distributed Communications,
the series of memoranda described in Appendix.
------------------------------<page break>-----------------------------
Attachment - page two
I.-i79300 8/30/(:5
V¾'e lack complete flexibility of connection between
users where secrecy is mandatory. W'e desire a system
in which it will be possible to speak quickly to anyone,
anywhere without prior arrangement for the communica-
tions. (See OI:%S VIII and IX.)
Military communication volume is increasing ra._'idly,
and greater efficiency would result from bet:er?_se_r-to-
user communications. %Ve want a system bet:e able
ß to handle larger volumes of traffic than commonly expe-
rienced today. Better military effectiveness could result
if communications were not treated as a scarce resource
that must be highly rationed. (See ¸De V and VIII.)
Because of the difficulty in assigning costs t.o commu-
nications in the military, it is difficult to determine
exactly how much we are presently paying. Estimates
range from one-half to three billion dollars annually,
with this cost increasing about t5 percent annually.
In the belief that markedly more military communication
is indicated in the future, it becomes mandatory to lay
the groundwork upon which to build future communi'ca-
tion networks that achieve the desired long range demand
for.digital and voice communications at minimum cost.
The maturity of digital technology allows us to think
broadly and differently regarding communications. The
concept of an all digital communications network (as
opposed to the traditional analog communication net-
works) allows the designer and user great freedom in
survivability, flexibility, growth, secure communica-
tions and in handling the expected exponential growth
in the transmission of digital data. (See ODC X.).
HOW THE PROPOSED NETWORK DIF,,-EKS FROM PRESENT DAY NETWORKS
The proposed network is described in detail in the OEX series of RAND
Nemoranda. Some oœ the key differences between [t and the present
concepts of building communication networks are:
Use of digital techniques in lieu of an91.g circuits.
throughout the entire network, both in transmission
and in switchingß
------------------------------<page break>-----------------------------
o
. 6.
_Cryptographic techniques:designed as an integral portion
of the swltchin. mechanism to provide secrecy for all
network users--:hether t.ransmltting clsslfied traffic
or not.
The communication net-work desi.9'ned as = "bit-transportation
_system_, "_rather than a "real-time" connection of.'.-series
of links.
Integral automatic error'detection and repeat (vl'a a dif-
ferent route} transmission to allow the use of low cost
a_nd unreliab19 links in the network.
Ada_ptt_ve routing for_hqh user reliability during peak load
periods including enemy attacks.
Pro._.Jps to allow each user to "carry his telephone
number" with him.
SPECIFIC RECO.M MENDATIO NS
V/e recommend that a research and development program be undertaken which
can lead eventually to the construction and implementation of the Distributed
Adaptive %4essage-Block Network described in the referenced RAND
L
We recommend that only the research and development phases be undeaken
at this tlme,wlth further implementation to follow the successful corr..ple:ioa
of the preliminary work phases.
Because of the complex nature of the undertaking and the intricacy of the sub-
systems, an unusual high level and continuit_y of technical expertise in digital
tech.nolo_gy is .required in the development and its system management. %Ve
suggest that the research and development tasks be divided into separate con-
trac'.s, each portion awarded to an organization particularly competent within
its own field. This division by task imposes stringent technical competence
on the system manager. To aid in this complex and technically difficult effort,
we feel that a technical advisory panel should be constituted to review on-going
study and development contracts and insure that necessary technical liaison
and inter-action among the various sub-contractors take place.
Because much of the proposed system is based on digital computer techniques,
or._g_a_niza_:lo.ns with_demonstrated experience in _he com..puter field will probably
brin greater pertinent competence to bear in key aspe.cts of the system, devel_op-
m"-en-than organizations experienced only in traditional analog communication
Drac*,lce s .
------------------------------<page break>-----------------------------
D'EVELOPk'iENT COST AND RATE
While no definite time-money expenditure rates are enumerated, the order
of magnitude of the effort is described in Volume XI of the ODC series.
Possibly five million dollars would be an appropriate amount for the initia-
tion of this development effort.
We are not presuming to outline the detailed sequence of acti_ons that might
be taken. 'Ate have, however, prepared such a list of specif!c tasks to be
performed, in what we believe to be a logical form, and I wi.i:b. happy to
submit it to you for your consideration and action if this would be helpful.
The need for a survtvable, high data-rate, flexible, user-to-user communi-
cations is of overriding importanc. Ve do not know of any comparable
alternative system proposals to attain this capability and we believe that
the Air Force should move swiftly to implement the research and develop-
ment program proposed herein.
The orderly development program outlined offers the promise of a marked
increase in communication capability with minimum risk. If the military
situation makes it necessary, the research and development program could
be accelera:ed. For highest payoff per dollar risked it should be remem-
bered that communication system dvelopment is normally a te'n y.a lead%_
m cy_cle proce.s_s  _S!I__u_ch beer work in this field appears to result fro
a mall nu_mbe,,,r of people working on a proJeq,t for a long eriod of time 
an from aarcje massive crash-elf Or: ar:empt. fng to generate complet_e.
_system__t_thl a shc time. ) f{ l hoped that 'me gui'dane of th su-ge:ed
technical advisory panel will provide for flexibility and redirection of th'.s
program in the event unforeseen problems or new developments'appear.
------------------------------<page break>-----------------------------
APPEN DD<
ON DISTRIBUTED COMMUNICATIONS:
List of Publications in the Series
Introductton to Distributed Communications Networks, Paul Baran,
Riv1-3420-PR. .......
Introduces the system concept and outlines the require--
ments for and design considerations of the distributed
digital data communications network. Considers espe-
cially thee use of redundan. c , as a means of withstandl.ng
hea_vy..e_nemv attacks., A general under}tahding of the
proposal may be obtained by reading this volume and
Vol. XI.
Digital Simulation of Hot-Potato Routing in a Broadband Distributed
Communications Network, Sharla P. Boehm and Paul Baran,
RM-3103-PR.
Describes a computer simulation of the message routing
scheme proposed. The basic routing doctrLne permitted
a network to suffer a large number of breaks, then recon-
stitute itself by r,apidly relearning to make best use of the
surviving links.
Determination of Path-Lengths in a Distributed Network, I. %'. Smith,
Rlvi-3$7 8-PR.
Continues model simulation reported in Vol. II. The pro-
gram was rewritten in a more powerful computer language
allowing examination of larger networks. 4odificatlon of
the routing doctrine by intermittently reducing the input
data rate of local traffic reduced to a low level the num-
ber of message blocks taking excessively long paths.
The level was so low that a deterministic equation was
required in lieu of N4onte Carlo to examine the now rare
event of a long message block ath. The results of both
the simulation and the equation agreed in the area of over-
lapping validity.
------------------------------<page break>-----------------------------
IV.
Priority, Precedence, and Overload, Paul Baran, Kv:-3638-PK
The creation of dynamic 0 flexible priority and prece-
dence structures within a communication system handling
a mixture of traffic with different data rate, urgency, and
importance levels is discussed. The goal chosen is opti-
mum utilization of the communications resource within a
seriously degraded and overloaded network.
History, Alternative Approaches, and Comparisons, Paul"B.ran,
R?, -3097-PR.' '. ..... -
A background paper acknowledging the efforts of people
in many fields working toward the development of large
communications systems where system reliability and
survivability are mandatccy. A consideration of termi-
nology is designed to acquaint the reader with the di-
verse, sometimes conflicting, definitions used. The
evolution of the distributed network is traced, and a
number of earlier hardware proposals are outlined.
Mini-Cost Microwave, Paul Baran, Rk';-3752-PR.
The technical feasibility of constructing an extremely
low-cost, all-digital, X- or K -band microwave relay
u
system, operating at'a multi-megabit per second data
rate, is examined. The use of newly developed vatactor
multipliers permits the design of a miniature, all-solid-
state microwave repeatmr powered by a thermo-electric
converter burning L-P fuel.
Tentative Engineering Specifications and Preliminary Design for a
High-Data-Rate Distributed Network Switching Node, Paul
Baran, RM-3763-PR.
High_-_speed, or "ho-potato," store-and forward mes-
_s_ag_e block r_e_layinq forms the heart of the proposed.
information transmission systems. The Switching Nodes
are the units in which the complex processing takes
place. The node is described in sufficient engineering
detail to estimate the components required. Timing cal-
culations, together with a projected implementation
scheme, provide a strong foundation for the belief that
the construction and use of the node is practical.
------------------------------<page break>-----------------------------
The ¾iultiplexlng Statlon, Paul Baran, RN,'-3764-PR.
A description of the .'.Multiplexing Stations which connect
subscribers to the Switching Nodes. The presentation
is in engineering detail, demonstrating how the network
will simultaneously process traffic from up to 1024 sepa-
rate users sending a mixture of start-stop teletypewriter,
digital voice, and other synchronous signals at various
rates.
Securlty, Secrecy, and Tamper-Free Considerations, Pau!._-aran,
RF.,1-376S-PR .. :'.'.: -_
Considers the security aspects of a system of the type-
proposed, m which secrecy is of paramount importance.
Describes the safeguards to be buil: into the network,
and evaluates the premise that the existence of "spies"
within the supposedly secure system must be anticipated.
Security provisions are ba.sed on the belief that protec-
tion is best obtained by raising the "price" of espied
information to a level which becomes excessive. The
treatment of the subject is itself unclassified.
Cost Estimate, Paul Baran, RM-3766-PR.
A detailed cost estimate for the entire proposed system,
based on an arbitrary network configuration of 400
Swilchlng Nodes, servicing 100,000 simultaneous users
via 200 Multiplexing Stations. Assuming a usable life
of len years, all costs, including operating costs, are
estimated at about $60 - 100,000,000 per year.
Summary Overview, Paul Baran, RM-3767-PR.
Summarizes the system proposal, highlighting the more
important features. Consider s the particular advantages
of the distributed network, and commen:s on disadvantages.
An outline is given of the manner in which future research
aimed at an actual implementation of the network might be
conduc:ed. Together with the introductory volume, it pro-
vides a general description of the entire system concept.
------------------------------<page break>-----------------------------
Attachment #.2.
" " .. DIPARTMENT OF THE AIR ORCE ','. . .
- ". . , "..c- ..... 1 .0CT .i "
,. '"'... ß . '. ' ,. " .' .'?;" ' .L: . ..:'." ß .'""  ' .
',....,.-?.}.:.,.. ... . ß .. ..:.;..,.,'.; . ; ; ;' . '. ß . .: . . '.
ß :..... .' .., ..'.. .. . ' ..... ;., ,'. .; ß ;....:.,..,.. ;,:.%. .. ,.. ., . ..: ..... ...
ß , ' . , .'.:. o.;' ß ."/.:.....,.....D, '. ':'.:.,.' ,."..;;..: ..ø"..' J, :.L. ß ß .'. . . .
..., .... , ,.,,,:.,:.,,.,.,.,'.,,. ,.'.,r,,':. L [,,',.',.,;,' ,,.'., ...,' ...:... ß ..., ,' .
, ,,..,., , .. .... : ...:l ,.,.,. ;i.. ,. .... .' ß ..; .... , ß . , , .
 , '$::ß ß : ;',';....!:,,.,.'.."..;....,.:.,'?, .;.'..'. ;..., .,.; .... . .:,. .,'....;. . :.
ß :,!.'..' '. Mr 'F, R. Collo . ß . . . -, ,. ..... ,v -... ß . ;..,,. '.... ,.' .. ;.
'"' "' ' Corportiou ' . ". ' ' "' ..' :' :' '. ." ß ' '. :;.".,'.'. "' ''
" '..": The. RAFB ;
i;' :;'' 1700 Main Street
"' '"':' Santa Monica, Clifornia 90406 ' ': '' ":" '"" '" '
';. .;' ....
,,: .. ß . ,;"' ß ; -/.:; '.,;" .,.;.'. , ,
":"'""';' 'De  Co33bo: :''"
" The RkND Recommendation to the Air Staff - Development of the .':'." - '" , "'"'.
i...:'.'.':' ß Distributcd Adaptive Me$.age-Block Network has bean reviewed by ""'.' ." '"' ":
.T!,.,'.'.:!..:. the Air.D-Z.-aOor James $ineton of my Comm& =d ControZ .'...{ .:; .... "::!. ] '
'.';...... oeet ß ß '. ß
- ;". :." zsc.to pre, :'.::. '
.. "; "" nfo 7o of loreas o is' roect, ,:', .::..-,"..";i:'..'...'- .i';..' :] ",'
ß . \, / . .- .;.,'..,... ,.., ..:.. :.:,., . ,. . .. '.'..; :;,.,,.::....; ..: .... ..
,': . ' ':.';:';'.',' ;..-'.':",' ,..!...' "..
) ', ' ' ' ß ' ß '...';'",, ' ,.-:' ;.' ,'. L, ß
 ; . ß - .' ' .,.' ',.,, .k,'. "5'..'.,. " .  . ;" .'.. . ".'.-' . ß ' ','" . - .
ß ..' ß -.,/3 .,-'  .' ', .':'.;.'..,'..%'..',.: : :r,.".': .',:''.v:..,: .;. '... ß .
ß . . . .... .. ... , ':,.;-,... ',:...-:. ;,. . ,..,  .... . .; .,.. .. .....  . ß , , . .
.... _lVn.. /, L, -.,,,,',A . . .. ......... . ..;.,.. ..... '...,....., .., .. .. , .-' ....... :..... .'. ....., -. ß
-', '   ,'T"'" ' ' .';; /: .. v'..".y,li::....,, :!.:.:.'.:..',?.....,:..q..':!:"::..,..'". .... '"' '".'.":';......" ....
OT': n " ß .' ' .'," '...' ' .' L:.':.t;.,";:....:."',: :.' !';.',,.:::.":.,,-.,' , '.:" ' ::" ' '-... ".. ";':. :" : .
ß .  . In he absenco of: '.."', :, .' :,r;:','..v ;.:',-'.". ;;"...:'..':?.':",-" . ',........: .  ß .
", ,., '. _ ' . ' ' '. " .' ;''' ';',',..: .;.{'".'.'"".' ' ;.'.'.,t L, ','. , ß -'. .'.'..'' .-'. . '. .' ' .
':'," C u, CAZZOK . .. .... ..:.. '.,: '; :" ,;.'..:,.:,';''.'.t,.' .;,;; :'_'"';i".; ":: ..:':r:X'.:':-. '...', ß .'- .': ß '" ß .
,,,; .... <,. ;,[aJor ½enerl, USAm" ..... ;".v;-'.;';,,:.',.::'..,;'',',: ..... , .... ::':'..:' ....... :-..' ß ': ........
.' "Dir,otor of 0 o,ationnl R,qismont k.:.";',;3',;' T.,L'.:; '.,'...;',:": ". '.' .'. ':: ":' .".: .; ......
.,' .;." ]3 - . .:.; :.,..=...',..:,:. f..: ,..'....,.";",:..-: .': ....,.,.,:..... -.. , ....... , ß
ß """. ' '"; . .... ,, :'. ,' 4.:'.' .-J .,'";.'':: ;';';':.,'.:.' ..".';;',';" '., ".', ' ß ,' '.-,;;,l ': .... '. ." ' ....
ß , ..... '., Developmont lnns DOS/PD ...'  :.,....,:,...:'." :.....!.'[,!.,'..j, ?..:;'.. . ....... .... ,' ,...,.,.'..'. ; .! '' ß .
,,:' .... ,, . . .... ........... ......,..., ...;:,..',,..:....,: ......,.-.:..:.'.-;,. :..:;......,:.,..-,.,,.,....., .., .. :. , .. ..... . ..: ß
ß .. .. ., . . . . .... ,.,,.:...,... :: ,,....:,.. ,-. , ...;:'-.y .,../ ,. :..,., ,;..,. ., .., , ..... ., , , . , ,, .
 . . . . . . ß . . .  .... ß ., ." ;. . .'., ., . -.. ..... .. . . ..;., -, . .  .. . . . . . ß . ,.
..... ' , . . . :, , ..... ,..:...', : ,;...'.:!: .; . ... ;. . .....:, .., . ,
.... ß :. ' ..':. ' :'.,'":7,:.".". ',, ,.'..'' ½ ....',' "..' ;:',-'.';.I ß ' ; ...... .. '" .., ., . . .... , . ß
.,,,.,. ß ',  ß ' <. " .""." ;. J ,?'.,:. '  ':', ' :.;',t.'"..:' : -. ' ,. ? ".', ß '.':'."-' .''.. .. .... : ß '... ' ' " ' ' ß
', "' '' '" . .... ' '7 "" .':..',/ "',L;.',"'.'-":":.:: ': " , ':. "."..'..'.. ".' ': ,- '." ' ",' ' '.,' ,..' ' . "
ß '.'.. ", ; . .' ' . ' ..,.; .... ;.',;" '::. '31,;.':'.,,,; r-'.'; 'J.::,'.' .'" 3: .- ;... :<,.':'..'< ..," ""., '!, .' ". ' '. .' ,'' ' ' . .'.'
ß ""-. . ". '...' .. : '. ,' ß ." ',':-.'".": :'.', ...... ;.'.: ß i.;',.;.. V';'. "."..".'. "."': ':..."..'".. '": '. ß ß ,'. ß .' '.
,;.. , . .. .... .. : .. . ,.. ... ,, .,:' .... ,.., ... :', ,...,.  ..... .., : .... ,..... :,. ;. ,. . ß . ß .
,',.:, . ß .'... ß .... ' .,..,....:4'' .'.'.';..';-. :' .,:.'." ',:.....:.., ' :,".'.4.: .,.'. ,,..' . ß . .. ". ' ß
...,, ', ' ..... , .':,F., ,.'J."L:'.?4 ",,: ,'"'': '", ,..
' ..; , ! .. . '..., ß .%.....,.; ...'.'.:.,/... .'.. ' ..,...,. .. .' '., . . . ..
ß .,.. ,.. ,-...., . .... . .... ,. '.,..,.,;......;,:.!:;:.,'...' :..;: :..: ,:....,:...,! '..: .;..:.., ... ..., ..
-,,,, ' ß '. ', '" ' '.',";V ",."""',',' ""i";'" ""'" ""'! ...... ' ' ' ' ' '
, ..... . .... ...... ;, ... , .,. . . ;'.,: . . . . ...... . ' . . .
ß .. . , ß .... -..., ';-., . .; ..'< , ; ß. .... .,. ;; , ß . .... : ß : .. ; ..' . . , ., . .. ",. .
ß ,,' .' . ' '.'. ,..':. .: ß .';. ,,..,.,-.',,, ß .: ß ß :'.',, r,.. ,,: , ... :., ß ....... ',;,.'. .
ß . -., , '., .,., ,. ', ..,',,'.'; ß ... . .., ', : ; . . . ,. , ß .. .
"i'.': - i ,: . . ' "::'."'""-'}'i::t:::',"::"'11'.": ':' :::.:::i',.':. ',"i'"':' .." ;. ' ' ß
------------------------------<page break>-----------------------------
,,, ,.,,., AtJLa. c hi
------------------------------<page break>-----------------------------
, ,"
  T  T 
10 I1 12 13 14 15
o
MAY 196
'' 13
:' : : '  :  Lincoln's Birthday
2B 2P 30 31
JUNE 1967
S  T W T F s
20 21 22 23 24 '
2 26 27 2B 29 0
S M T W T [ S '
2 3   6 7 B
lB I 20 21 22 ' ,.
I 2 3 4 5
6 7 B 9 I0 II 12
27 2B 29 31 ß '
s M T W T F
3 4 5 6 7 B 9
10 11 12 13 14 I
17 lB IP 20 21 22 23
24 2fi 26 27 28 2 30
i OCTOBER 1967
'j S M T W T F S
4 5 6 7 ,
S M T W I F S  D .
19 20 21 22 23 24
26 27 2B 2, 30
DECEMBER 1967
S M I W T F
' 17 18 19 20 21
24 2 26 27 28 29 30 : "
JANUARY
7 B P 10 11 12
21 22 23 24 23 26 27 ß '
j 2B 29 30 31
'J FEBRUARY 168
] S M T W T F
4 5 6 7 B 9 10
tB 19 20 2l 22 23 24  -.
23 26 27 28 2 ,..,..?,
------------------------------<page break>-----------------------------
NOVEMBER 1967
5 M T W T
3 4 $ 6 7 8
10 II 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 'tO
"31
JANUARY 1968
S tA T W T F S
I 2 3 4 5 6
7 B 9 10 11 12 13
14 J 16 17 18 19 20
:21 :23:24:25 26 27
2B 29 'tO ,31
JE8 RUA RY 196B
S A', T W T F 5
.I
3 4 $ 6.7 8 9
lO 11 12 13 14 15 16
17 1] 19 20 21 22 23
24 :26 27 28 29 30
31
APRI I ß 1968
5 M T W T F 5
15 16 18 19 20
21 22 23 24 2 ' 26 27
28 29 "30
},RAY 196B
S M T W T F S
20 21 22 23 24 25
JUNE 196B
S M T v,-' T F
1
2 3 41 5 6 7 B
9 10 12 13 14 I.
i! 17 18 19 20 21 22
ß 24 25 26 27 28 29
JuLY 1968
1 2 3 5 6
7 B 9 10 1
I?
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 "30 31
AUGUST 1968
S M I W T F S
123
45678910
11121314151617
18192021222324
25 26 27 28 29 "30 "31
SEPTEMBER 196B
STWTS
1 16 17 18 19 20 21
22 23 24 25 26:27 2B
29 "30
OCTO8 ER 1968
S M T w T F 5
'1
29
I 2 3 4 5
6 ? B 9 10 11 12
13 f 15 16 17 lB 9
20 22 23 24 25 26
27 28 29 30 31
------------------------------<page break>-----------------------------
"Nsvsm]sr
lg7
$ M T W T f
3 4 $ 6 7 8 9 ....
1 2 3 4 5 6
7 8 9 10 i1 12 13
1 2 23 24 25 26 27
28 39 30 31
MARCH 1968
S M  W T F S
1
3 4 5 6 7 B 9
2 5 26 2 28 29
S M T W T  S
8
19 20
MAY 1766
5 6 7 B ? 1o
1 2o 21 22 23 4 25  .
26 7 28 29 3o 31
JUNE 168
S  T W T F
JUL 16B
S  T W T F
1 2 3 4 5 6
s  T W  F S "'
, 5 6 7 B 9 10
18 I? 20 21 22 23 24 ..
2 26 27 2a 2 30 3t
SEPTEMBER 166
S M T W T F S '  .--
23 24 25 26 27 28
OCTOBER 1968
S M T W T F
1 2 3 
6 7 8  10 II 12
0 21 22 23 24 2 26 .. -.
27 28 29 30 31
NOVEM8ER 1968
S M T W T F S
3 4 5 6 7 8
17 18 1 20 21 22 23
24 25 26 27 28 29 30
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
Attachment #3
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
III II
NOIIVOINnlAIIAIO0 NI SNOIIVIS dO NOIlOVd
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
]
]
]
]
]
]
v
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
]
]
]
------------------------------<page break>-----------------------------
c
------------------------------<page break>-----------------------------
I
1
I
]
]
1
N
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
SJ. NdNI
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
REJIABT,W, DIGITAL COMMUNICATIONS_-$YSTS
USING UNRELIABLE NETORK REPEA'R NODES
Paul Baan
Mathematics Division
The RAND Corporation
P-19S5
May 27, 1960
Reproduced by
The RAND Corporation ß Santo Monica ß Californ a
The views expressed in Ihis paper are not necessarily those of lhe Corporalion
Attachment #4
------------------------------<page break>-----------------------------
RELIABLE DIGITAL COMMUNICATIONS SYSTEMS
UTILIZING UNRELIABLE NETWORK REPETER NODES
I}.RODUCTION
The cloud-of-doom attitude that nuclear war spells the
end of the earth is slowly lifting from the minds"0.-the ,many_j_
Better quantitative estimates of post-attack destribn to-
gether with a less emotional discussion of the alternatives
may mark the end of the "what the hell--what's the use?" era.
A new view emerges:
the possibility of a war exists but there
is much that can be done to minimize the consequences.
If war does not mean the end of the earth in a black and
white manner, then it follows that we should do those things
that make the shade of grey as light as possible: to plan now
'to minimize potential destruction and to do all those things
necessary to permit the survivors of the holocaust to shuck
their ashes an reconstruct the economy swiftly.!
PURPOSE
This paper is directed toward the communications
system designer in the hope that he will not neglec these
problems of survival n the design of his future
systems. Unfortnately no panacea can be presented.
------------------------------<page break>-----------------------------
There is none. Only a single lim_ited technique is described,
that of using one form of distributed redundancy to minimize
vulnerability A detailed examination demonstrates that the
problem is not an nscivable one, but that much money and much
work are necessary t,o provide all-the elements needed for the
rue solution to the vulnerability problem. .
The timing forsuch thinking is particularly appropriate
now. for we are just beginning to design and lay out designs
fcr he digital data transmission systems of the future. We
s. re Jus5 approaching the initiation state of the designs of
systems where computers speak So one another. When this day
comes, the digital communication s õystem capacity needed will
be crders of magnitude greater than our chSef present-day
digital transmission technique; nmely, telet)Te. For example,
a 60 WPM teletype channel would take about two weeks oerating
24 ho,'.rs.per day without breakdown to transmit a single standard
size reel of magnetic tape used on a digital computer. This ,.may
De compared to a writing time of about 6 minutes,.on the computer.
Millimicrosecond clock rate data links will be needed in
the  ..... . 
i_t .... Our prosen5 co,-ununicasions olant wi TM not suffice.
As there does not seem to be any fundmenta! technical prcblem
that prohibits the operation of digital comnications links at
the clock rate of digital computers the view is taken that it
is only a matter of time before such design requirements be-
come hardware. The systems designs of the nex few years will
be the systems-in-being of the next few decades. It is hoped
------------------------------<page break>-----------------------------
that these systems will be foresighted enough to include an
ins]trance policy purchasing invulnerability.
CONTENTS
This first paper of this series describes a rai-onale for
the use of red,andancy in networks and the oayofr
destees of such redundancy.
The second paper, by Frank Yates and Paul Baran, describes
..
a ..io.. speed automatic switching teclmnique for digital
messages. It. this paper the time of arrival of messages ic
used to determine the shortest.routing paths over a severely
damaged network.
The third paper, by Paul Baran and Robert Hameriy, describes
a lower speed digital system using a tag containing the number
of times a message has been relayed to provide a method for
authenticating point of origin.
These two separate systems exhibit different properties;
it is possible to multiplex them together to
ui_z.e the ad-
vmntages of both and devise a third system. It is thus pcssible
to visualizs a new set of systems based upon a distributed
organization.
Atcmat_ic route selection is not new, but these papers seek
a._s a comon goal synthesis of systems where the intelligence
requi,____red to switch signals to surviving linxs is at the link
nodes and not at one or a few centralized switching centers.
This paper thus describes what we choose to call distributed
craauizasion as compared to the more fa.m_iiiar cntra!ized
swltc.,ing center type orgo-nizaticn.
------------------------------<page break>-----------------------------
Attachment 6
]
][
IEEE publications
scnnin the issues
special publications
Scanning the issues
{!
Distributed Communications. A great
deal of thought has necessarily been
given ?n recent years to the question
of the survivability of communications
systems in the event of attack. Un-
pleasant though such a prospect may be,
the studies themselves have led to some
interesting suggestions as to what form
our future networks might take.
OF particular interest is the distributed
communication network concept in
which each station is connected to all
adjacent stations rather than to a few
switching points, as in a centralized
system.
Although one can draw a wide variety
of networks, they all factor into two
components: centralized (or star) and
distributed (or grid or mesh). (See
,pes (A) and (C), respectively in Fig.
The r:entralized network is obviously
Fig. 1. (A) Centralized. ,rB) Decenlralized
(C} Distribution networks.
,' ' ' Link
  o o o o o o
o o o
0 O O O  0
R i =J.5 R=2
vulnerable, as destruction of a single
central node deslroys communication
between the end stations. In practice,
i,
 mixture'of star and mesh components
is used to form communications net-
works. For example, type (B) in Fig. l
shows the hierarchical structure of a set
of stars connected in the form of a larger
star with an additional link forming a
loop. Such a network is sometimes called
a "decentralized" network, because
'complete reliance upon a single point is
not always required.
Since destruction of a small number
of nodes in a decentralized network can.
destroy communications, the properties,
problems, and hopes or building "dis-
tributed" communications networks are
of paramount interest.
In a recent study of the effects of an
attack on a distributed network, an
investition was made of the per-
centage of surviving linked stations
versus the probability of station de-
struction for various levels of network
redundancy. Redundancy level is, a
measure of connectivity and is used here
as defined in Fig. 2.
A key point revealed by the study
is that extremely survivable networks
can be built with the use of a moderately
low redundancy of connectivity level.
Redundancy levels on the order of only
three permit the withstanding of ex-
tremely heavy level attacks with neg-
ligible additional communications loss.
In add}don, when one studies what
happens when the links rather than the
nodes of the network are destroyed, h
becomes evident that highly reliable
networks can be achieved with low. cost
unrelhble communication links, even
links so unreliable. as to be unusable in
present networks.
Our future systems design problem is
that of building at lowest cost very.
reliable systems out of the described set
of unreliable elements, For communi-
cations or the future, digital links appear
increasingly attractive by permitting
tow-cost switching and low-cost links.
In any event the network must be built
with the expectation of heavy damage.
Powerful error removal methods exist.
Some of the communication con-
s/ruction methods that look attr.,ctivc
for the na-r--fuiure include pulse
regenerative ,r:?P_CatCr line, minimum-
cost or "rninico v microw.ve, TV
broadre, st station i:ligital transmission;
and satellites.
In communimtmions, ;s in mmsporta-
Fig. 3, All.digital network composed
mixture of links,
don, it is most economicall for many
users to share a common rource
rather than for each to build his own
system. particularly when intermittent or
occasional service is supplied. This
intermlttency of service is highly char-
acteristic of digital communication re-
quiremeats. Therefore, we would like to
consider one day the interconnection
of many aIl-dig/tul links to provide a
resource optimized for the handling of
data for many potential intermittent
users: a new common-user system. '
Fig. 3 demonstrates the basic notion:
A wide mixture of different diglmI
transmission links is combined to form
a common resource divided among
many potential users. (P. aran. "On
Distributed Communications Net-
works," IEEE Trrm$. rm C,mu,umicr-
/ions Systems, March 1964.)
Repetition in Learning. Learning iS
facilitated when subject matter is pre-
sented in extremely small segments and
the student is stimulated to respond
immediately, at frequem but increasingly
------------------------------<page break>-----------------------------
Attachment #7
u
u
------------------------------<page break>-----------------------------
II
il
il
------------------------------<page break>-----------------------------
------------------------------<page break>-----------------------------
F""
U
0
U
U
------------------------------<page break>-----------------------------
u
u
------------------------------<page break>-----------------------------