A Novel Behavior-Based Virus Detection Method for Smart Mobile Terminals

Algorithm 1

Virus detection algorithm.

Input:

  The traffic data instance 𝑋 collected on one day (m is the number of user subprofiles; n is

the number of instances in the user subprofile into which 𝑋 can be classified.);

Output:

  The answer to the question: “Is this terminal infected with virus?”;

(1)   for i = 1 to m do

(2)   Calculate the Similarity degree sim(𝑋, class(i)) between 𝑋 and user subprofile i;

(3)   for  i = 1 to m do

(4)   Find the maximum value MaxClassSim among all sim(𝑋, class(i)) and the

  corresponding class class(x) that yields MaxClassSim;

(5)   for  j = 1 to n do

(6)   Calculate the similarity degree sim(𝑋,𝑋_𝑗) between 𝑋 and the jth instance in class x;

(7)   for j = 1 to n do

(8)   Find the minimum value MinSim and the maximum value MinSim among all sim(𝑋,𝑋_𝑗);

(9)   if   MinSim > a&&MaxSim > b then (a, b are predetermined thresholds.)

(10)  Update subprofile x by including 𝑋 into it;

(11)   return there’s no virus;

(12)  else

(13)    return the terminal is infected with virus;

(14)  end if