Table 2 Notations used in FTTHDRL system architecture.
Notation | Meaning |
|---|---|
\(loa{d}_{{v}_{j}}\) | Running tasks on \(j\) VMs |
\(loa{d}_{p{n}_{k}}\) | Running tasks on \(k\) physical nodes |
\(tota{l}_{pr{o}_{v{m}_{j}}}\) | Total processing capacity of all \(j\) VMs |
\({ta}_{i}^{len}\) | Length of \(i\) tasks |
\({ta}_{i}^{pri}\) | Priorities of \(i\) tasks |
\({v}_{j}^{pri}\) | Priorities of \(j\) VMs |
\(Re{s}_{cost}\) | Resource cost in cloud environment |
\(ms\left(t{a}_{i}\right)\) | Makespan of \(i\) tasks |
\(Fai{l}_{rate}\) | Fault tolerance or rate of failures of \(i\) tasks |
\(avail\left({v}_{j}\right)\) | Availability of \(j\) VMs |
\(su{c}_{rate}\left({v}_{j}\right)\) | Success rate of \(j\) VMs |
\(tur{n}_{eff}\left({v}_{j}\right)\) | Turn around efficiency of \(j\) VMs |
\(trs{t}_{cp}\) | Trust on Cloud provider |
\(stat{e}_{t{a}_{i}}\) | State space of considered \(i\) tasks |
\(Ac{t}_{t{a}_{i}}\) | Action space of considered \(i\) tasks |
\(Re{w}_{fn}\) | Reward function |
\(\Delta\) | Rate of learning |
\(\omega\) | Discount factor |
\(rm\) | Replay memory |
