Dynamic Request Scheduling Optimization in Mobile Edge Computing for IoT Applications

Problem 1: Power Allocation

PA.m (PA function)

SubGrad.m (Subgradient function)

NCGG.m (Implementation of NCGG algorithm)

SA.m (Implementation of SA algorithm)

PSO.m (Implementation of Inertia Weighted Particle Swarm Optimisation algorithm)

Results

1. Energy consumption vs Number of Mobile Users (pmax = 5W).

2. Energy consumption vs Maximum Power (pmax).

3. Convergence Property of NCGG vs Number of Mobile Users (pmax = 5W, y axis is E, not Error of E in the paper).

Problem 2: Joint Request Offloading and Computing Resource Scheduling

JRORS.m (JRORS function)

Welfare_PSO.m (Implementation of Binary Particle Swarm Optimisation)

Results

1. Performance versus different number of mobile users ($I_q = 700KB$).

2. Performance versus different request workload ($U = 60, I_q = 700 KB$).

3. Performance versus different request input ($U = 60, w_q = 1500$ Magacycles).