CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 51, 2016 

A publication of 

 
The Italian Association 

of Chemical Engineering 
Online at www.aidic.it/cet 

Guest Editors: Tichun Wang, Hongyang Zhang, Lei Tian 
Copyright © 2016, AIDIC Servizi S.r.l., 

ISBN 978-88-95608-43-3; ISSN 2283-9216 

Optimization Design of Computer Network Reliability Based 
on Genetic Algorithms 

Lijie Liu 
Jilin Engineering Vocational College, Siping, Jilin 136001, China 
liulijie9957@126.com 

With the acceleration of the process of information society, not only the user's computer communication 
networks is increasing, but also the rapid expansion of computer communication network connecting regional 
scale and network connections. Network reliability optimization of network design is a classic problem. Due to 
the complexity of the network reliability with the number of network nodes increases exponentially, it takes too 
much time to accurate calculation takes. It appears genetic algorithms, neural network, fuzzy neural network 
intelligent algorithm for solving this problem provides a new ideas and approaches. Due to the complexity of 
the network reliability with the number of network nodes increases exponentially, we want accurate calculation 
takes too much time, not even the result. Meanwhile, in order to make a better search algorithm performance, 
the paper also introduces the concept of co-evolution according to the schema theorem. The introduction of 
the test group, using a test group to retain better pattern, while the interaction between test groups and 
reconciliation group, thereby achieve the common purpose of evolution. Finally, the proposed algorithm 
simulation comparison, the results show that the algorithm has good convergence and search results. 

1. Introduction 

Computer communication network reliability concepts appear in the 1970s due to the rapid development of 
communications technology (Balog, 2007). Optical fiber technology, computer network technology, making the 
network function of growing, promote research computer communication network reliability problems continue 
depth direction (Putha and Sahoo, 2012).  
With the acceleration of the process of information society, not only the user's computer communication 
networks is increasing, but also the rapid expansion of computer communication network connecting regional 
scale and network connections (Bureerat, 2013). Due to a computer communication network is widely used in 
enterprises and institutions (Liu, 2015). Important areas of banking, transportation, communications, industrial, 
defense, computer communication network powerful complexity of its structure or not is directly proportional 
(Ali, 2013).  
In order to ensure the system work safely and reliably, theoretical research focused on the need for other 
indicators to optimize the design of the network reliability and improve network reliability based on the 
optimization of the system, such as the link cost, maintenance cost, delay, obstruction (Mohan, 2012). Rate 
and response time in practical application, the designer in the construction of the network system (Karaboga 
and Zhang, 2014).  
The main consideration is two aspects: performance and investment cost of the network due to network 
systems are generally costly, and therefore costs are often become a network topology design is a very 
important constraint (Gandomi, 2013). Currently, for most networks, as long as the small improvement of its 
topology design, you can save a certain amount of cost, "by re-optimizing the system performance without 
reducing, the application of new ideas, new technology, you can get considerable cost savings (Montalvo, 
2014).  
Based on a comprehensive study it has been proposed based on genetic algorithm computer communications 
network reliability analysis and optimization. To discuss the establishment of the reliability of the model and 
reliability analysis in theory, to provide the necessary engineering applications security service strategy, and 
network reliability study multi-objective optimization, in order to establish optimization algorithm based on the 

                               
 
 

 

 
   

                                                  
DOI: 10.3303/CET1651130

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Please cite this article as: Liu L.J., 2016, Optimization design of computer network reliability based on genetic algorithms, Chemical 
Engineering Transactions, 51, 775-780  DOI:10.3303/CET1651130   

775



method of construction of intelligent computer communication network reliability purposes, to give a specific 
implementation cases. Due to the complexity of the network reliability with the number of network nodes 
increases exponentially, which is a NP problem, you want accurate calculation takes too much time. Based on 
previous research on genetic algorithm is proposed based on the improved heuristic algorithm, genetic 
algorithm strong global search ability, with some of the local search strategy, to overcome the shortcomings of 
traditional methods at the same time has made good effect. 

2. The related theory and method 

2.1 Algorithm for computing network reliability 
Complete state enumeration method is to calculate the reliability of the network the most common and easiest 
way (Afonso and Gandomi, 2013). The main idea is: according to the requirements listed envoy network 
during normal operation of the S event occurs, all possible mutually exclusive events Bt, i=1,2,…n. Then S 
can be expressed as (Nguyen, 2014): 

1 2
( )

k
AS k A A A     (1) 

Therefore, the network reliability: 

   
1

Pr Pr
n

i

i

S B


   (2) 

For a given graph G (V, E) on each side there are two states: running or failure. If your network has old edges, 
then it would have 2E-1 states. For larger networks, with the increase in the number of edges, the number of 
states increases exponentially, so complete state enumeration method is only suitable for small computing 
network reliability (Yang, 2012). 
Such methods are mainly seeking network in accordance with the principle of inclusion-exclusion formula 
combinatorial mathematics reliability. Let Am is the talk events. By the inclusion-exclusion principle known, A1, 
A2, ... Am. The probability that at least one event is: 

   

       

1 2

1

1 2 1 2

Pr Pr

Pr Pr 1 Pr

m

m

i m

G A A A

A A A A A A


  

    
 (3) 

The topology of the network and the principle of inclusion-exclusion formula combining the introduction of the 
concept of free p circle graphs prove the inclusion-exclusion principle of two-terminal reliability of the formula 
not just destructive items and networks p circle diagram No-one correspondence, giving an inclusion-exclusion 
principle can be obtained directly in the formula is not destructive term formulas: 

     
1

Pr 1 Pr
j i

i

n v

a

i

G G
 

   (4) 

Wherein, J represents a network of non-G circle diagram, n, v denote the network G vertices and edges. This 
result makes the principle of inclusion and exclusion algorithm simplified. 

2.2 The basic principles of genetic algorithms 

Genetic Algorithm (GA) is a draw on natural selection and genetic mechanisms of thinking global random 
search algorithm. It is the solution of the problem might POPULATION, put every possible solution seen as 
individuals within a population, running the algorithm in the whole population random search space, according 
to a certain assessment strategies for each individual evaluation, constantly using the selection, crossover and 
mutation of these three genetic operators, the problem of evolving solution until an optimal solution .GA 
execution process was shown in Figure 1. 
GA basic elements including structure, population initialization code, select Options, genetic operations 
(crossover and mutation), the objective function and fitness function design, the termination condition 
selection. Among them, selection, crossover and mutation three operations is the core of genetic algorithms, 
Options are selected or eliminated based on individual fitness function value in the size of the parent to ensure 
that the direction of optimal search algorithm; crossover operator is the main method to produce new 
individuals, which determines the global search ability of GA; mutation operation is to generate new individual 
helper method, which determines the local search ability of GA. 

776



Parameter Set 
Determination

Encode the Parameter Set

Initialize 
Populations

Meet the 
conditions

Genetic Operator

Population P(t+1)

Population P(t)

Start

Evaluate 
Populations

1.Decode Parameters
2.Caculate Objective 
function
3.Mapped to fitness value
4.Adapt fitness value

1.Selcetion
2.Cross
3.Mutate
4.Other Operator

EndY

N

 

Figure 1: GA execution process 

3. Experiments and results 

3.1 The impact of network topology analysis 

Congenital factor programming computer communication network topology belong to computer 
communication networks, but also affect the reliability of computer communication network practice shows 
that: different application areas, different levels of scale computer communication networks must have 
different network topologies, otherwise improve computer communication network reliability is just empty talk. 
impact analysis of computer communication network topology of the computer network reliability of this 
chapter to be studied core interconnection network topology is the main connection computer communication 
networks between the components, which can be used diagram, graph theory is the study and therefore the 
most powerful mathematical tool interconnection network performance.  
Connected graph any edge or a node failure will bring down the network. Despite the low cost of such a 
computer communication network, but from the reliability point of view to consider its small fault tolerance, 
reliability is poor. For more important computer communications networks, not the use of such a computer 
communication network topology network topology bus structure shown in Figure 2 (a), the network topology 
star structure shown in Figure 2 (b), in addition to there is a ring network topology, as shown in Figure 2 (c). 

Node 1

Terminal 
controller

No
de

 2

Node n+1

Node n

Node 1

No
de

 2

Node n+1

Node n

Node 1 Node 2 Node n+1Node n

Terminal 
controller

...

...

...

(a)Bus 

(b)Star (c)Ring  

Figure 2: Computer network topology of the communication network 

777



3.2 Multi-objective optimization 
Significance of multi-objective optimization problem is to find problems one or more solutions to enable 
designers to accept all of the target value. Therefore, it is considered the single objective optimization problem 
is a multi-objective optimization problem in a special case, the various departments in the engineering, 
production management and national defense building society the problem encountered by most multi-
objective optimization problem. for example, in the design of computer communication network backbone, to 
consider how to make the general cost delay as small as possible, reliability and survivability to be as large as 
possible, which optimization is a three indexes can be said that multi-objective optimization problem in real life 
is abundant, and even everywhere. 
General mathematical form of multi-objective optimization problem as follows: 

         

   

   

1 2
min min , , ,

0 0,1, 2, , m
. .

0,1, 2, , m

T

m

i

i

V F x f x f x f x

g x i
s t

h x R i

  

  


 

 (5) 

Let optimization system parameters (variable) have n parameters for optimization constitute selected 
(variable) set x: 

  1 2, , , , 1, 2, ,n iX x x x x R i n    (6) 

X is the solution space optimization problem in which a set of parameters: 

        1 2, , , nx k x k x k x k  (7) 

Evaluation of optimization problem set performance indicators with m constituting performance indicators set 
H: 

 1 2, , , , 1, 2, , mm kH h h h h R i    
(8) 

3.3 Application of genetic algorithms in computer communications network reliability optimization 
In this paper, include a computer network reliability optimization simulation example calculation. Computer 
network nodes N = 6, the computer network node reliability constraint constant α = 2, ᵝ = 2, the number of 
iterations for the genetic manipulation 100, computer network link cost matrix C, and computer network 
reliability matrix R, respectively As follows: 

0

0 5 8 12 14 7

5 0 9 4 8 11

8 9 0 5 10 9

12 4 5 0 3 7

14 8 10 3 0 6

7 11 9 7 6 0

C

 
 
 
 

  
 
 
 
   

0

0 0.98 0.95 0.99 0.999 0.96

0.98 0 0.97 0.988 0.996 0.99

0.95 0.97 0 0.96 0.95 0.988

0.99 0.988 0.96 0 0.93 0.92

0.999 0.996 0.95 0.93 0 0.96

0.96 0.99 0.998 0.92 0.96 0

R

 
 
 
 

  
 
 
 
 

 

(9) 

778



Genetic algorithm according to flow and flow adjustment algorithm genetic algorithms, simulated gradually 
solved. When the genetic manipulation of the number of iterations is 100 times, the simulation process is 
terminated. The simulation solved a minimum cost of 45 computer network links, and to ensure that the 
computer maximum network reliability is 0.875. 
If the network is slightly more expensive than the average delay, then Wc = 0.8, Wd = 0.2, Wr = 0, optimal 
results are shown in Table 1; the results showed that the central node: 1,2, converted into the tree structure 
has three sides, They are: (3,1), (1,2), (2,4), 3,1,3,3,1,2,2,2,2 workstation. 

Table 1: Performance Indices Under different Weights and Optimized result by GA 

Weight Network Cost Reliability 
Cost 
Satisfaction 

Reliability 
Satisfaction 

Synthetical 
Satisfaction 

Wc=Wr=Wd=1/
3 

601 1 0.995 1 0.984 

Wc=Wr=0.5,W
d=0 

609 0.9354 0.974 0.962 0.995 

Wc=0.8 541 0.9137 1 0.935 0.996 

 
Through the above analysis shows that under these conditions the reliability of different weights, can get a 
better satisfaction can be said that these multi-objective optimization algorithm combined genetic and after, in 
the shortest possible time to find a satisfactory solution can be successful high reliability and low cost to solve 
the NP problem, and resolve quickly implement topology optimization computer communication network. 

4. Conclusions 

Due to the complexity of the network reliability with the number of network nodes increases exponentially, we 
want accurate calculation takes too much time, not even the result. It appears genetic algorithms, neural 
network, fuzzy neural network intelligent algorithm for solving this problem provides a new ideas and 
approaches. Due to the complexity of the network reliability with the number of network nodes increases 
exponentially, we want accurate calculation takes too much time, not even the result. It appears genetic 
algorithms, neural network, fuzzy neural network intelligent algorithm for solving this problem provides a new 
ideas and approaches. Meanwhile, in order to make a better search algorithm performance, the paper also 
introduces the concept of co-evolution, according to the schema theorem, the introduction of the test group, 
using a test group to retain better pattern, while the interaction between test groups reconciliation group, 
thereby achieve the common purpose of evolution. 

Acknowledgments 

This work is supported by Jilin Province Department of Education teaching reform of vocational education and 
adult education research in 2015: The Research on the Development and Application of Curriculum Resource 
Library of Higher Vocational “Computer Application Foundation”.(No. 2015ZCY198). 

Reference 

Afonso L.D., Mariani V. C., & dos Santos C.L., 2013, Modified imperialist competitive algorithm based on 
attraction and repulsion concepts for reliability-redundancy optimization. Expert Systems with 
Applications,40(9), 3794-3802. 

Ali A.M., Zendehboudi S., Lohi A., Elkamel A., & Chatzis I., 2013, Reservoir permeability prediction by neural 
networks combined with hybrid genetic algorithm and particle swarm optimization. Geophysical 
Prospecting,61(3), 582-598. 

779



Balog K, Rijke M, 2007, Finding similar experts. In: Proceedings of the 30th annual international ACM SIGIR 
conference on Research and development in information retrieval, Amsterdam, The Netherlands, p 23–27. 
doi:10.1145/1277741.1277926. 

Bureerat S., & Sriworamas K., 2013, Simultaneous topology and sizing optimization of a water distribution 
network using a hybrid multiobjective evolutionary algorithm. Applied Soft Computing, 13(8), 3693-3702. 

Gandomi A. H., Yang X. S., & Alavi A. H., 2013, Cuckoo search algorithm: a metaheuristic approach to solve 
structural optimization problems. Engineering with computers, 29(1), 17-35. 

Gandomi A. H., Yang X. S., Talatahari S., & Alavi A. H. (Eds.)., 2013, Metaheuristic applications in structures 
and infrastructures. Newnes. 

Karaboga D., Gorkemli B., Ozturk C., & Karaboga N., 2014, A comprehensive survey: artificial bee colony 
(ABC) algorithm and applications. Artificial Intelligence Review, 42(1), 21-57. 

Liu X., Croft W.B., Koll M., 2005, Finding experts in community based question-answering services. In: 
Proceedings of the 14th ACM international conference on information and knowledge management, 
Bremen, Germany, p 315–316. doi:10.1145/1099554.1099644 

Mohan B. C., Baskaran R., 2012, A survey: Ant Colony Optimization based recent research and 
implementation on several engineering domain. Expert Systems with Applications, 39(4), 4618-4627. 

Montalvo I., Izquierdo J., Pérez‐ García R., Herrera, M., 2014, Water Distribution System Computer‐ Aided 
Design by Agent Swarm Optimization. Computer‐ Aided Civil and Infrastructure Engineering, 29(6), 433-
448. 

Nguyen A. T., Reiter S., & Rigo P., 2014, A review on simulation-based optimization methods applied to 
building performance analysis. Applied Energy, 113, 1043-1058. 

Putha R., Quadrifoglio L., & Zechman E., 2012, Comparing ant colony optimization and genetic algorithm 
approaches for solving traffic signal coordination under oversaturation conditions. Computer‐ Aided Civil 
and Infrastructure Engineering, 27(1), 14-28. 

Sahoo L., Bhunia A. K., & Kapur P. K., 2012, Genetic algorithm based multi-objective reliability optimization in 
interval environment. Computers & Industrial Engineering, 62(1), 152-160. 

Yang X. S., Gandomi A. H., Talatahari S., & Alavi A. H. (Eds.)., 2012, Metaheuristics in water, geotechnical 
and transport engineering. Newnes. 

Zhang Z., Long K., Wang J., & Dressler F., 2014, On swarm intelligence inspired self-organized networking: its 
bionic mechanisms, designing principles and optimization approaches. Communications Surveys & 
Tutorials, IEEE, 16(1), 513-537. 

 
 
 

780