Skip to main content
. 2019 Apr 8;19(7):1674. doi: 10.3390/s19071674
Algorithm 1 Reward function of the proposed method
1 Current load of connecting eNB: Lt+1
2 Ideal load that is control target value: Lideal
3 Available bandwidth of connecting eNB: ABW
4 Select action at time t: at
5 Episode end time: tend
6 Normalization variable: β
7 ifLt+1 Lidealthen
8 r1+tβ
9 else
10 if ABW = 0 and at = 4 then
11   r0
12 else
13   r(Lt+1 LidealLideal+tendtβ)
14 end if
15 end if
16 returnr