Review of Probability Theory Review of Probability Theory Experiments, Sample Spaces and Events Axioms of Probability Conditional Probability Bayes’s Rule Independence Discrete & Continuous Random Variables © Tallal Elshabrawy 2 Random Experiment It is an experiment whose outcome cannot be predicted with certainty Examples: Tossing a Coin © Tallal Elshabrawy Rolling a Die 3 Random Experiment in Communications Transmission of Bits across a Communication Channel v Waveform Generator x Channel y Waveform r Detection +A V. vi vi=1 vi=0 0 T 0 T xi -A V. + zi ]-∞, ∞[ yi 0 yi>0 yi<0 ri=1 ri ri=0 Why is this a random experiment? We do not know © Tallal Elshabrawy The amount of noise that will affect the transmitted bit Whether the bit will be received in error or not 4 Random Experiment in Networks Transferring a Packet across a Communication Network Packet Packet Why is this a random experiment? We do not know © Tallal Elshabrawy Whether the packet will reach the destination or not If the packet reaches the destination, how long would it take to get there? 5 Sample Space The set of all possible outcomes Tossing a coin Heads Tails S = {H,T} Rolling a die S = {1,2,3,4,5,6} The AWGN in a Communication Channel S = ] -∞, ∞ [ xi + yi zi © Tallal Elshabrawy 6 Event An event is a subset of the sample space S Examples Let A be the event of observing one head in a coin flipped two times A = {HT,TH} Let B be the event of observing two heads in a coin flipped twice B = {HH} © Tallal Elshabrawy 7 Axioms of Probability Probability of an event is a measure of how often an event might occur no. of sample pts in A P( A) no. of sample pts in S Axioms of Probability 1. 0 P A 1 2. P 0,P S 1 3. P A B P A +P B -P A, B © Tallal Elshabrawy 8 Example Let Event A characterize that the outcome of rolling the die once is smaller than 3 A = {1,2} P(A) = 2/6 = 1/3 Let Event B characterize that the outcome of rolling the die once is an even number B = {2,4,6} P(B) = 3/6 = 1/2 © Tallal Elshabrawy S A B 1 2 5 6 4 3 P A, B 1/ 6 P A B 1/3 1/ 2 1/ 6 9 Conditional Probability Probability of event B given A has occurred P A, B P B A P A Probability of event A given B has occurred P A, B P A B P B © Tallal Elshabrawy 10 Example Two cards are drawn in succession without replacement from an ordinary (52 cards) deck. Find the probability that both cards are aces Let A be the event that the first card is an ace Let B be the event that the second card is an ace P A, B =P A P B A 4 3 1 P A, B = 52 51 16 17 © Tallal Elshabrawy 11 Conditional Probability in Communications +A V. vi vi=1 vi=0 0 T 0 T xi -A V. yi + 0 yi<0 Pr[error v=1]= Pr[r=0 v=1] Pr[error v=1]= Pr[y<0 x=1] Pr[error v=1]= Pr[x+z<0 x=1] ri ri=0 0.8 0.7 0.6 0.5 © Tallal Elshabrawy yi>0 zi ]-∞, ∞[ Conditioned on v=1, what is the probability of making an error? ri=1 0.4 0.3 r=0 Decision Zone 0.2 0.1 0 -5 -4 -3 -2 -1 0 1 2 3 4 5 12 Bayes’s Rule P A, B P B A P A P A, B P A B P B P(B A)P( A) P( A B) = P(B ) © Tallal Elshabrawy 13 Theorem of Total Probability Let B1, B2, …, Bn be a set of mutually exclusive and exhaustive events. P( A) = ∑i =1 P A Bi P (Bi ) n © Tallal Elshabrawy ( ) 14 Bayes’s Theorem Let B1, B2, …, Bn be a set of mutually exclusive and exhaustive events. ( ) P(B A) = ∑ P(A B )P(B ) P A Bi P(Bi ) i n i =1 © Tallal Elshabrawy i i 15 Independent Events A and B are independent if P(B|A) = P(B) P(A|B) = P(A) P(A,B) = P(A)P(B) © Tallal Elshabrawy 16 Example Let A be the event that the grades will be out on Thursday P(A) Let B be the even that I will get A+ in Random Signals and Noise P(B) So What is the probability that I get A+ if the grades are out on Thursday P(B|A) = P(B) © Tallal Elshabrawy 17 Random Variable Characterizes the experiment in terms of real numbers Example X is the variable for the number of heads for a coin tossed three times X = 0,1,2,3 Discrete Random Variables The random variable can only take a finite number of values Continuous Random Variables The random variable can take a continuum of values © Tallal Elshabrawy 18 Bernoulli Discrete Random Variable Represents experiments that have two possible outcomes. These experiments are called Bernoulli Trials Associates values {0, 1} with the two outcomes such that P[X = 0] = 1-p P[X = 1] = p Examples Coin tossing experiment maps a ‘Heads’ to X = 1 and a ‘Tails’ to X = 0 (or vice versa) such that p=0.5 for a fair coin Digital communication system where X = 1 represents a bit received in error and X = 0 corresponds to a bit received correctly. In such system p represents the channel bit error probability © Tallal Elshabrawy 19 Binomial Discrete Random Variable A random variable that represents the number of occurrences of ‘1’ or ‘0’ in n Bernoulli trials The corresponding random variable X may take and values from {0, 1, 2, …, n} The probability mass function PMF for having k ‘1’ in n Bernoulli trials is P[X = k] = nCk pk(1-p)n-k Examples In a digital communication system, the number of bits in error in a packet depicts a Binomial discrete random variable © Tallal Elshabrawy 20 Geometric Discrete Random Variable Geometric distribution describes the number of Bernoulli trials in succession are conducted until some particular outcome is observed (lets say ‘1’) The corresponding random variable X may take and values from {1, 2, 3, …, ∞} The probability mass function PMF for having k Bernoulli trials in succession until an outcome of ‘1’ is observed P[X = k] = (1-p)k-1p Examples: In a communication network, the number of transmissions until a packet is received correctly follows a Geometric distribution © Tallal Elshabrawy 21