本帖最后由 WSHXPY 于 2015-2-5 20:13 编辑
8 I0 ?8 @/ @) {1 t$ Y4 c: K6 u1 x3 U t
An Introduction to CELLULAR AUTOMATA ! V8 p% W" T( m5 p( v- s
# ?0 [3 v# D, W' N% R+ `5 hBy H′el`ene Vivien" P. R/ e. b j) i0 b5 T
- ~, R( N# D* e
% B5 }: u* Y, ^+ _! s2 E o* o. M% e9 Y" `3 }+ R
1 h3 Y5 L+ Q" i6 Y( l+ v3 G3 @0 m
The order of the book is from the simple to the more complex, introducing a question as soon as we have the frame to put it and the means to solve it.% [ g. Y* W5 R% v( D" \, C
Chapter 1 starts with the finite linear c.a’s. They are precisely the required framework for introducing the Firing Squad Synchronization Problem, and giv-ing the first and simplest solution, that of Minsky.4 }$ |8 X* b+ c! b/ F+ d' @
Chapter 2 presents Mazoyer’s solution which is elegant and sophisticated. The needed display of technics also introduces to the abstract notions to be devel-oped in later chapters." _3 k& m4 m/ t- A
Chapter 3 clarifies some important notions related to semi-infinite linear c.a’s : inputs and outputs, computing a function, recognizing a language. It also gives fundamental examples of language recognition.
' T/ n D. ?- g4 h2 wIt is then possible to relate c.a’s with Turing machines, which is done in Chapter 4, which ends with Atrubin’s real time algorithm to multipliy two integer numbers. Semi-infinite linear c.a’s are the required framework for many developments in the next chapters.& Q) Y! n1 s4 n( Q: H% q
Chapter 5 gives a precise definition of a signal. Different notions then appear, those of waves and networks of signals, with the beautiful example of Fischer’s c.a for recognizing prime numbers, and the gap theorem for waves.9 k3 p, S+ B8 a6 `0 q
In chapter 6, slowing down is studied. It may seem absurd at first view. But we shall see that it is useful, next to speeding up, for intelligently driving our c.a's. This incidentally leads us to build a c.a which computes particular word morphisms.
: Y" M6 w; K: D* f- gChapter 7 is an exhaustive study of speeding up, for recognizers and for syn-chronizers.
1 u5 _9 p# q" m& UChapter 8 is devoted to the study of the family of synchronization times. The richness of this family allows for flexible use of the synchronization process.
: i ^. \3 \# m4 aChapter 9 introduces the geometrical c.a’s, i.e. c.a’s the states of which are geo-metrical pieces. Numerous examples give new and elegant methods for speeding up.
" S" o5 }8 a5 b$ |* X$ }Chapter 10 is a general study of n-dimensional c.a’s and of the languages that they recognize. It is based on Cole [6].
4 _( h5 { R% k- G9 x/ n# d( uIn the last three chapters, communication delays are introduced. They consid-erably complexify the behavior of c.a’s, in particular their synchronization.# Y6 F" y. }" d6 O$ q4 y$ F
Chapter 11 considers a line of two automata. Solutions for synchronizing these two automata are presented, and they are proved to be almost optimal.
+ [+ A6 F5 _; {7 Q# o) _Chapter 12 considers a finite line of automata. A c.a synchronizing such a line is presented, which mimics Minsky’s method of chapter 1.% `0 A- f1 I3 p! R/ p$ z \
In chapter 13 we first give the general definition of a network, without or with communication delays. We show how automata placed at the nodes of the net-work may solve graph problems. We then use this possibility to transform the network into a fictive line. Applying the result of the preceding chapter to this line, we obtain a c.a for synchronizing the network.$ I- J+ E* i. r* E( I
& o. y& L- a' N- |! ~
- H5 m( }8 J+ X$ o L0 [# A: H
3 g8 ?; U) n3 {: F+ ^0 l9 `$ w: Q
4 r0 x& T0 [( W% A
: c1 u) H7 u. Y: s. f. k7 \
: Z3 o+ ~1 J/ ^: l# p) h4 l
- d( F9 G3 h' g1 _- A, r% T6 N" J# y$ a V
3 V9 M4 z* R+ s3 ~$ h5 i, t
! V. M8 _" k: V! o: _+ @ |