Computer Science, Informatik 4 Communication and Distributed Systems Simulation “Discrete-Event System Simulation” Dr. Mesut Güneş
Computer Science, Informatik 4 Communication and Distributed Systems Chapter 5 Random-Number Generation
Computer Science, Informatik 4 Communication and Distributed Systems Purpose & Overview � Discuss the generation of random numbers. � Introduce the subsequent testing for randomness: • Frequency test • Autocorrelation test. Dr. Mesut Güneş Chapter 5. Random-Number Generation 3
Computer Science, Informatik 4 Communication and Distributed Systems Properties of Random Numbers Two important statistical properties: � • Uniformity • Independence. Random Number, R i , must be independently drawn from a uniform � distribution with pdf: ≤ ≤ ⎧ x 1 , 0 1 = f x ⎨ ( ) ⎩ 0 , otherwise 1 2 x 1 = ∫ 1 = = E R xdx ( ) 2 2 0 0 Figure: pdf for random numbers Dr. Mesut Güneş Chapter 5. Random-Number Generation 4
Computer Science, Informatik 4 Communication and Distributed Systems Generation of Pseudo-Random Numbers “Pseudo”, because generating numbers using a known method � removes the potential for true randomness. Goal: To produce a sequence of numbers in [ 0,1 ] that simulates, or � imitates, the ideal properties of random numbers (RN). Important considerations in RN routines: � • Fast • Portable to different computers • Have sufficiently long cycle • Replicable • Closely approximate the ideal statistical properties of uniformity and independence. Dr. Mesut Güneş Chapter 5. Random-Number Generation 5
Computer Science, Informatik 4 Communication and Distributed Systems Techniques for Generating Random Numbers Linear Congruential Method (LCM). � Combined Linear Congruential Generators (CLCG). � Random-Number Streams. � Dr. Mesut Güneş Chapter 5. Random-Number Generation 6
Computer Science, Informatik 4 Communication and Distributed Systems Linear Congruential Method To produce a sequence of integers, X 1 , X 2 , … between 0 and m-1 � by following a recursive relationship: = + = X aX c m i ( ) mod , 0 , 1 , 2 ,... + i i 1 The The The modulus multiplier increment The selection of the values for a , c , m , and X 0 drastically affects � the statistical properties and the cycle length. The random integers are being generated [ 0,m-1 ], and to convert � the integers to random numbers: X = = i R i , 1 , 2 ,... i m Dr. Mesut Güneş Chapter 5. Random-Number Generation 7
Computer Science, Informatik 4 Communication and Distributed Systems Linear Congruential Method – Example Use X 0 = 27 , a = 17 , c = 43 , and m = 100 . � The X i and R i values are: � X 1 = (17*27+43) mod 100 = 502 mod 100 = 2, R 1 = 0.02; X 2 = (17*2+43) mod 100 = 77, R 2 = 0.77 ; X 3 = (17*77+43) mod 100 = 52, R 3 = 0.52; X 4 =(17*52+43) mod 100 = 27, R 3 = 0.27; … Dr. Mesut Güneş Chapter 5. Random-Number Generation 8
Computer Science, Informatik 4 Communication and Distributed Systems Characteristics of a Good Generator Maximum Density � • Such that he values assumed by R i , i = 1,2,… , leave no large gaps on [0,1] • Problem: Instead of continuous, each R i is discrete • Solution: a very large integer for modulus m - Approximation appears to be of little consequence Maximum Period � • To achieve maximum density and avoid cycling. • Achieve by: proper choice of a , c , m , and X 0 . Most digital computers use a binary representation of numbers � • Speed and efficiency are aided by a modulus, m , to be (or close to) a power of 2 . Dr. Mesut Güneş Chapter 5. Random-Number Generation 9
Computer Science, Informatik 4 Communication and Distributed Systems Random-Numbers in Java � Defined in java.util.Random private final static long multiplier = 0x5DEECE66DL; private final static long addend = 0xBL; private final static long mask = (1L << 48) - 1; protected int next(int bits) { long oldseed, nextseed; ... oldseed = seed.get(); nextseed = (oldseed * multiplier + addend) & mask; ... return (int)(nextseed >>> (48 - bits)); } Dr. Mesut Güneş Chapter 5. Random-Number Generation 10
Computer Science, Informatik 4 Communication and Distributed Systems Combined Linear Congruential Generators Reason: Longer period generator is needed because of the � increasing complexity of simulated systems. Approach: Combine two or more multiplicative congruential � generators. Let X i,1 , X i,2 , …, X i,k , be the i-th output from k different multiplicative � congruential generators. • The j-th generator: - Has prime modulus m j and multiplier a j and period is m j - 1 - Produces integers X i,j is approx ~ Uniform on integers in [1 , m j – 1] - W i,j = X i,j - 1 is approx ~ Uniform on integers in [0 , m j -2 ] Dr. Mesut Güneş Chapter 5. Random-Number Generation 11
Computer Science, Informatik 4 Communication and Distributed Systems Combined Linear Congruential Generators • Suggested form: ⎧ X > i X , 0 ⎪ ⎛ ⎞ = ∑ k i ⎪ m ⎜ ⎟ − − − j X X m = 1 R ⎨ ( 1 ) mod 1 1 ⎜ ⎟ Hence, − i i j m , 1 i 1 ⎝ ⎠ ⎪ = = j X 1 1 , 0 ⎪ i m ⎩ 1 The coefficient: Performs the subtraction X i,1-1 − − − m m m ( 1 )( 1 )...( 1 ) • The maximum possible period is: = P k 1 2 − k 1 2 Dr. Mesut Güneş Chapter 5. Random-Number Generation 12
Computer Science, Informatik 4 Communication and Distributed Systems Combined Linear Congruential Generators Example: For 32-bit computers, combining k = 2 generators with � m 1 = 2,147,483,563 , a 1 = 40,014 , m 2 = 2,147,483,399 and a 2 = 20,692 . The algorithm becomes: Step 1: Select seeds X 1,0 in the range [ 1, 2,147,483,562] for the 1 st generator - X 2,0 in the range [ 1, 2,147,483,398] for the 2 nd generator. - Step 2: For each individual generator, X 1,j+1 = 40,014 X 1,j mod 2,147,483,563 X 2,j+1 = 40,692 X 1,j mod 2,147,483,399 . Step 3: X j+1 = ( X 1,j+1 - X 2,j+1 ) mod 2,147,483,562 . Step 4: Return ⎧ X + j > 1 X ⎪ , 0 + j 1 2,147,483, 563 ⎪ = R ⎨ + j 1 ⎪ 2,147,483, 562 = X ⎪ , 0 + j 1 ⎩ 2,147,483, 563 Step 5: Set j = j+1 , go back to step 2. • Combined generator has period: (m 1 – 1)(m 2 – 1)/2 ~ 2 x 10 18 Dr. Mesut Güneş Chapter 5. Random-Number Generation 13
Computer Science, Informatik 4 Communication and Distributed Systems Random-Numbers in Excel 2003 � In Excel 2003 new Random Number Generator ∈ X, Y, Z {1,...,300 00} = ⋅ X X 171 mod 30269 = ⋅ Y Y 172 mod 30307 = ⋅ Z Z 170 mod 30323 ⎛ ⎞ Z X Y = + + ⎜ ⎟ R mod 1.0 ⎝ ⎠ 30269 30307 30323 � It is stated that this method produces more than 10^13 numbers Dr. Mesut Güneş Chapter 5. Random-Number Generation 14
Computer Science, Informatik 4 Communication and Distributed Systems Random-Numbers Streams The seed for a linear congruential random-number generator: � • Is the integer value X 0 that initializes the random-number sequence. • Any value in the sequence can be used to “seed” the generator. A random-number stream: � • Refers to a starting seed taken from the sequence X 0 , X 1 , …, X P. • If the streams are b values apart, then stream i could defined by starting seed: ⎣ ⎦ = = S X i P K 1 , 2 , , − i b i b ( 1 ) • Older generators: b = 10 5 ; Newer generators: b = 10 37 . A single random-number generator with k streams can act like k distinct � virtual random-number generators To compare two or more alternative systems. � • Advantageous to dedicate portions of the pseudo-random number sequence to the same purpose in each of the simulated systems. Dr. Mesut Güneş Chapter 5. Random-Number Generation 15
Computer Science, Informatik 4 Communication and Distributed Systems Tests for Random Numbers Two categories: � • Testing for uniformity: H 0 : R i ~ U[0,1] H 1 : R i ~ U[0,1] / - Failure to reject the null hypothesis, H 0 , means that evidence of non- uniformity has not been detected. • Testing for independence: H 0 : R i ~ independently H 1 : R i ~ independently / - Failure to reject the null hypothesis, H 0 , means that evidence of dependence has not been detected. Level of significance α, the probability of rejecting H 0 when it is true: � α = P( reject H 0 | H 0 is true) Dr. Mesut Güneş Chapter 5. Random-Number Generation 16
Computer Science, Informatik 4 Communication and Distributed Systems Tests for Random Numbers When to use these tests: � • If a well-known simulation languages or random-number generators is used, it is probably unnecessary to test • If the generator is not explicitly known or documented, e.g., spreadsheet programs, symbolic/numerical calculators, tests should be applied to many sample numbers. Types of tests: � • Theoretical tests: evaluate the choices of m , a , and c without actually generating any numbers • Empirical tests: applied to actual sequences of numbers produced. - Our emphasis. Dr. Mesut Güneş Chapter 5. Random-Number Generation 17
Computer Science, Informatik 4 Communication and Distributed Systems Frequency Tests Test of uniformity � Two different methods: � • Kolmogorov-Smirnov test • Chi-square test Dr. Mesut Güneş Chapter 5. Random-Number Generation 18
Recommend
More recommend
