TrafficScheduling

Traffic Scheduling for Time Sensitive Networks

Run the program schedulegenwithdeadline.py

Overview of the approach

Traffic_Scheduling-GeneticAlgoApproach.pdf

A brief description of code structure

Scheduling being an NPHard problem many methods have been applied. Genetic Algorithms, Metaheuristics all have found its advantages. I found it easier with Genetic algorithms as finally all you do is try creating different populations and check the fitness (whether rules or constraints are met) more relatable. This is the approach in Genetic.

Inline image

An initial flow sequence is provided by JSON for 10 Links

The Traffic Model is here. The flows (4) are described in code here

Main method

Include calls to prepare initial polpulation
Multiple Iterations involved to find the best

Perform mutations (rearrangements) and check fitness

Identifying the best schedule by checking makespan. The Shortest Makespan is considered best

Here is where we check deadline. Main rule here is deadline