Research Article

Joint Optimization for MEC Computation Offloading and Resource Allocation in IoV Based on Deep Reinforcement Learning

Table 1

Notation description.

NotationDescription

LThe length of the selected road
The coverage range of the RSU
Set/number of RSUs
Set/number of vehicles
i,jThe vehicle index /the RSU index
The arrival rate of vehicles
Vehicle’s position/speed
Required computation resource/input data size/output data size/maximum tolerable delay of the computation task
The time available by the vehicle before leaving the communication range of RSU--
The equivalent speed of vehicles
The available uplink/downlink transmission rate of vehicle-i
The number of vehicles offloads task to MEC server via RSU-j
The probability of vehicle-i connects to the RSU-j in a random time slot
The duration of a time slot
The propagation delay
The success transmission period between vehicle-i and RSU-j
The bandwidth of RSU-j
The transmission power of vehicle-i
The channel gain between vehicle-i and RSU-j
The uplink/downlink transmitting time
The two-way transmission time between vehicle and RSU
The binary offloading strategy of vehicle-i/vehicles
The task execution time locally/in the MEC server
Total time for processing task locally/in the MEC server
Computing resource of the vehicle/allocated to MEC-j/the MEC server
The cost of vehicle-i locally/in MEC processing/vehicle-i under different task offloading decisions
The penalty for offloading failure
The weighted parameters of delay and computation resource cost
The task’s actual tolerant delay
The unit cost of the computing resource of the MEC server