Conference PaperPDF Available

EigenTrust++: Attack Resilient Trust Management

Authors:

Abstract and Figures

This paper argues that trust and reputation models should take into account not only direct experiences (local trust) and experiences from the circle of “friends”, but also be attack resilient by design in the presence of dishonest feedbacks and sparse network connectivity. We first revisit EigenTrust, one of the most popular reputation systems to date, and identify the inherent vulnerabilities of EigenTrust in terms of its local trust vector, its global aggregation of local trust values, and its eigenvector based reputation propagating model. Then we present EigenTrust++, an attack resilient trust management scheme. EigenTrust++ extends the eigenvector based reputation propagating model, the core of EigenTrust, and counters each of vulnerabilities identified with alternative methods that are by design more resilient to dishonest feedbacks and sparse network connectivity under four known attack models. We conduct extensive experimental evaluation on EigenTrust++, and show that EigenTrust++ can significantly outperform EigenTrust in terms of both performance and attack resilience in the presence of dishonest feedbacks and sparse network connectivity against four representative attack models.
Content may be subject to copyright.
8th International Conference Conference on Collaborative Computing: Networking, Applications and Worksharing , Collaboratecom Pittsburgh, PA, United States, October 
  

       
     
   
       
         
          
         
       
         
         
         
       
      
      
        
        
       
       
      
      
         
       
   
 

        
      
       
    
       
       
       
       


  
    
      
    
        
          
  
         

  
         
    
         
     
        
       
        
         
        
         
     
      
         
      
     
   
        
          
 
    
         
          
   
       
     
       
        
       
   
    
       
         
      
        
        
         
         
  

        
   
        
        
         
         
      
      
        
         
        
 

      

         
          
         
       
          
        
          
    
       
         
     
        
         
          
      
        
      
        
       
      
        


        
        
        
     
   
   
       
     
       
       
      
        
      
        
       
     
        
   
       
       
         
        

       
        

      
    
      
         
        
      
    

 
  
          
        
       
      
        
     
 

          
        
        
       

       
    
A. EigenTrust Overview
       
        
        

      

 
        
         
         
          
        
          
        
         
       
          
         
   
       
        
        
          
 
          
        
     
   

     


 
  
        
          
       
     

           
            
   
       



         
            
          
          
    
        
          
           
         
  
 Threat Models and Trust based Service Selection Methods
         

    
 
Threat Model A. 
        
        

Threat Model 
       
          
   
        

Threat Model C.   
         
          
     
Threat Model D.     
        
    
         
        
          
       
  
Threat Model      
          
          
    
   
Threat Model     
       
       

        
    

 

  


 




   
 








   
 
 


 
 




           
Deterministic v.s. Probabilistic Service Selection. 
    

        
 
        
       
         
       
     

  
   
 
 
        
      
  
   
        
        
     
    
 Vulnebilities Inherent in EigenTrust
      
 
 
Local ust Rating.    
 


        
 
       
        
      
       
          
       
            

          
            
            
         
      
  
Feedback Credibility.     
        
 
    
        
          
  
   
     
        
 
          
       
        
         
         
    
      




    
      

           
        


        
         
     
        
        
           
       
        
         
         
      
       
  

  
          
       
Utilizing Circle of Friends.   
      
          
  

  
   
   
   
  
     
  
  
  






 
 
 
 


 




   
  


0
9
.
rtDFP=O%
-
 
  

�ig&nTrustDFP=10%      
08
E>genTrustDFP=2O%        
_Eig&nTrustDFP=3Q%      


    
05 


 
04  












Percentageofmalbouspeers(%)

       
         

        
   
     
         
        

           
         
 

           
 
      
           
 
    
   
       
        
   
         
   
     
         
        
        
          
           
       
         
           

 

 


     
    
        
     







       
  
             


   

 



 


         



  

   


            


      
        

   
        

            
 
         
          
           
       
     
     
        
         
       
   
          
      

 
 

        
    
          
        
          
          
   
  
       
    
       
   
  
        
         

 
   
  
           
    
        
        
       
          
        
      
         
         
       
 
        
       
        
         
         
           
           
          

          
       

 

  
       
   

        

     
         
      
        
       
      
      
         

    

A. Attack Resilient Local Trust Computation
          
 
  
   
       
        
  
    
 
      
       
        
 
      


 
 
   


       
      
     



 
  
         
       
 Aggregating Local Trust Values Using Feedback Credibili
    
        
  
       

       
       
        
    
           

     
        
  
        
         
       
      
 
         
         
     
         
         
          
       
           
          
   
        
     
 
         
  
          

       
        


       
      
       

 
       
 



    


   


 
          
          
         

  

         
        
        
     
 
     

 

   

  
           
      
     

   
  
 
 

 

  
 

 
  
          
      
        
   
 
 

       
        
  
 


 
       

 Pbabilistic Trust Ppagation thugh Circle of Friends
         
        
         

 
         
        
       
        
         

       

         
       
      
        
       

     
Combining local trust and feedback similarity. 
         
        
         
       
       
       
        
       
 
         
  
 
     
      

 
         




          
  
      
        
       
Threshold-based Probabilistic Trust Propagation. 
        
       
 
            
           
       
        
        
        
      

 
         
       
        
          
     
     
  

       
        
 
         

            
       
           

       
        
     
           
         
    
       
          
        
        

        
        
         
        
    

  
 
      
 
        
       
          
        
  
       
     
  

 


[
  
_ 

    


          
 
  

 




 
  



  
 

       
 

 
           
      
   
_
    
  

          
 
Algorithm Complexity Analysis.   
        
     
         
    
        
        
      
         
         
      
         
   
       
 
     
         
   
 

        
       
       
       

 
A. Pameter Conguration
        
        
       
   
       
    
         
         
   
   
           
 
        
           
           
     
        
         
          
   

 

  
    
         
       
       
    
    
          
             

        

         
     

       
        
      
   
           

 Eectiveness of Feedback Credibili
   
       
        
      
      
   
       
         
    
         
      




 

 
  
    


   








 
    
       
  
 

  
   
 
 


  

 
 


         
      
    
   
      
         
        
 

         
       
         
       
       
      
        
        
         
         
  
 
     
   
     
 Eectiveness of Circle of Friends


  
      
        
  
 
        
     
       


    
      
  


      






   




   

        

   


      



        

        
      
         
          
       
     
      
  
     

  
  
        
        
        
   
       
         
          
       
        
         
       
        
  
  
       

     
         
     

         

 

      
        
      

      



   
       
        
   


        
       
       
   
        

  


           
      
 
          
         
    
        

           
      

         
       
  
   
 
         
     
          
 
       
    
         
          
        

      
     
 
        
 
  
     
      
 
             
        
 
        
        
        
     
 
        
         
        

        
  
   
         
     
  

... In the real-world communication networks, many attack behaviors exist, such as packet drop attacks, 17 gray-hole and black-hole attacks, [18][19][20] probabilistic service attacks, 6,21 bad-mouthing, 22 spoofing, free-riders, slandering, 23 denial of service, 11 ballot-stuffing attack, 24,25 on-off attack, 22,26 colluding attack, 6 and self-promoting. 21,27 According to their behavioral properties and interactive principles, we can analyze and summarize these misbehaviors into four attacks referring to the studies 6,21,27 : (i) Independent attack; (ii) collective attack; (iii) camouflage attack; and (vi) spy attack. ...
... In the real-world communication networks, many attack behaviors exist, such as packet drop attacks, 17 gray-hole and black-hole attacks, [18][19][20] probabilistic service attacks, 6,21 bad-mouthing, 22 spoofing, free-riders, slandering, 23 denial of service, 11 ballot-stuffing attack, 24,25 on-off attack, 22,26 colluding attack, 6 and self-promoting. 21,27 According to their behavioral properties and interactive principles, we can analyze and summarize these misbehaviors into four attacks referring to the studies 6,21,27 : (i) Independent attack; (ii) collective attack; (iii) camouflage attack; and (vi) spy attack. ...
... To articulate the independent and collusive malicious interactional behaviors, we define four categories of malicious behaviors through applying the trust-enabled interactive networks as research background wherein each participant has two roles, that is, service provider and feedback rater. In accordance with the communication principles and interactive-behavior properties as reported in previous works, 6,21,27,28 we define the four attack behaviors as follows. ...
Article
Full-text available
The interactive risks of different devices, serving as clients or servers, have increasingly attracted huge attention in various communication systems, such as wireless sensor networks, wireless communication networks, and mobile crowd‐sensing. At present, lots of countermeasures had been proposed and deployed accordingly. Nevertheless, the investigation on the interaction risks between different devices is still very limited to date. In this paper, we propose a novel adverse effect inference mechanism TAEffect for malicious behaviors of devices emerged in various decentralized and open communication systems/networks through network‐percolation theory. At first, four typical malicious interactional behaviors are mapped into four topologies, then, upon which a network influence‐inspired approach is employed to quantify the adverse effect. Finally, multifacet experiments using five real‐world datasets and a synthetic testbed are performed to validate the efficiency and effectiveness. The experimental results show our proposed approach is significant and rational to quantitatively calculate and qualitatively mirror the four kinds of malicious interactional behaviors in diverse misbehavior‐emerged communication systems. This work aims to explore how to evaluate the adverse effect for various malicious participants in an objective manner. First, four typical malicious interactional behaviors are defined and mapped into four corresponding network topologies, then, upon which an effective method to quantitatively calculate the adverse effect is figured out using network‐percolation theory.
... Many models have been derived from EigenTrust to try to solve these issues. Some integrate the similarity to evaluate the reliability of feedback like SimiTrust [13] or EigenTrust++ [14]. Others integrate new trust factors such as contribution quantity or a context and/or quality factors like PeerTrust [15], [16] or CuboidTrust [17]. ...
Article
Full-text available
Within large and growing human communities where interactions occur, trust is a key factor to consider. Computational trust models have then been widely studied since the 2000s targeting items ratings (e.g. in e-commerce) or M2M (e.g. in IoT network). Among these models, EigenTrust is today one of the most popular and studied ones. It provides a global reputation calculation and is efficient in distributed networks, but not fully satisfactory for human interactions. On the opposite, the Bi-lattice model is well suited for human networks interactions such as solidarity networks and/or human services exchange networks but is limited to local trust results. In this paper, we propose a new aggregator that extends the Bi-lattice model to enable a global reputation calculation. This new aggregator discovers the trust links from the member whose score is to be evaluated to every other members he is connected to on the trust network. It then computes the global reputation of this member based on these trust links. Furthermore, it enables a lightweight approach, as it is able to compute a global reputation based only on a partial knowledge of the trust network. Throughout the paper, the proposed aggregator is presented, evaluated and compared to Eigentrust to show its effectiveness.
... To minimize the influence of colluding activities, some peers are assumed as trusted. This RS is susceptible for peers' feedback manipulations, 39 and several changes have been introduced to increase its robustness 1,21,46 or adapt it to new scenarios. 28,49 Due to its performance, Eigentrust is a popular benchmark for evaluating other TRRs. ...
Article
Full-text available
The introduction of trust-based approaches in social scenarios modeled as multi-agent systems (MAS) has been recognized as a valid solution to improve the effectiveness of these communities. In fact, they make interactions taking place in social scenarios much fruitful as possible, limiting or even avoiding malicious or fraudulent behaviors, including collusion. This is also the case of multi-layered neural networks (NN), which can face limited, incomplete, misleading, controversial or noisy datasets, produced by untrustworthy agents. Many strategies to deal with malicious agents in social networks have been proposed in the literature. One of the most effective is represented by Eigentrust, often adopted as a benchmark. It can be seen as a variation of PageRank, an algorithm for determining result rankings used by search engines like Google. Moreover, Eigentrust can also be viewed as a linear neural network whose architecture is represented by the graph of Web pages. A major drawback of Eigentrust is that it uses some additional information about agents that can be a priori considered particularly trustworthy, rewarding them in terms of reputation, while the non pre-trusted agents are penalized. In this paper, we propose a different strategy to detect malicious agents which does not modify the real reputation values of the honest ones. We introduce a measure of effectiveness when computing reputation in presence of malicious agents. Moreover, we define a metric of error useful to quantitatively determine how much an algorithm for the identification of malicious agents modifies the reputation scores of the honest ones. We have performed an experimental campaign of mathematical simulations on a dynamic multi-agent environment. The obtained results show that our method is more effective than Eigentrust in determining reputation values, presenting an error which is about a thousand times lower than the error produced by Eigentrust on medium-sized social networks.
... Eigentrust++ [25] identifies several attack vectors and modifies Eigentrust to improve performance and reliability in the presence of malicious nodes. Eigen-trust++ is currently implemented in NEM [26]. ...
Article
Full-text available
A peer-to-peer cloud storage network implementing client-side encryption would allow users to transfer and share data without reliance on a third party storage provider. The removal of central controls would mitigate most traditional data failures and outages, as well as significantly increase security, privacy, and data control. Peer-to-peer networks are generally unfeasible for production storage systems, as data availability is a function of popularity, rather than utility. We propose a solution in the form of a challenge-response verification system coupled with direct payments. In this way we can periodically check data integrity, and offer rewards to peers maintaining data. We further propose a model for addressing access and performance concerns with a set of independent or federated nodes
... To begin, we assume that the reader is familiar with the standard equations for trust computation in EigenTrust (see [24]). We modify the equations to address its limitations and the inherent vulnerabilities identified by Fan et al. in their work, Eigentrust++ [29] to make it more attack resilient. The Eigentrust approach has been well established in the literature [24,[30][31][32], as an effective trust model in P2P systems due to its algorithmic approach to enhancing the overall system security by detecting malicious peers, making it easy to punish such behavior, and encouraging honest peers, thereby encouraging cooperative behavior and enforcing trust in the system. ...
Chapter
Full-text available
The voluntary gathering and pooling of personal data by individuals via legal fiduciaries called data cooperatives is gaining a lot of attention as an approach to secure data management. Data cooperatives and blockchain are an excellent combination since they share fundamental features like decentralization and democratic design. In this paper, we leverage the power of blockchain to design a trusted news-sharing system for social media. We prove our concept by implementing a consumer news coop network on the Ethereum blockchain where members can voluntarily pool news information about their neighborhood for their benefit and also receive incentives in the form of an improved reputation for sharing credible news stories. We enforce honest behavior among the participants by implementing a trust and reputation scheme based on EigenTrust. Our results show that the blockchain approach to implementing a data cooperative is efficient with respect to memory consumption, scalability, and cost while also providing improved trust among participants. Furthermore, the reputation mechanism is effective in ensuring that malicious participants are severely penalized and removed from the system, while honest participants are rewarded. This approach can be used in a much bigger setup like Twitter so that the credibility of a shared post can be verified by a consensus before being shared on the network, thereby mitigating the spread of misinformation.KeywordsBlockchainDistributed Ledger TechnologyData cooperativeReputation SystemSocial mediaNews sharing
Article
Full-text available
One of the most common types of malicious behavior in social networks is represented by collusion, which consists of a secret cooperation between two or more agents providing mutual, highly positive feedback to each other. This collusion creates misleading advantages for the involved agents, deceiving others and distorting the actual reputation perception of the colluding members. Although the well-known EigenTrust algorithm can be fruitfully used to detect colluded agents, two important issues arise which limit its effectiveness: 1) it requires input information about which agents can be a-priori considered particularly trustworthy; and 2) it is not designed to handle situations in which we have several, different groups of colluded agents. These problems lead EigenTrust, to produce a significant number of false positives in some real situations. In this paper, we address the aforementioned issues. We introduce an automatic procedure to provide EigenTrust with the necessary inputs, and we propose an appropriate algorithm that combines EigenTrust with a clustering process. This procedure groups agents based on their reputation scores to tackle the presence of different groups of colluded agents. Through experiments, we demonstrate that our method, while maintaining the same effectiveness as EigenTrust in detecting malicious agents, is significantly more capable of avoiding the generation of false positives.
Preprint
The emergence of Consumer-to-Consumer (C2C) platforms has allowed consumers to buy and sell goods directly, but it has also created problems, such as commodity fraud and fake reviews. Trust Management Algorithms (TMAs) are expected to be a countermeasure to detect fraudulent users. However, it is unknown whether TMAs are as effective as reported as they are designed for Peer-to-Peer (P2P) communications between devices on a network. Here we examine the applicability of `EigenTrust', a representative TMA, for the use case of C2C services using an agent-based model. First, we defined the transaction process in C2C services, assumed six types of fraudulent transactions, and then analysed the dynamics of EigenTrust in C2C systems through simulations. We found that EigenTrust could correctly estimate low trust scores for two types of simple frauds. Furthermore, we found the oscillation of trust scores for two types of advanced frauds, which previous research did not address. This suggests that by detecting such oscillations, EigenTrust may be able to detect some (but not all) advanced frauds. Our study helps increase the trustworthiness of transactions in C2C services and provides insights into further technological development for consumer services.
Article
Full-text available
E-commerce markets can increase their efficiency through the usage of intelligent agents which negotiate and execute contracts on behalf of their owners. The measurement and computation of trust to secure interactions between autonomous agents is crucial for the success of automated e-commerce markets. Building a knowledge sharing network among peer agents helps to overcome trust-related boundaries in an environment where least human intervention is desired. Nevertheless, a risk management model which allows individual customisation to meet the different security needs of agent–owners is vital.The calculation and measurement of trust in unsupervised virtual communities like multi-agent environments involves complex aspects such as credibility rating for opinions delivered by peer agents, or the assessment of past experiences with the peer node one wishes to interact with. The deployment of suitable algorithms and models imitating human reasoning can help to solve these problems.This paper proposes not only a customisable trust evaluation model based on fuzzy logic but also demonstrates the integration of post-interaction processes like business interaction reviews and credibility adjustment. Fuzzy logic provides a natural framework to deal with uncertainty and the tolerance of imprecise data inputs to fuzzy-based systems makes fuzzy reasoning especially attractive for the subjective tasks of trust evaluation, business-interaction review and credibility adjustment.
Conference Paper
Full-text available
Trust is an essential component for secure collaboration in uncertain environments. Trust management can be used to reason about future interactions between entities. In reputation-based trust management, an entity's reputation is usually built on ratings from those who have had direct interactions with the en- tity. In this paper, we propose a Bayesian network based trust management model. In order to infer trust in di erent aspects of an entity's behavior, we use multi- dimensional application specific trust values and each dimension is evaluated us- ing a single Bayesian network. This makes it easy both to extend the model to involve more dimensions of trust and to combine Bayesian networks to form an opinion about the overall trustworthiness of an entity. Each entity can evaluate his peers according to his own criteria. The dynamic characteristics of criteria and of peer behavior can be captured by updating Bayesian networks. Risk is explicitly combined with trust to help users making decisions. In this paper, we show that our system can make accurate trust inferences and is robust against unfair raters.
Article
Full-text available
The original publication is available at http://www.springer.com/ To what extent can social structure result from evolutionary processes as popposed to being deliberately organised? To begin to answer this questions five different but releated social simulations are reviewed, and a map of which mechanisms might results in what structures under what conditions being started. These show that different structures can be brought about by evolutionary processes based on the abilities and propensities of the individuals. The article ends with some challenges—to construct a credible simulations of more sophisticated structures: social group selection and self-organised value chains.
Article
Reputation-based trust systems provide important capability in open and service-oriented computing environments. Most existing trust models fail to assess the variance of a reputation prediction. Moreover, the summation method, widely used for reputation feedback aggregation, is vulnerable to malicious feedbacks. This paper presents a general trust model, called RLM, for a more comprehensive and robust reputation evaluation. Concretely, we define a comprehensive reputation evaluation method based on two attributes: reputation value and reputation prediction variance. The reputation predication variance serves as a quality measure of the reputation value computed based on aggregation of feedbacks. For feedback aggregation, we propose the novel Kalman aggregation method, which can inherently support robust trust evaluation. To defend against malicious and coordinated feedbacks, we design the Expectation Maximization algorithm to autonomously mitigate the influence of a malicious feedback, and further apply the hypothesis test method to resist malicious feedbacks precisely. Through theoretical analysis, we demonstrate the robustness of the RLM design against adulating and defaming attacks, two popular types of feedback attacks. Our experiments show that the RLM model can effectively capture the reputation's evolution and outperform the popular summation-based trust models in terms of both accuracy and attack resilience. Concretely, under the attack of collusive malicious feedbacks, RLM offers higher robustness for the reputation prediction and a lower false positive rate for the malicious feedback detection.
Article
It is argued that social-network based (or group-based) trust metric is effective in resisting various attacks, which evaluates groups of assertions “in tandem”, and generally computes peers’ reputation ranks according to peers’ social positions in a trust graph. However, unfortunately, most group-based trust metrics are vulnerable to the attack of “front peers”, which represents malicious colluding peers who always cooperate with others in order to increase their reputation, and then provide misinformation to promote actively malicious peers. In traditional reputation ranking algorithms, like Eigentrust and Powertrust, etc., front peers could pass most of their reputation value to malicious friends, which leads to malicious peers accruing an improperly high reputation ranking. This paper proposes an alternative social-network based reputation ranking algorithm called Poisonedwater, to infer more accurate reputation ranks then existing schemes, when facing front peers attack. Our contributions are twofold: first we design the framework of the Poisonedwater approach including the following three procedures: (1) the propagation of Poisoned Water (PW): through direct transactions or observations, several malicious users are identified, termed as the poisoned seeds, and the PW will iteratively flood from those poisoned seeds along the reverse indegree direction in the trust graph; (2) the determination of adaptive Spreading Factor (SF) from PW level: based on the logistic model, PW level will correspondingly shrink each peer’s adaptive SF, which can determine how much percentage of each peer’s reputation could be propagated to its neighbors, and can be regarded as indicative of the peer’s recommendation ability; (3) the enhanced group-based reputation ranking algorithm with adaptive SF which seamlessly integrates peers’ recommendation ability to infer the more accurate reputation ranking for each peer; second, we experimentally analyze the mathematical implication of the Poisonedwater approach, and investigate the effect of various parameters on the performance of Poisonedwater. Simulation results show that, in comparison with Eigentrust and Powertrust, Poisonedwater can significantly reduce the ranking error ratio up to 20%, when the P2P environment is relatively hostile (i.e., there exists a relatively high percentage of malicious peers and front peers).
Article
Trust and reputation systems represent a significant trend in decision support for Internet mediated service provision. The basic idea is to let parties rate each other, for example after the completion of a transaction, and use the aggregated ratings about a given party to derive a trust or reputation score, which can assist other parties in deciding whether or not to transact with that party in the future. A natural side effect is that it also provides an incentive for good behaviour, and therefore tends to have a positive effect on market quality. Reputation systems can be called collaborative sanctioning systems to reflect their collaborative nature, and are related to collaborative filtering systems. Reputation systems are already being used in successful commercial online applications. There is also a rapidly growing literature around trust and reputation systems, but unfortunately this activity is not very coherent. The purpose of this article is to give an overview of existing and proposed systems that can be used to derive measures of trust and reputation for Internet transactions, to analyse the current trends and developments in this area, and to propose a research agenda for trust and reputation systems.
Article
For buyers and sellers alike, there's no better way to earn one another's trust in online interactions.
Article
Social trust and recommendation services are the most popular social rating systems today for service providers to learn about the social opinion or popularity of a product, item, or service, such as a book on Amazon, a seller on eBay, a story on Digg or a movie on Netflix. Such social rating systems are very convenient and offer alternative learning environments for decision makers, but they open the door for attackers to manipulate the social rating systems by selfishly promoting or maliciously demoting certain items. Although a fair amount of effort has been made to understand various risks and possible defense mechanisms to counter such attacks, most of the existing work to date has been devoted to studying specific types of attacks and their countermeasures. In this article, we argue that vulnerabilities in social rating systems and their countermeasures should be examined and analyzed in a systematic manner. We first give an overview of the common vulnerabilities and attacks observed in some popular social rating services. Next, we describe three types of attack strategies in two types of social rating systems, including a comprehensive theoretical analysis of their attack effectiveness and attack costs. Three context-aware countermeasures are then presented: (i) hiding user-item relationships, (ii) using confidence weight to distinguish popular and unpopular items, and (iii) incorporating time windows in trust establishment. We also provide an in-depth discussion on how these countermeasures can be used effectively to improve the robustness and trustworthiness of the social rating services.
Article
In overlay networks, the interplay between network structure and dynamics remains largely unexplored. In this paper, we study dynamic coevolution between individual rational strategies (cooperative or defect) and the overlay network structure, that is, the interaction between peer's local rational behaviors and the emergence of the whole network structure. We propose an evolutionary game theory (EGT)-based overlay topology evolution scheme to drive a given overlay into the small-world structure (high global network efficiency and average clustering coefficient). Our contributions are the following threefold: From the viewpoint of peers' local interactions, we explicitly consider the peer's rational behavior and introduce a link-formation game to characterize the social dilemma of forming links in an overlay network. Furthermore, in the evolutionary link-formation phase, we adopt a simple economic process: Each peer keeps one link to a cooperative neighbor in its neighborhood, which can slightly speed up the convergence of cooperation and increase network efficiency; from the viewpoint of the whole network structure, our simulation results show that the EGT-based scheme can drive an arbitrary overlay network into a fully cooperative and efficient small-world structure. Moreover, we compare our scheme with a search-based economic model of network formation and illustrate that our scheme can achieve the experimental and analytical results in the latter model. In addition, we also graphically illustrate the final overlay network structure; finally, based on the group selection model and evolutionary set theory, we theoretically obtain the approximate threshold of cost and draw the conclusion that the small value of the average degree and the large number of the total peers in an overlay network facilitate the evolution of cooperation.