Skip to main content

Table 1 Summary of notations

From: Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach

Notation

Description

\(\mathcal {M}\)

The set of mobile users

\(\mathcal {T}\)

Index set of time slots

hm(t)

Channel vector between user m and BS at slot t

y(t)

Received signal of BS at slot t

ρm

Normalized temporal channel correlation coefficient of user m

H(t)

Channel matrix from all the users to BS at slot t

gm(t)

User m’s ZF detection vector used by BS at slot t

γm(t)

User m’s receiving SINR at BS of slot t

Bm(t)

Queue length of user m’s task buffer at slot t

am(t)

Amount of task arrivals of user m in bit at slot t

fm(t)

CPU frequency scheduled for local execution of user m at slot t

Lm(Fm)

CPU cycles required per one task bit (allowable CPU-cycle frequency) at user m

po,m(t)

Transmission power of user m for computation offloading at slot t

do,m(t)

Data transmitted by user m for computation offloading at slot t

pl,m(t)

Power consumption of user m for local execution at slot t

dl,m(t)

Data processed by user m via local execution at slot t

Po,m(Pl,m)

Maximum transmission power (local execution power) of user m

τ0

Length of one time slot