<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Sdrka VavredkovS</string-name>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Institute of Computer Science,Faculty of Philosophy and Science,Silesian University in Opava</institution>
          ,
          <country>Czech Republic sarka</country>
        </aff>
      </contrib-group>
      <fpage>235</fpage>
      <lpage>242</lpage>
      <abstract>
        <p>Eco-coloniesare new grammar systems with very simple regular glarrunars called agents. Every agent gerreratesits own finite language, all agents cooperate on the shared environment. The environment is not changed only by agents, but it can develope itself. The generating power of eco-colonieswas discussed in several papers, eco-colonieswere compared especially with various types of colonies,but not all relations were proved. In this paper we summarize the published results and we present some other non-published results about the generating power of eco-colonies. I{eywords: Eco-colony,0L eco-colony,EOL eco-colony,property, colony, agent, component, grammar system</p>
      </abstract>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>Introduction</title>
      <p>Colonies were introduced in [ ] as collections of simple grammars (called
components) working on common environment. A component is specifiedby its start
symbol (an object or another symbol from the environment, the component can
find its start symbol in the environment and processit) and by its finite language.
This language determines actions to do with the start symbol, it is usually a list
of words, the component substitutes its start symbol by some of these words.
The environment is static, only the components can modify it.</p>
      <p>There are several variants of colonies with various types of derivation. The
original model was sequential(only one component works in one derivation step),
the other basic types of derivation are sequentialwith parallely working
components or parallel. Parallel colonieswere introduced in [3], parallel behaviour of
a colony meansworking all the componentsthat can work (the component whose
start symboi is in the environment and no other component is occupying tiris
symbol for the actual derivation step), one component processesone occurrence
of its start symboi.</p>
      <p>Eco-colonieswere first studied in 17],their EOL form in [5,6]. trco-colonies
are colonieswith developingenvironment. The concept of developingof the
environment is inspired by another type of grammar systems,eco-grammar systems
(t2]) The environment is not specifiedonly by its aipliabets I/ and ? but as 0L
(one alphabet) or EOL (two alphabets) scheme.Every symbol of the environment
not processedby agents (components) is overwritten by some of the developing
rules of this scheme.</p>
      <p>Eco-coloniesare useful for modelling some simple processesin nature and it
is not so hard to programme the model. Here are some elementary examples of
models:
- rabits (agents) on a meadow (environment), the developementof the
environment means growth of grass,the piecesof soil without grass are (or are
not) replaced by the grass-blades,the grass-bladesnot eaten by rabbits can
grow,
colony of ants (the ants * agents- work on the shared environment - ant-hill
and its neighbourhood),
drug (agents) acting on a bacterial cuiture.</p>
    </sec>
    <sec id="sec-2">
      <title>Definitions</title>
      <p>In this section we define colonies,two types of eco-coloniesand then two types
of derivation in eco-colonies.</p>
      <p>Definition 1. A colonyi,san /1-tupleC: (V,T,R,ws), where</p>
      <p>V is a finite non-emptg alphabetof the colony,
- T is a non-ernpty termi,nal alph,abeotf the colony, T C V,
- R i,s a fi"nite (multi)set of components,</p>
      <p>R : { ( ^ 9 , F ) l S e ( V - T ) , F g V - { ^ 9 } ) . , F ' i s f t n ' i t ea n d n o ' n - e m p t y } ,
S is the start symbol of the coTnponen(.9,.F) and F ts the finite languageof
th'is component,
- us is the atiom.</p>
      <p>We know severaltypes of derivations for colonies,here are tire ba-sictypes:</p>
      <p>The exact definitionsof basic types of derivation in coloniesare in 1I,3,4,7.,
5], definitions of an eco-gralnmar system and its type of derivation are in [2].
Deflnition 2. LetC be a colonUC,: (V,T,R). The languagegeneratedby the
deriuat'ionstepr,r e {b,t,wp,sp} in C i,s</p>
      <p>L ( C , r ) : { w e T * | u s $
r y
Definition 3. An E7L eco-colony of degreen, n
E : ( E , A t , A z , , . . . , A n , w o ) ,w h e r e
- E - (V,7, P) i,sE}L scheme,wltere
c V is a fini,te non-empty alphabet,
o T 'is a non-empty termi,nal alphabeLT g V,
. P 'is a fin"ite set of E}L rew'nting 'rulesouer V ,
- Ai: (S,, Ft), L &lt; i 1n, 'isthe i,-th agent,where
o ,.9t€ V i.sthe start symbol of the agent,
. Fi Q (V - {5r}). is a finite set of action rules of the agent (tlte language
of the agent),
- ws i,sthe ariom.</p>
      <p>An 0L eco-colongis defined similarly, only the environment is 0L scherne
E: (V,P), P is a finite set of 0L rewriting rules over I/.</p>
      <p>As we can see, agents are defined such as components in colonies, an
environment is determined by the alphabets in coionies,and by EOL or 0L scheme
in eco-colonies.</p>
      <p>\AIedefine two derivation modes for eco-colonies - the first one, tlp, is
inspired by the urp mode for coionies,we only add possibility of developing for
the environment. In every derivation step each agent (S, .F) looks for its start
symbol ,5. If it finds some occurrenceof this symbol not occupied by any other
agent, the agent becomesactive, occupies this symbol and rewrites it by some
of words of its IanguageF.</p>
      <p>Definition 4. We d,efi"nae weaklqcompeti,tiueparallel deriuation step in an
ecoc o l o n yD : ( E , A t , A z , . . . , ,A n , w o ) b y a 3 1 p , w l t , e r e
- a - ? o S r r l r S U ^ / 2... ^ 1 , - t 5 i l " 1 , , r ) 0 ,
- P : l ' o f i r l l f i r l z " t ' r - t f + l ' , , A i u : ( S r * , F t * ) , f t o € F i u , 1 &lt; k 1 r ( t h e
ugentA;u r,sactive in this deriuation step),
- { i t , i 2 , . . . , ' i , l q { 1 , 2 , . . . , n } , i r * i , n f o r e u e r yk * m , I ( k , m 1 r ,
- fo, euer?JsymbolS e V t llo^tt...^t,ls &gt; O then euery agent wi,thth'estart
symbolS must be actiue (i,f agents can work tltey must work),
- ^yx&amp; -.rf, ^{t,€V*, 0 &lt; k &lt; r, is tlte d,eriuat,iosntep of the schemeE.</p>
      <p>The second type of derivation step, ap, means that all agents must work in
every derivation step and if some agent is not able to work (there is not any free
occurrenceof its start symbol), the derivation is blocked.This type of derivation
is inspired by the basic type of derivation in eco-grammar systems.
Definition 5. We def,ne o, deri,uationstep ap (all are work'ingparallely) in an
e c o - c o l o nDA : ( 8 , A r , A 2 , . . . , A n , w o ) b y a + p , w h e r e
- o - 7oSi,^ltSU^lz. . . ^ln-rSi-^ln,
- B - ^ r ' o f i r ' i f u 1 z " ' ^ r ' r - t h - 1 ' n , A i o : ( S 0 . ,F t u ) ,f u e F i o , l S k S n ,
{ 1 , 2 , . . . , n } ( e u e r ga g e n tw o r k s ' i ne u e r yd e r i u a t i o ns t e p ) ,
- { i r , i 2 , . . . , i n } :
- ^lt 4 il, jt" € V* ,, 0 &lt; k &lt; n, is the d,eriuationstep of the schemeE.
D e f i n i t i o n 6 . L e t D b e a n 0 L e c o - c o l o n yD, - ( E , A t , A 2 , . . . , A n t l o ) . T h e
languagegeneratedby the deri,uattonstep r,,r e {wp,ap} i'n D i,s</p>
      <p>L ( E , r ) : { , e V * | * o 4 . }
Erample /. We createan EOL eco-colonyD : (E,Ar,A2,AbB), where
E - - ( { A , B , a , b } , { o ,b } , { o , e , , ,-b- b b } ) ,A r : ( A , { a B , e } ) , A z - ( 8 , { a A , e } )</p>
      <p>Let us construct derivations with ap and tlp types of derivations:
A b B 4 a B b 2 a A 4 a z A b a a 2 B$ q s g 6 8 a 3 A $ q + n 6 r 6 o a g4 . . .
AbB 4 aBbzaA ry a2Abaa2B 3 a2b8a4s $ o26r6oaBry . . .</p>
      <p>The wp derivation allowes "resting" of non-active agents. If we use the ap
type of derivation, a terminal word is generated only if the both agents use the
E-rule in the same derivation step, otherwise the derivation is blocked without
creating thc final word.</p>
      <p>\Aiegenerate these languages:</p>
      <p>L ( E , a p ) :</p>
      <p>{ a ' b 2 - a n I n &gt; 0 }</p>
      <p>L ( D , w p ) : { o t u z ^a i I n &gt; o , o &lt; i , i = " }
3</p>
      <sec id="sec-2-1">
        <title>Generating power of eco-colonies</title>
        <p>We compare the generating power of eco-coloniesand colonies, eco-grammar
systems and a special type of colonieswith the terminai alphabet equal to the
alphabet of the system. We use this notation:
\EC,
EEC,
r e {up,op} classof 0L eco-colonieswith r type of derivation
r Q {up,op} classof EOL eco-colonieswith r type of derivation
COL, r e {b,t,wp,sp} classof colonieswith the r type of derivation
COLT r e {b,t,up,sp} classof colonieswith the r type of derivation,V -T
EG class of eco-grammar systems
Properties of Eco-Colonies
3.1 Known results</p>
      </sec>
      <sec id="sec-2-2">
        <title>We proved in [6] these results:</title>
      </sec>
      <sec id="sec-2-3">
        <title>In [5] are proved these resuits:</title>
      </sec>
      <sec id="sec-2-4">
        <title>In [7] we proved the results:</title>
        <p>COL-e c EEC-e
}EC-e c EEC-e
E E C - e - E G + A</p>
        <sec id="sec-2-4-1">
          <title>COLb C EEC*e</title>
          <p>
            E E C * T - C O L I + a
C O L Tc C O L * , r e { b , t , w p , s p }
jECoe+ \EC-e
C O L Tc T E C - , , r e { b , - p }
aEcoe- COLT*A, * e {b,wp)
} E C a C E G , y e { a p , w p }
(
            <xref ref-type="bibr" rid="ref1">1</xref>
            )
(
            <xref ref-type="bibr" rid="ref2">2</xref>
            )
(
            <xref ref-type="bibr" rid="ref3">3</xref>
            )
(
            <xref ref-type="bibr" rid="ref4">4</xref>
            )
(
            <xref ref-type="bibr" rid="ref5">5</xref>
            )
( 6 )
(7)
(B)
(e)
( 1 0 )
3.2 New results
In the equation (
            <xref ref-type="bibr" rid="ref2">2</xref>
            ) we can find the relation }EC-e C EEC*'. W" prove the
equivalent relation for the ap derivation.
          </p>
          <p>Theorem 1.</p>
          <p>}ECoe c EECoe
( 11 )
Proof. 0L eco-coloniesare special types of trOI, eco-colonieswhere V : T , so the
relation 1EC"e g EEC"e is trivial.</p>
          <p>To prove the proper subset we use the language</p>
          <p>L t : { o 'L" ' ) l n &gt; 1 }
This languageis generatedby the EOL eco-colonyE: (E,At,A2,(JVa), where
E : ( { o , U , V } , { o } , { o - - a a , ( J' - n( J , V - - +V } ) , A t - ( U , { V , r } ) ,
A z : ( V ,{ U , e } ) .</p>
          <p>UVa 4 Vfla2 + UVaa $ yuq8 -sg UVal6 g ... $ oz"</p>
          <p>The agentsdo not generateany word, they exist only becausethe number of
agentsmust be greater than 0, and with using the ap derivation all agents must
work in every derivation step, so the agents A1 and A2 work with the symbols
LI,V untll the final word is gcnerated.</p>
          <p>Supposethat this languagecan be generatedby some 0L eco-colonyrvith op
derivation. We need at least one agent and this agerrt rnust be active in every
derivation step. We have orrly one alphabet V : {a}, so the start symboi of this
agent is o. But the agent generatesonly a finite language,so the symbol a is not
aiowedin the set of action rules,the agent is definedas.4.: (o, {u}).</p>
          <p>The language is exponential and the agent(-s) does not help with growing
of the language,so the rule a ---+aa is in the environment. It can generate the
language,but the agent "eats" one symbol a in every derivation step and the
environment cannot correct it (it is only 0L scheme,any other correcting rule
such as a '---Q,aawould be used freely any times and anywhere in one derivation
step), so the languageZ1 would not be generated,Lt f \ECoe. tr</p>
        </sec>
        <sec id="sec-2-4-2">
          <title>Theorem 2.</title>
          <p>I E C , - C O L " * A
0 2 )
w h e r e r e { 0 , - E } , y e { w p , , a p } ,z e { b , t , w p , s p } .</p>
          <p>Proof. In this proof we use the languagesimiiar to L1,</p>
          <p>L z : {Lc d . a z 'b" " " 'l n -) - )0 ) o- t {- a r o ' 2 n * tU 2 2 n +lt,'t" -&gt;" )o }</p>
          <p>T h i s l a n g u a g e c a n b e g e n e r a t e d b y t h e e c o . c o l o n y r
w h e r eE : ( { a , b , c , d } , { o- * a a , b- - b b , c- , c , d - - -d } ) , A r : ( c , { d } ) ,
Az : (d,{.}).
cdab+ dca2bz+ cdaaba+ d,ca8b+8 cd,arobro+ .. .</p>
          <p>Two agentswork in every derivation step and we use only one alphabet, so
it may be EOL as well as 0L eco-colonyand the derivation step may be tlp or
ap.</p>
          <p>The languageL2 LSnot context-free,so L2 f COL;, and it grows
exponentially so L2 f COL-e and L2 4 COL,e (proved in [O]and [5]).</p>
          <p>Colonieswith t type of derivation can generate some exponentially growing
languages,but only one component work in one derivation step. This component
rewrites all occurrencesof its start symbol, and there is only one start symbol
in one component. So we can rewrite a or b in one derivation step, but not both.
There is not any way to control generating the same exponential number of o-s
and b-s,so L2 e COLI. tr</p>
        </sec>
        <sec id="sec-2-4-3">
          <title>Corollary 1.</title>
          <p>L E C | - C O L T * A
w h e r er € { 0 , 8 } , , y e { w p , a p } , z e { b , t , w p , s p } .
Proof. Follows from Theorem 2 and Equation (6).</p>
        </sec>
        <sec id="sec-2-4-4">
          <title>Theorem 3.</title>
        </sec>
      </sec>
      <sec id="sec-2-5">
        <title>Proof. In [7] we proved that the language</title>
        <p>COL, - }EC-e + A, r e {b,t,wp, sp}
L s : { a r 5 - 2 n 6 n c b "ld0 ( n I T , n i s e v e n }</p>
      </sec>
      <sec id="sec-2-6">
        <title>U {als-z"bncb'clI o &lt; n I T,n is odd}</title>
        <p>( 1 3 )</p>
        <p>tr
(14)
is not in}EC-o. It is a finite language,so.L3e CoL, fbr r € tb, t,wp,sp\. r</p>
        <sec id="sec-2-6-1">
          <title>Propertiesof Eco-Colonies 241</title>
          <p>Corollary 2. The set of languages}EC-o i,s i,ncomparableto the sets of
languagesCOL6, COLI, COL-, o,ndCOL,r.</p>
        </sec>
        <sec id="sec-2-6-2">
          <title>Proof. Follows from Theorems 2 and 3.</title>
          <p>Corollary 3.</p>
        </sec>
        <sec id="sec-2-6-3">
          <title>Proof. Follows from Equations (6) and (2).</title>
          <p>Theorem 4.</p>
        </sec>
      </sec>
      <sec id="sec-2-7">
        <title>Proof. In [7] we proved that the language</title>
      </sec>
      <sec id="sec-2-8">
        <title>COLTc EEC-., COLT,,c EEC-,</title>
        <p>CO L, - \ECoe * A, r e {b,t,rup, sp}</p>
        <p>L+ : {a, aa}</p>
        <sec id="sec-2-8-1">
          <title>COLa c EECoe</title>
          <p>is not generatedby any 0L eco-colony.But this languageis finite, so La e COL"
f"orr € {b, t,wp, sp}. !
Theorem 5.</p>
          <p>Proof. We have a colony with the b mode of derivation C - (V,T,R,tus) and we
createan EOL eco-colonywith the ap derivationD: (E,At,Az,BC otlo).</p>
          <p>The agents ,41 and A2 work very simply - they rewrite onlv one symbol to
another one: ,41 : (B,{C,u}), A2 : (C,{B,,e}), such as in the proof of the
Theorem 1.</p>
          <p>We create rules of the environment from the components of. R. Suppose
that all components in R have various start symbol, there is not any couple of
componentswith the same start symbol.</p>
          <p>For everycomponent(cl,{a1,e2,...,(tk}) *" createdevelopingrulesfor the
environment:
androreverysymboblwhicht:;i't1'llj], ,r',l1",inanycomponewnetcreate
one rule b --- b.</p>
          <p>So the environment simulates working of the components in the colony. To
simulate the sequential derivation we must allowe to rewrite every symbol to
itself. tr
Erample 2. We demonstrate the construction of the proof on tire colony
generating this language:</p>
          <p>L s : { w a w R a i l u e { 0 ,1 } " , i &gt; 0 }</p>
          <p>We havea colonyC : ({S,H,H' , A, A' ,0, 1,o}, {0, 1,a},R,S) generatin gthe
language,the set of conrponentsR is
n : { ( s , u / A i ) , ( H , t 0 . H ' 0L,H ' r , a } ) ,( H " { H } ), ( A , { o A ' , . t r } )( ,A " { A i ) }
!
(15)</p>
          <p>n
(16)
(17)
B C S4 C p n A 4
4 CntlHlrA4
$ lgsglsq
B C S4 C A n A I B C r H , r a H4
4 C a t o H o r a a g l g s g l s s</p>
        </sec>
        <sec id="sec-2-8-2">
          <title>Corollary 4.</title>
          <p>Now we create an EOL eco-colony with op derivation X : ( E , A t , A z , B C S ) ,
E : ( { B , C , S , H , H ' , A , A ' , 0 , 1 , o } , { 0 ,1 , a } , P ) ,
A t : ( B, { C ,e } ) , A z : ( C ,{ 8 , e } ) ,
the set of rulesP in the environmenits
P : { H - - H l 0 H ' 0 l I H ' I l A , H ' - - H ' I H , 1 _ , 1 ,</p>
          <p>A -, AlaA'la, A' -. A'IA, 0 -- 0,
A - + Q t
S - , S I H A |</p>
        </sec>
        <sec id="sec-2-8-3">
          <title>One of derivations in C:</title>
          <p>s,4 na 4 rH,LA+ :HLA4 roa,oL+A r0H0rA4rooora4
4 toootoA' + roaoLaA4 roootoo</p>
          <p>Two of possible derivations of the same word in I:</p>
          <p>B C T H , L A4
nCtlalrA9</p>
          <p>C n t H r A 4
CntlalraA, g
s C t l v , l r A g</p>
          <p>nCtlalraAJ4
C B r H r a Ag</p>
          <p>B c r y H , l r a a4</p>
        </sec>
        <sec id="sec-2-8-4">
          <title>COLT c EECoe</title>
          <p>Proof. Follows from Theorem 5 and Equation (6).
(18)
!
This work has beensupportedby Grant Agency of CzechR,epublic the grant IVo.
201/ 06/0 56T, "Bzo,inforntat'ika a biouypoit'y:so'u'u'islostm'i,odely, aplikace".</p>
        </sec>
      </sec>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          1. csuhaj-varjri, E.,
          <string-name>
            <surname>Dassow</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Kelemen</surname>
          </string-name>
          , J., pd,un, G.:
          <article-title>Gramrnar systerns. A Grammatical Approach to Distribut'ion and Cooperat,ion</article-title>
          .Gordon &amp; Beach, London (
          <year>1994</year>
          ).
          <source>ISBN 2887249574</source>
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          2.
          <string-name>
            <surname>csuhaj-Varjri</surname>
            ,
            <given-names>E.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Kelemen</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Kelemenovd</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          , piun, G.:
          <article-title>Eco-grar-nrnarSystems. Gramat'ical Frarnework for Studying Lifelike Interactions</article-title>
          .
          <source>Artificial Life</source>
          <volume>3</volume>
          (
          <issue>igg7</issue>
          ) pp.
          <fpage>l</fpage>
          -
          <lpage>28</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          3.
          <string-name>
            <surname>Dassorv</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Kelemen</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          , PXun, G.:
          <article-title>On Parallelism i,n Coloni,es</article-title>
          .
          <source>Cybernetics and Systems</source>
          <volume>24</volume>
          (
          <year>1993</year>
          ) pp.
          <fpage>37</fpage>
          -
          <lpage>49</lpage>
          . ISSN 0196-
          <fpage>9722</fpage>
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          4.
          <string-name>
            <surname>Kelemetr</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          , Kelenrenovd,,
          <string-name>
            <surname>A.</surname>
          </string-name>
          :
          <article-title>A Grammar-theoret,ic treatment of Multiagent System,s</article-title>
          .
          <source>Cybernetics and Systems</source>
          <volume>23</volume>
          (
          <issue>Ig9Z</issue>
          ) pp.
          <fpage>621</fpage>
          -
          <lpage>633</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          5.
          <string-name>
            <surname>Vavreckovd</surname>
            ,
            <given-names>S.:</given-names>
          </string-name>
          <article-title>EOL el;o-kolonie</article-title>
          .In:
          <article-title>Kognice a umelf Zivot Vi</article-title>
          . Silesian University. Opava (
          <year>2006</year>
          ), pp.
          <fpage>413</fpage>
          -
          <lpage>419</lpage>
          . ISBN 80-7248-355-2
          <article-title>English abstract on http:l /www.cs.cas.czf kuz2A06/ABSTRACTS-complete.doc o . V'avreikov6</article-title>
          ,
          <string-name>
            <surname>S.</surname>
          </string-name>
          : Eco-coLoniesI,
          <source>n: IvIEMICS</source>
          <year>2006</year>
          ,
          <article-title>proceedings of the 2nd Doctoral Wbrkshop</article-title>
          .
          <source>Uni'ersity of Technology, FIT, Brno</source>
          ,
          <year>2006</year>
          , pp.
          <fpage>253</fpage>
          -
          <lpage>25g</lpage>
          7. VavreckovS, S.:
          <article-title>Eko-kolonie</article-title>
          .
          <source>In: Kognice a umelli Zivot</source>
          V. Silesian University, Opava (
          <year>2005</year>
          ), pp.
          <fpage>601</fpage>
          -
          <lpage>612</lpage>
          . ISBN 80-2248-310-2
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>