A Mobility-Aware and Sociality-Associate Computation Offloading Strategy for IoT

Algorithm 1

AP-based mobility-aware and sociality-associate clustering algorithm (AP-MASACA).

Input:

  •The user set and edge servers set ;

  •The positions of edge servers ;

  •History connection relationship between and ;

Output:

  •: the partition of all nodes into clusters;

  •: the set of cluster centers;

 1: function KALMANFILTERING

 2:  Initialize and ;

 3:  for Each do

 4:   Time Update according to (19);

 5:   Measurement Update according to (20);

 6:  end for

 7:  Output mobility prediction results:

 8:  .

 9: end function

 10: function PROBABILITYCOMPUTATION

 11:  Build matrix and ;

 12:  Calculate , , then compute according to (22);

 13:  Calculate , , then compute according to (23);

 14:  Output and .

 15: end

 16: function APCLUSTERING,,,

 17:  Calculate the distances based on (24);

 18:  Calculate the similarities based on (25);

 19:  Initialize to be zero;

 20:  Initialize according to the type of nodes;

 21:  Update the responsibility based on (26);

 22:  Update the availability based on (27);

 23:  ;

 24:  Calculate ;

 25:  ;

 26:  Output and

 27: end function