Cellular Automata

Waleed Khalid
Outline

 What is a Cellular Automaton?

 Wolfram’s Elementary CA
 Conway's Game of Life
 Applications and Summary
3
INTRODUCTION
What are Cellular Automata?
A cellular automaton (plural: cellular automata) is a discrete model studied in
computability theory, mathematics, theoretical biology and microstructure modeling

 CA are discrete dynamic systems.

 CA's are said to be discrete because they operate in finite space and time and
with properties that can have only a finite number of states.
 CA's are said to be dynamic because they exhibit dynamic behaviors.
 Basic Idea: Simulate complex systems by interaction of cells following
 Moore easy von Neumann
neighborhood neighborhood
rules.
 4
COMPONENTS
 Grid
 Mesh of cells.
 Simplest mesh is one dimensional.
 Cell
 Basic element of a CA.
 Cells can be thought of as memory elements that store state information.
 All cells are updated synchronously according to the transition rules.
 Rules
 5
BEHAVIOR
 Local interaction leads to global dynamics.
 One can think of the behavior of a cellular automata like that of a “wave” at a sports event.
 Each person reacts to the state of his neighbors (if they stand, he stands).
CELLULAR AUTOMATON

 An initial state (time t=0) is selected by assigning a state for each cell

 A new generation is created (advancing t by 1), according to some fixed rule that
determines the new state of each cell in terms of:
 the current state of the cell
 the states of the cells in its neighborhood
 Typically, the rule set is
 the same for each cell von Neumann
 Moore
neighborhood neighborhood
 does not change over time
 applied to the whole grid simultaneously
 Background
Ulam

 Originally discovered in the 1940s by Stanislaw Ulam and John von Neumann

 Ulam was studying the growth of crystals and von Neumann was imagining a world of
self-replicating robots
 Studied by some throughout the 1950s and 1960s
 Conway's Game of Life (1970), a two-dimensional cellular automaton, interest in the
subject expanded beyond academia
 In the 1980s, Stephen Wolfram engaged in a systematic study of one-dimensional cellular
von Neumann
automata (elementary cellular automata)
 Wolfram’s research assistant Matthew Cook showed that one of these rules has a VERY
cool and important property
 Wolfram published A New Kind of Science in 2002 and discusses how CA are not simply
cool, but are relevant to the study of many fields in science, such as biology, chemistry,
physics, computer processors and cryptography, and many more
General One-Dimensional CA
 Grid

 Initial state

 Neighborhood

 Rules
Average,
Sum
…
Wolfram’s Elementary CA
 Simplest possible model
 One-dimensional cellular automaton 
 The standard Wolfram model is to initialize the grid with all cells assigned with 0, except for
the middle cell which is assigned with 1

 The neighborhood is the two immediate neighbors

 Rules
i

i+1
Wolfram’s Elementary CA

 Set of rules:

 These set of rules are called “Rule 90”

 Wolfram’s Elementary CA – Rule 90

 Rules: XOR of the two neighboring values

 "the simplest non-trivial cellular automaton“ (Martin, Odlyzko & Wolfram, ‘84)
 Interesting properties:
 When started from a random initial configuration, its configuration remains random at each time
step
 Any configuration with only finitely many nonzero cells becomes a replicator that eventually fills
all of the cells with copies of itself
 Additive cellular automaton: if two initial states are combined by computing the XOR of each their
states, then their subsequent configurations will be combined in the same way
Wolfram’s Elementary CA – Classes

 There are only 256(= 28) different ways for a Wolfram Elementary CA can be defined!

? ? ? ? ? ? ? ?
 They can be categorized into 4 classifications:
 CA which rapidly converge to a uniform state (rule 222)
 CA which rapidly converge to a repetitive or stable state (rule 190)
 CA which appear to remain in a random state (rule 30)
 CA which form areas of repetitive or stable states, but also form structures that interact with each
other in complicated ways – a complex system (rule 110)
 These are thought to be computationally universal, or
capable of simulating a Turing machine
 Wolfram’s Elementary CA – Rule 110

  

 From the ‘Complexity’ class

 Interesting behavior on the boundary between stability and chaos
 Rule 110 is known to be Turing complete. This implies that, any calculation or computer
program can be simulated using this automaton (like calculus)
 Rule 110 is arguably the simplest known Turing complete system
 This result therefore provides significant support for Wolfram's view that class 4 systems are
inherently likely to be universal
Conway's Game Of Life - History

 von Neumann attempted to find a hypothetical machine that could build copies of itself
and succeeded when he found a mathematical model for such a machine with very
complicated rules on a rectangular grid
 The Game of Life emerged as Conway's successful attempt to drastically simplify von
Neumann's ideas and still achieve similar “lifelike” result
 Conway's Game of Life has the power of a universal Turing machine: that is, anything
that can be computed algorithmically can be computed within the game
 The game made its first public appearance in the October 1970 issue of Scientific
American, in Martin Gardner's "Mathematical Games" column
 15
GAME OF LIFE
 John H. Conway developed “the Game of Life” in the 1970’s.
Conway's Game Of Life
  A two-dimensional CA
 Each cell is initialized with a random state: 0 or 1.
 In this case, 0 means dead and 1 means alive
 Each cell will have a bigger neighborhood
 With nine cells as the neighborhood, we have 9 bits, or 512 possible configurations -
impractical to define an outcome for every single possibility (This is Moore neighborhood,
the total number of automata possible would be , or )
Conway's Game Of Life – The Rules

 A cell can die if

 The cell is alive and has fewer than two live neighbors (loneliness)
 The cell is alive and has more than three live neighbors (overpopulation)
 A cell can reborn if
 The cell is dead and has exactly three live neighbors (reproduction)
 A cell remains in its current state if
 The cell is alive and has two or three live neighbors
 The cell is dead and has anything other than three live neighbors
Life Rules

 Each step: cell lives or dies

 Alive Cell = 1, Dead cell = 0

 Three simple rules

 dies if # of alive neighbour cells =< 2 (loneliness)
 dies if # of alive neighbour cells >= 5 (overcrowding)
 lives is # of alive neighbour cells = 3 (procreation)
 Examples

Still Lifes

Block Beehive Loaf Boat

Oscillators

Blinker Toad Beacon

Glider
Examples
Rule Examples

• loneliness (dies if #alive =< 2)

• overcrowding (dies if #alive >= 5)

• procreation (lives if #alive = 3)

Life Patterns
Stable

block pond ship eater

Periodic

time = 1 time = 2

Moving

Time = 1 time = 2 time = 3 time = 4 time = 5

 23
CHARACTERISTICS
 Discrete lattice of cells.
 Homogeneity – all of the cells of the lattice are equivalent.
 Discrete states – each cell takes on one of a finite number of possible discrete states.
 Local interactions – each cell interacts only with cells that are in its local neighbourhood.
 24
CLASSES
 CA typically fall into 4 classes:
 Class 1: system freezes into a fixed state after a short time.
 Class 2: system develops periodic behaviours, which repeat continuously.
 Class 3: system becomes a periodic, in which it continuously changes in unpredictable ways.
 Class 4: system can develop in highly patterned but unstable ways.
Encryption
Rule 30 Randomness

Pattern produced by rule 30

25
Simple Encryption

A simple encryption scheme [9].

26
Repeated Encryption Sequence

Repeated encryption key [9].

27
Applying Rule 30

30:
Rule 30 Evolution
Rule 30 Ad Infinitum

While one side has repetitive

patterns, the other side appears
random.
Zoom of the Regular Region
Zoom of the Random Area
Other Rules
Rule Atlas (1)
Rule Atlas (2)
Rule Atlas (3)

 256 rules
 Same initial condition
Cellule Automata Encryption cont.

Encryption using rule 30 [9].

37
Randomness using RULE 30 (Wolfram)

 One can find the form of a row in the cellular automaton, if given some segment of
the encrypting sequence, corresponding to a particular column.
 Rule 30 is not additive and to determine the color of a cell from the colors of its
neighbor columns is the same as enumerating all possible initial conditions.

38
 Applications

 Simulation of biological processes

 Simulation of cancer cells growth
 Image processing
 Pattern recognition
 Simulations of social movement
 Urban modeling
 Universal computers (embedded Turing machines)
 Cryptography
 …

"Cellular automata are sufficiently simple to allow detailed mathematical analysis,

yet sufficiently complex to exhibit a wide variety of complicated phenomena.“ Stephen Wolfram (1983)
40
CONCLUSION
 Discrete dynamical system simulator.
 Allow for a systematic investigation of complex phenomena.
 Original models of fundamental physics.
 Instead of looking at the equations of fundamental physics, consider modelling them with CA.
 Can mimic complex operations
 Problem – How to find the exact CA rules which will model a particular application
Thank you