MEC	Mobile edge computin
D2D	Device-to-device
MDP	Markov decision process
DQN	Deep Q network
AR	Augmented reality
VR	Virtual reality
UE	User equipment
BS	Base station
SCA	Successive convex approximation
GP	Geometric programming
DCN	Distributed computing node
MIP	Mixed integer programming
RL	Reinforcement learning
VFC	Vehicular fog computing
TD	Task device
RD	Resource device
D2D RD	D2D Resource device
Notations
$u_i$	The index of the TD i
$k_i$	The index of the RD i, where $k_0$ represents the BS, and the others represent D2D RD
$\mathrm{U}$	The number of TDs
$\mathrm{K}$	The number of D2D RDs
$\mathcal{U}$	The set of all TDs
$\mathcal{K}$	The set of all RDs
${\varphi }_i$	The index of the computing task on $i\in \mathcal{U}$
${\mathrm{Q}}_i$	The data size of the task ${\varphi }_i$
${\mathrm{C}}_i$	CPU cycles per bit required for task ${\varphi }_i$
${\tau }_i$	The maximum delay of the task ${\varphi }_i$
$f_i$	The local computing capacity of the TD i
${\mathrm{F}}_i$	The computing resource of the RD i
$R_{ij}$	The transmission rate between the TD i and the RD j
$B_{ij}$	The bandwidth allocated to the channel between TD i and RD j
$p_i^c$	The cellular transmission power from TD i to BS
$p_i^d$	The transmission power of D2D from TD i to a D2D RD
$p_i^{max}$	The maximum uplink power of TD i
$x_{ij}$	User association between TD i and RD j
${\alpha }_i$	The proportion of a computing task on TD i that is offloaded to the BS
${\beta }_i$	The proportion of a computing task on TD i that is offloaded to D2D RD
$D_i^{l,c}$	The local computation delay of the task on TD i
$D_i^e$	The delay of TD i to complete edge cloud computing
$D_i^{e,t}$	Cellular transmission delay of TD i
$D_i^{e,c}$	The computation delay of task ${\varphi }_i$ at BS
$D_i^D$	The delay of TD i to complete D2D RD computing
$D_i^{D,t}$	D2D transmission delay of TD i
$D_i^{D,c}$	The computation delay of task ${\varphi }_i$ at D2D RD
$D_i$	The total delay for completing the task ${\varphi }_i$ on TD i
$o_{u_i}$	The completion of the computing task on TD i