Self-Learning-Based Data Aggregation Scheduling Policy in Wireless Sensor Networks

Algorithm 1

Node behaviour on time slot scheduling.

 Data:

 N_R, SR_c, , , SR_h

1 whilea new frame starts to executedo

2  for i = 1; i ≤ N_R; i + + do

3   for k = 1; ; k + + do

4    .disable (SR_c(k));

5   end

6   t^R = Δε – greedy ;

7   SR_c .add(t^R);

8  end

9  SR_h.add(SR_c);

10  ST_c = Δε - greedy;

11  if SR_h.comparison() is not stablethen

12   Set the state of all slots as ACTIVE;

13  else

14   Set t_k as ACTIVE where , and set other slot as SLEEP

15  end

16  for j = 1; j ≤ F; j + + do

17   ifthe state of t_j is ACTIVEthen

18    if ST_c == t_j then

19     send packet and wait for ACK;

20     ifACK is receivedthen

21      Update (t_j) with rewards;

22     else

23      Update (t_j) with punishment;

24     end

25    else ifreceive from i^th upstream nodethen

26     Update (t_j) with rewards;

27    end

28    else if SR_c (i) == t_j and not receive any packet from i^th upstream nodethen

29     Update (t_j) with punishment;

30     Set t_j+1 to t_j+p as ACTIVE;

31    end

32   end

31  end

32 end