Energy Efficiency Performance Improvements for Ant-Based Routing Algorithm in Wireless Sensor Networks

Algorithm 1

IEEABR algorithm pseudocode.

Function 1: Initialization

(1) //Function: Initialization

(2) //prototype of Procedures

(3) Next-Hop-Selection;

(4) Calculate-Pheromone;

(5) //initialize the global variable

(6) S = Source ID;

(7) //initialize probability distribution table

(8) 𝑁 = Number of nodes in network;

(9) 𝑁_𝑘 = the set of neighboring nodes of node 𝑘;

(10) 𝑛=𝑁_𝑘;

(11) dp = distribution probability;

(12) 𝑆_PDT [𝑛][dp] = Source probability distribution table;

(13)𝑃_id=1𝑁_𝑘; // 𝑃_id is the probability of jumping from node 𝑖 to node 𝑑

(14) //Initiate the routing table

(15) For (𝑖=0; 𝑖=𝑛; 𝑖++){

(16) [dp]·S_PDT←𝑃_id;

(17)}

(18) D = Destination ID (sink ID);

(19) FA [S, M[2], D] = forward ant [source ID, memory of forward ant, Destination ID];

(20) Ph = amount of Pheromone;

(21)𝐶 = initial energy;

(22) BA [S, M[2], Ph, D] = Backward Ant [Source ID, Memory of backward ant,

   Pheromone value, Destination ID];

(23)𝐸[] = visibility array;

(24)𝑟 = Intermediate node ID;

Function 2: Decision making

(25) //If the intermediate node is equal to the destination node,

   then calculate the pheromone and construct the backward ant

(26) L1:   If (𝑟=𝑑)

(27)       Calculate-Pheromone;

(28)       Construct the BA [S, M[2], Ph, 𝑑];

(29) L2:       Next-Hop-Selection;

(30)          If (𝑟=𝑠)

(31)       Eliminate BA [𝑠, M[2], Ph, 𝑑];

(32)            //update the routing table

(33)            𝜌=0.8

(34)            ∅=0.3

(35)            𝐵𝑑_𝑘  = number of visited nodes by the backward ant

(36)          𝜏=1−𝜌∗𝜏(𝑟,𝑠)Δ𝜏∅𝐵𝑑_𝑘

(37)          Goto L3;

(38)         Else

(39)          Goto L2;

(40)      Else

(41)       Next-Hop-Selection;

(42)       Goto L1;

(43) L3: End

Function 3: Next hop selection

(44) //procedure Next-Hop-Selection

(45) Proc Next-Hop-Selection {

(46)     𝑠 = next Intermediate node

(47)     𝑒 = actual energy;

(48)     𝑁 = Number of nodes;

(49)     𝐸_𝑠 = visibility of 𝑠;

(50)     𝐸_𝑠 = 1𝐶−𝑒_𝑠;

(51)     𝐸[]←𝐸_𝑠;

(52)      𝜏 = pheromone routing table;

(53)      𝑃(𝑟,𝑠) = probability of 𝑟 jump to s as a next hop;

(54)      𝛼=0.9

(55)      𝛽=0.2

(56)      𝑃(𝑟,𝑠) = [𝜏(𝑟,𝑠)]^𝛼∗[𝐸(𝑠)]^𝛽∑[𝜏(𝑟,𝑢)]^𝛼∗[𝐸(𝑠)]^𝛽;

(57)      𝑋 = number of neighbors which are located in the destination direction;

(58)      𝑃[𝑋] = array for storing probability amount of neighbors;

(59)      𝑃[𝑋]←𝑃(𝑟,𝑠);

(60)      𝑃_max = 0;

(61)      For (𝑖=0; 𝑖=𝑋; 𝑖++){

(62)           If (𝑃[𝑖]>𝑃_max)

(63)           𝑃_max=𝑃[𝑖];

(64)      }

(65)      𝑟=𝑠⋅𝑃_max;

(66)      If (𝑟∈M[]·FA)

(67)           Loop happens then eliminate FA;

(68)      Else

(69)           {LastV= Count the member of M[];

(70)            If (LastV = 2)

(71)                Delete M[i];

(72)                M[𝑖]←M[𝑖+1];

(73)                M[𝑖+1]←𝑟;

(74)           }

(75) }

Function 4: Update of source probability distribution table

(76) //update the source probability distribution table

(77) For (𝑖=0; 𝑖=𝑛; 𝑖++){

(78)      Check the RoutingTable · S;

(79)      Find nodes which could be 𝐷 simultaneously Then

(80)           𝑃_dk = 9𝑁_𝑘−54𝑁_𝑘²;

(81)           Update [dp]·S_PDT←𝑃_dk;

(82)     For other neighbor

(83)           𝑃_ik = 4𝑁_𝑘−54𝑁_𝑘²;

(84)           Update [dp]·S_PDT←𝑃_ik;

(85)      //check the link failure

(86)      If ([dp]·S_PDT=0) //it means the link is lost

(87)      {

(88)          𝑃𝑍=^id1−𝑃_id;

(89)          𝑃_id=𝑃_id∗(1+𝑍);

(90)          Update [dp]·S_PDT←𝑃_id;

(91)      }

(92)}

Function 5: Pheromone calculation

(93) // procedure Calculate-Pheromone

(94) Proc Calculate-Pheromone{

(95)      𝐸_min← Min.𝐸[];

(96)      𝐸_avg← Avg.𝐸[];

(97)      𝑁_𝑗 = number of visited nodes by forward ant

(98)      1Δ𝜏=𝐶−𝐸_min−𝑁_𝑗/𝐸_avg−𝑁_𝑗;

(99)      BA·Ph ←Δ𝜏

(100)}