View
222
Download
0
Category
Preview:
Citation preview
7/31/2019 Ms 51 Complete
1/22
1. Solve the following L P problem through SIMPLEX method
Maximize Z = 5x1 + 6x2
Subject to
2x1 + 3x2 3000
5x1 + 7x2 1000
x1 + x2 500
x1 and x2 0
Introducing slack variables, x3 0, x4 0, x5 0, so that the constraints now become equations of
the form:
2x1 + 3x2 + x3 = 3000
5x1 + 7x2 + x4 = 1000
x1 + x2 + x5 = 500
Also, z 5x1 6x2 = 0
Now, constructing Table 1,
x1 x2 x3 x4 x5 z
2 3 1 0 0 0 3000
5 7 0 1 0 0 1000
1 1 0 0 1 0 500
_________________________________________
-5 -6 0 0 0 1 0
The pivot element should now be determined. For this, first the pivot column is selected which will
be the one with the most negative element in the objective row which is the fourth row.
1
7/31/2019 Ms 51 Complete
2/22
Here, -6 is the most negative element.
Now, divide the constant column, which is the last column after the vertical line, with the
corresponding values from the pivot column and select the smallest non-negative element. The row
with this smallest non-negative element is the pivot row.
The values from dividing the constant column with the pivot column are:
3000 / 3
1000 / 7
500 / 1
Here, 1000 / 7 is the smallest non-negative value. The pivot element will be the element at the
intersection of the pivot row and pivot column.
So from Table 1, the pivot element is 7.
x1 x2 x3 x4 x5 z
2 3 1 0 0 0 3000
5 7 0 1 0 0 1000
1 1 0 0 1 0 500
_________________________________________
-5 -6 0 0 0 1 0
Now the pivot column in Table 1 is made into a Unit Column by converting the Pivot Element to 1
and the other elements to 0.
R2 = R2 / 7
R1 = R1 3R2
R3 = R3 R2
R4 = R4 + 6R2
2
7/31/2019 Ms 51 Complete
3/22
Now constructing, Table 2,
x1 x2 x3 x4 x5 z
-1/7 0 1 -3/7 0 0 18000/7
5/7 1 0 1/7 0 0 1000/7
2/7 0 0 -1/7 1 0 2500/7
_________________________________________
-5/7 0 0 6/7 0 1 6000/7
Since there is a negative value in the objective row, the process should be carried out again by
finding the pivot element.
Here, -5/7 is the most negative element in the objective row and dividing the constant column with
the pivot column gives the following values:
-18000
200
1250
Here, 200 is the smallest non-negative value and thereby, from Table 2, the pivot element is 5/7.
x1 x2 x3 x4 x5 z
-1/7 0 1 -3/7 0 0 18000/7
5/7 1 0 1/7 0 0 1000/7
2/7 0 0 -1/7 1 0 2500/7
_________________________________________
-5/7 0 0 6/7 0 1 6000/7
Now the pivot column in Table 2 is made into a Unit Column by converting the Pivot Element to 1
and the other elements to 0.
3
7/31/2019 Ms 51 Complete
4/22
R2 = R2 / (5/7)
R1 = R1 + (1/7)R2
R3 = R3 (2/7)R2
R4 = R4 + (5/7)R2
Now, constructing Table 3,
x1 x2 x3 x4 x5 z
0 1/5 1 -2/5 0 0 2600
1 7/5 0 1/5 0 0 200
0 -2/5 0 -1/5 1 0 300
_________________________________________
0 1 0 1 0 1 1000
Now all the values in the objective row are non-negative.
Therefore, maximum value of the objective function,
z = 1000
when, x1 = 200
x2 = 0
x3 = 2600
x4 = 0
x5 = 300
4
7/31/2019 Ms 51 Complete
5/22
2. Write the Dual of the Problem 1 and give the economic interpretation of Dual Variables.
Problem 1: Maximize z = 5x1 + 6x2
Subject to:
2x1 + 3x2 3000
5x1 + 7x2 1000
x1 + x2 500
x1 and x2 0
The above problem is the primal problem in the standard form, which means that the primal
problem:
Considers maximization of the objective function.
Has less than or equal to type constraints.
Has non-negative constraints on the variables.
Since the above primal problem is in the standard form, the dual problem can be formulated by
using the following rules:
The number of constraints on the primal problem is equal to the number of dual variables.
The number of constraints in the dual problem is equal to the number of variables in the primal
problem.
The primal problem is a maximization problem while the dual problem is a minimization
problem.
The profit coefficients of the primal problem appear on the right hand side of the
constraints of the dual problem.
The primal problem has less than or equal to type constraints while the dual problem has
greater than or equal to type constraints.
5
7/31/2019 Ms 51 Complete
6/22
The coefficients of the constraints of the primal problem which appear from left to right are
placed from top to bottom in the constraints of the dual problem and vice versa.
Therefore, the corresponding dual problem of the primal problem 1, is,
Minimise 3000w1 + 1000w2 +500w3
Subject to:
2w1 + 5w2 + w3 5
3w1 + 7w2 + w3 6
w1 0, w2 0, w3 0
6
7/31/2019 Ms 51 Complete
7/22
3. Write the generalized form of transportation problem as LP Problem. Also explain how to
solve the degeneracy in transportation problem.
GENERALIZED FORM OF TRANSPORTATION PROBLEM
In order to obtain the generalized form of transportation problem, we shall make use of an
example. A manufacturer operates three factories and dispatches his products to four different
dealers. The table shows the total units that each factory can supply, the total units required at each
dealer and the cost of transportation of one unit from each factory to each dealer.
DEALERS
FACTORY 1 2 3 4 SUPPLY
1 2 2 2 4 1000
2 4 6 4 3 700
3 3 2 1 0 900
REQUIREMENT 900 800 500 400
The total units that factories 1, 2 and 3 can supply are 1000 units, 700 units and 900 units, denoted
by a1, a2 and a3, respectively.
The total units required at dealers 1, 2, 3 and 4 are 900 units, 800 units, 500 units and 400 units,
denoted by b1, b2, b3 and b4, respectively.
The cost of transportation of one unit from Factory 1 to Dealer 1 is 2, from Factory 1 to Dealer 2 is
2 and so on.
A transportation problem is formulated as an LP problem using variables with two subscripts.
Let,
X11 = Amount to be transported from Factory 1 to Dealer 1.
X12 = Amount to be transported from Factory 1 to Dealer 2, and so on until,
X34 = Amount to be transported from Factory 3 to Dealer 4
7
7/31/2019 Ms 51 Complete
8/22
Let,
C11 = Cost of Transportation of one unit from Factory 1 to Dealer 1, and so on until,
C34 = Cost of Transportation of one unit from Factory 3 to Dealer 4
Then the transportation problem can be formulated as:
Minimise C11X11 + C12X12 = ..C34X34
Subject to,
X11 + X12 + X13 + x14 = a1
X21 + X22 + X23 + X24 = a2
X31 + X32 + X33 + X34 = a3
X11 + X21 + X31 = b1
X12 + X22 + X32 = b2
X13 + X23 + X33 = b3
X14 + X24 + X34 = b4
X11 0, X12 0,.X34 0
The generalized form of the above transportation problem in Linear Programming problem can be
written as:
m nMinimise CijXij (where, m = number of factories, n = number of dealers)
i=1 j=1
n
Subject to Xij = ai, where i = 1 to mj=1
m
Xij = bj, where j = 1 to ni=1
Xij 0, where i = 1 to m, j = 1 to n
8
7/31/2019 Ms 51 Complete
9/22
DEGENERACY IN TRANSPORTATION PROBLEM
If a basic feasible solution of a transportation problem with m origins and n destinations has fewer
than m + n 1 positive Xij (occupied cells) the problem is said to be a degenerate transportation
problem.
Degeneracy can occur at two stages:
At the initial solution.
During the testing of the optimum solution.
A degenerate basic feasible solution in a transportation problem exists if and only if some partial
sum of availabilitys (row) is equal to a partial sum of requirements (column).
We shall make use of the previous example to solve the degeneracy in transportation problem. A
manufacturer operates three factories and dispatches his products to four different dealers. The
table shows the total units that each factory can supply, the total units required at each dealer and
the cost of transportation of one unit from each factory to each dealer.
DEALERS
FACTORY 1 2 3 4 SUPPLY
1 2 2 2 4 1000
2 4 6 4 3 700
3 3 2 1 0 900
REQUIREMENT 900 800 500 400
Here, S1 = 1000, S2 = 700, S3 = 900R1 = 900, R2 = 800, R3 = 500, R4 = 400
Since R3 + R4 = S3 so the given problem is a degeneracy problem.
Now we will solve the transportation problem by Matrix Minimum Method.
To resolve degeneracy, we make use of an artificial quantity, d.
The quantity, d, is so small that it does not affect the supply and demand constraints.
9
7/31/2019 Ms 51 Complete
10/22
Degeneracy can be avoided if we ensure that no partial sum of S i (supply) and Rj (requirement) are
the same. We set up a new problem where:
Si = Si + d i = 1, 2, ....., m
Rj = Rj
Rn = Rn + md
DEALERS
FACTORY 1 2 3 4 SUPPLY
1 2900 2100+d 2 4 1000 + d
2 4 6700d 42d 3 700 + d
3 3 2 15002d 0400+3d 900 + d
REQUIREMENT 900 800 500 400 + 3d
Substituting d = 0.
DEALERS
FACTORY 1 2 3 4 SUPPLY
1 2900 2100 2 4 1000
2 4 6700 40 3 700
3 3 2 1500 0400 900
REQUIREMENT 900 800 500 400
Initial basic feasible solution:
2 * 900 + 2 * 100 + 6 * 700 + 4 * 0 + 1 * 500 + 0 * 400 = 6700
Now degeneracy has been removed.
10
7/31/2019 Ms 51 Complete
11/22
4. Arrivals at an STD booth is considered poisson with an average time of 10 minutes
between them. Length of phone call is assumed exponential with mean 3 minutes. Determine
a. The probability that a person arriving at the booth will have to wait.
b. Average Length of queue that form time to time
c. Average utilization of STD booth.
Average arrival time = 1 = 10 minutes = 1 hour
6
Therefore, = 6 / hour
Average service time = 1 = 3 minutes = 1 hour
20
Therefore, = 20 / hour
Here, < , so a steady state solution exists.
a. The probability that a person arriving at the booth will have to wait
Average waiting time for a person arriving at the booth,
Wq = = 6
( - ) 20 (20 - 6)
= 6 x 1 hours
20 14
= 1.28 minutes
Therefore, the average waiting time for a person arriving at the booth is 1.28 minutes
11
7/31/2019 Ms 51 Complete
12/22
b. Average length of queue that form time to time
Expected number of customers in the queue,
Lq = 2 = 62
( - ) 20 (20 - 6)
= 36
20 x 14
= 9 / 70
Therefore, Lq = 9 / 70
c. Average utilization of STD booth
Average time spent in the booth,
Ls = 1
-
= 1
20 - 6
= 1 hour
14
= 4.28 minutes
Therefore the average utilization of the STD booth is 4.28 minutes
12
7/31/2019 Ms 51 Complete
13/22
5. Write short notes on:
i) Dynamic Programming
ii) Integer Programming
iii) Pure and Mixed Strategy
iv) Queue Parameters.
i. DYNAMIC PROGRAMMING
Dynamic programming is an optimization approach that transforms a complex problem into a
sequence of simpler problems. The main characteristic of dynamic programming is the multistage
nature of the optimization procedure and provides a general framework for analyzing many
problem types. Within this framework a variety of optimization techniques can be employed to
solve particular aspects of a more general formulation. Dynamic programming solves problems by
combining the solutions into sub-problems and can be used when a problem can be divided into
sub-problems and when these sub-problems are not independent. There are three important
characteristics of dynamic programming problems explained below.
Stages
The essential feature of the dynamic programming approach is the structuring of optimization
problems into multiple stages, which are solved sequentially one stage at a time. Although each
one-stage problem is solved as an ordinary optimization problem, its solution helps to define the
characteristics of the next one-stage problem in the sequence. Often, the stages represent different
time periods in the problems planning horizon but also, sometimes the stages do not have time
implications.
States
Associated with each stage of the optimization problem are the states of the process. The states
reflect the information required to fully assess the consequences that the current decision has upon
13
7/31/2019 Ms 51 Complete
14/22
future actions. The specification of the states of the system is perhaps the most critical design
parameter of the dynamic programming model. There are no set rules for doing this but the
essential properties that should motivate the selection of states are that the states should convey
enough information to make future decisions without regard to how the process reached the current
state and the number of state variables should be small, since the computational effort associated
with the dynamic programming approach is prohibitively expensive when there are more than two,
or possibly three, state variables involved in the model formulation.
Principle of Optimality
In dynamic programming terminology, each point where decisions are made is usually called a
stage of the decision-making process. At any stage, knowledge of the intersection currently
available is needed to be able to make subsequent decisions. Information that summarizes the
knowledge required about the problem in order to make the current decisions, such as the
intersection at a particular stage, is called a state of the decision-making process. On these terms,
the solution to the minimum delay problem involves what is referred to as the principle of
optimality.
Any optimal policy has the property that, whatever the current state and decision, the remaining
decisions must constitute an optimal policy with regard to the state resulting from the current
decision. To make this principle more concrete, the optimal-value function can be defined in the
context of the minimum delay problem.
Vn(Sn) = Optimal value or minimum delay over the current and subsequent stages or intersections,
given that we are in state Sn in a particular intersection with n stages to go.
A recursive relationship for computing the optimal value function by recognizing that, at each
stage, the decision in a particular state is determined simply by choosing the minimum total delay.
If we number the states at each stage as Sn = 1, which is the bottom intersection, up to Sn = 6 which
is the top intersection, then
14
7/31/2019 Ms 51 Complete
15/22
Vn (Sn) = Min { tn (Sn) + Vn 1(Sn 1)} , (1)
subject to:
Sn 1 = Sn + 1 if we choose up and n even,
Sn 1 = Sn 1 if we choose down and n odd,
Sn 1 = Sn otherwise,
Where, tn (Sn) is the delay time in intersection Sn at stage n.
ii. INTEGER PROGRAMMING
An integer programming problem is a mathematical optimization orfeasibility program in which
some or all of the variables are restricted to be integers. When formulating linearprograms it is
often found that, certain variables should have been regarded as taking integer values but, for the
sake of convenience, they are allowed to take fractional values reasoning that the variables were
likely to be so large that any fractional part could be neglected. While this is acceptable in some
situations, in many cases it is not, and in such cases we must find a numeric solution in which the
variables take integer values.
Problems in which this is the case are called integer programs and the subject of solving such
programs is called integer programming. Integer programming occurs frequently because many
decisions are essentially discrete in that one or more options must be chosen from a finite set of
alternatives. Problems in which some variables can take only integer values and some variables can
take fractional values are called mixed-integer programs.
Branch-and-Bound
Branch-and-bound is essentially a strategy of divide and conquer. The idea is to partition the
feasible region into more manageable subdivisions and then, if required, to further partition the
subdivisions. In general, there are a number of ways to divide the feasible region, and as a
15
http://en.wikipedia.org/wiki/Optimization_(mathematics)http://en.wikipedia.org/wiki/Optimization_(mathematics)http://people.brunel.ac.uk/~mastjjb/jeb/or/lp.htmlhttp://en.wikipedia.org/wiki/Optimization_(mathematics)http://en.wikipedia.org/wiki/Optimization_(mathematics)http://people.brunel.ac.uk/~mastjjb/jeb/or/lp.html7/31/2019 Ms 51 Complete
16/22
consequence there are a number of branch-and-bound algorithms. An integer linear program is a
linear program further constrained by the integrality restrictions. Thus, in a maximization problem,
the value of the objective function, at the linear-program optimum, will always be an upper bound
on the optimal integer-programming objective. In addition, any integer feasible point is always a
lower bound on the optimal linear-program objective value.
The idea of branch-and-bound is to utilize these observations to systematically subdivide the linear
programming feasible region and make assessments of the integer-programming problem based
upon these subdivisions.
Implicit Enumeration
A special branch-and-bound procedure can be given for integer programs with only binary
variables. The algorithm has the advantage that it requires no linear programming solutions. One
way to solve such problems is complete enumeration. List all possible binary combinations of the
variables and select the best such point that is feasible. The approach works very well on a small
problem, where there are only a few potential 01 combinations for the variables.
In general, though, an n-variable problem contains 2n 01 combinations; for large values of n, the
exhaustive approach is prohibitive. Instead, one might implicitly consider every binary
combination, just as every integer point was implicitly considered, but not necessarily evaluated,
for the general problem via branch-and-bound. Here, we adopt the opposite tactic of always
maintaining the 01 restrictions, but ignoring the linear inequalities. The idea is to utilize a branch-
and-bound (or subdivision) process to fix some of the variables at 0 or 1. The variables remaining
to be specified are called free variables.
Cutting Planes
The cutting plane algorithm solves integer programs by modifying linear-programming solutions
until the integer solution is obtained. It does not partition the feasible region into subdivisions, as
in branch-and-bound approaches, but instead works with a single linear program, which it refines
16
7/31/2019 Ms 51 Complete
17/22
by adding new constraints. The new constraints successively reduce the feasible region until an
integer optimal solution is found. In practice, the branch-and-bound procedures almost always
outperform the cutting-plane algorithm. Nevertheless, the algorithm has been important to the
evolution of integer programming.
In addition, even though the algorithm generally is considered to be very inefficient, it has
provided insights into integer programming that have led to other, more efficient, algorithms.
iii. PURE STRATEGY AND MIXED STRATEGY
A Pure Strategy in game theory, also known as Nash equilibrium is a fundamental concept in the
theory of games and the most widely used method of predicting the outcome of a strategic
interaction in the social sciences. A game in strategic or normal form consists of the following
three elements:
A set of players.
A set of actions or pure-strategies available to each player.
A payoff or utility function for each player.
The payoff functions represent each players preferences over action profiles, where an action
profile is simply a list of actions, one for each player. A pure strategy Nash equilibrium is an action
profile with the property that no single player can obtain a higher payoff by deviating unilaterally
from this profile. This concept can best be understood by looking at some examples.
Consider first a game involving two players, each of whom has two available actions, called A and
B. If the players choose different actions, they each get a payoff of 0. If they both choose A, they
each get 2, and if they both choose B, they each get 1. This coordination game may be represented
as follows, where player 1 chooses a row, player 2 chooses a column, and the resulting payoffs are
listed in parentheses, with the first component corresponding to player 1s payoff. The action
17
7/31/2019 Ms 51 Complete
18/22
profile (B,B) is an equilibrium, since a unilateral deviation to A by any one player would result in a
lower payoff for the deviating player. Similarly, the action profile (A,A) is also an equilibrium.
Mixed Strategy is another concept in game theory. A game in strategic form does not always have
a Nash equilibrium in which each player deterministically chooses one of his strategies. However,
players may instead randomly select from among these pure strategies with certain probabilities.
Suppose a consumer purchases a license for a software package, agreeing to certain restrictions on
its use. The consumer has an incentive to violate these rules. The vendor would like to verify that
the consumer is abiding by the agreement, but doing so requires inspections which are costly. If the
vendor does inspect and catches the consumer cheating, the vendor can demand a large penalty
payment for the noncompliance. The standard outcome, defining the reference payoff zero to both
vendor (player 1) and consumer (player 2), is that the vendor chooses Dont Inspect and the
consumer chooses to Comply. Without inspection, the consumer prefers to Cheat since that gives a
positive payoff, with resulting negative payoff to the vendor. The vendor may also decide to
Inspect and if the consumer complies, inspection leaves the consumers payoff unchanged, while
the vendor incurs a cost resulting in a negative payoff. If the consumer cheats, however, inspection
will result in a heavy penalty and still create a certain amount of hassle for player 1.
In all cases, player 1 would strongly prefer if player 2 complied, but this is outside of player 1s
control. However, the vendor prefers to inspect if the consumer cheats. If the vendor always
preferred Dont Inspect, then this would be a dominating strategy and be part of a unique
equilibrium where the consumer cheats. If any of the players settles on a deterministic choice, like
Dont Inspect by player 1, the best response of the other player would be unique, here Cheat by
player 2, to which the original choice would not be a best response. The strategies in Nash
equilibrium must be best responses to each other, so in this game this fails to hold for any pure
strategy combination.
iv. QUEUEING THEORY AND ITS PARAMETERS
18
7/31/2019 Ms 51 Complete
19/22
The main object of Queueing Theory is to develop formulae, expressions, or algorithms
for performance metrics, such as the average number of entities in a queue, mean time spent in the
system, resource availability, probability of rejection, and the like. The results from queueing
theory can directly be used to solve design and capacity planning problems, such as determining
the number of servers, an optimum queueing discipline, and schedule for service, number
of queues, system architecture, and the like. Besides making such strategic design decisions,
queueing theory can also be used for tactical as well as operational designs and controls. The
queueing parameters can be listed as:
Arrival pattern/process
The arrival process or input process to a queueing system is often measured in terms of the average
number of arrivals per unit of time or by the average time between successive arrivals. Customers
may arrive for service individually or in groups. Single arrivals are illustrated by customers
visiting a beautician, students reaching at a library counter, and so on. On the other hand, families
visiting restaurants, ship discharging cargo at a dock are examples of bulk, or batch, arrivals.
Customers can also arrive in the system at regular or irregular times, or they might arrive in a
random way. The queueing models wherein customers arrival times are known with certainty are
categorized as deterministic models and are easier to handle. On the other hand, a substantialmajority of the queueing models are based on the premise that the customers enter the system
stochastically, at random points in time. With random arrivals, the number of customers reaching
the system per unit might be described by a probability distribution. Generally, the queueing
models are based on the assumption that arrival pattern follows Poisson distribution.
Speed of service / service mechanism
In a queueing system, the speed with which service is provided can be expressed in either of two
ways, as service rate and as service time. The service rate describes the number of customers
serviced during particular time period. The service time indicates the amount of time needed to
service a customer. Service rates and times are reciprocals of each other and either of them is
sufficient to indicate the capacity of the facility. Thus, if a cashier can attend, on average, to 10
19
7/31/2019 Ms 51 Complete
20/22
customers in an hour, the service rate would be expressed as 10 customers per hour and service
time would be equal to 6 minutes per customer. Generally, however, we consider the service time
only. If these service times are known exactly, the problem can be handled easily. But, as generally
happens, if these are different and not known with certainty, we have to consider the distribution of
the service times in order to analyze the queueing system. Generally, the queueing models are
based on the assumption based on the assumption that service times are exponentially distributed.
Queue discipline
It refers to the manner by which customers are selected for service when a queue has formed. The
most common discipline that can be observed in everyday life is first come, first served order, or
first in-first out (FIFO), as it sometimes called some other in common usage are last in-last out
(LIFO), which is applicable to many inventory systems when there is no obsolescence in stored
units as it is easier to reach the nearest items which are the last in, and a variety of priority
schemes, where customers are given priorities upon entering the system, the ones with higher
priorities to be selected for service ahead of those with lower priorities, regardless of their time of
arrival to the system.
Number of servers
The number of servers or service channels refers to the number of parallel service stations which
can service customers simultaneously. There can be single servers or multiple server queueing
system. The two multichannel systems differ in that the first has a single queue, while the second
allows a queue for each channel. A barber shop with many chairs is an Example of the first type of
multichannel system, while number of queues before the railway counters is an example of the
second type of multichannel system.
System capacity
In some queueing processes there is a physical limitation to the amount of waiting room, so that
when the line reaches a certain length, no further customers are allowed to enter, until space
20
7/31/2019 Ms 51 Complete
21/22
becomes available by a service completion. These are referred to as finite queueing situation; that
is; there is a finite limit to the maximum queue size.
Size of the population
The source of customers for a queueing system can be infinite or finite. For example, all people of
a city or state (and others) could be the potential customers at a super bazaar. The number of
people being very large, it can be taken to the infinite. On the other hand, there are many situations
in business and industrial conditions where we cannot consider the population to be infinite-it is
finite. Thus, the ten machines in the factory requiring repairs and maintenance by the maintenance
crew would exemplify finite population.
REFERENCES
Dynamic Programming, [Online].
Available at: http://www.cs.berkeley.edu/~vazirani/algorithms/chap6.pdf
21
http://www.cs.berkeley.edu/~vazirani/algorithms/chap6.pdfhttp://www.cs.berkeley.edu/~vazirani/algorithms/chap6.pdf7/31/2019 Ms 51 Complete
22/22
John, R., A Dynamic Programming, [Online].
Available at: http://gemini.econ.umd.edu/jrust/research/papers/dp.pdf
Integer Programming, [Online].
Available at: http://web.mit.edu/15.053/www/AMP-Chapter-09.pdf
Wiesemann, W.Integer Programming, [Online].
Available at: http://www.doc.ic.ac.uk/~br/berc/IPlecture1.pdf
Mixed Strategy, [Online].
Available at: http://www.columbia.edu/~rs328/MixedStrategy.pdf
Nash Equilibrium, [Online].
Available at: http://www.columbia.edu/~rs328/NashEquilibrium.pdf
Game Theory, [Online],
Available at: http://www.cdam.lse.ac.uk/Reports/Files/cdam-2001-09.pdf
Queueing Theory, [Online],
Available at: http://www.win.tue.nl/~iadan/queueing.pdf
http://gemini.econ.umd.edu/jrust/research/papers/dp.pdfhttp://web.mit.edu/15.053/www/AMP-Chapter-09.pdfhttp://www.doc.ic.ac.uk/~br/berc/IPlecture1.pdfhttp://www.columbia.edu/~rs328/MixedStrategy.pdfhttp://www.columbia.edu/~rs328/NashEquilibrium.pdfhttp://www.cdam.lse.ac.uk/Reports/Files/cdam-2001-09.pdfhttp://www.win.tue.nl/~iadan/queueing.pdfhttp://gemini.econ.umd.edu/jrust/research/papers/dp.pdfhttp://web.mit.edu/15.053/www/AMP-Chapter-09.pdfhttp://www.doc.ic.ac.uk/~br/berc/IPlecture1.pdfhttp://www.columbia.edu/~rs328/MixedStrategy.pdfhttp://www.columbia.edu/~rs328/NashEquilibrium.pdfhttp://www.cdam.lse.ac.uk/Reports/Files/cdam-2001-09.pdfhttp://www.win.tue.nl/~iadan/queueing.pdfRecommended