A Dijkstra Based Algorithm for Optimal Splitter Location in Passive Optical Local Area Network (POLAN)

Abstract

Passive Optical Local Area Networks (POLANs) are integral to modern broadband communication systems, offering high bandwidth and immunity to electromagnetic interference. Designing an efficient POLAN requires careful consideration of splitter placement to minimize network costs. This paper presents an algorithmic approach using Dijkstra's algorithm and the Google Maps API to optimize splitter locations in a POLAN. By treating Optical Network Terminals (ONTs) as nodes in a graph and calculating walking distances between them, the algorithm identifies potential splitter locations that minimize fiber length. Using the Dijkstra's algorithm, the total fiber length used to connect every optical network unit is approximately 274km. Finally, a simulation of the full PON network was carried out and the BER and Q-Factor for each ONU was gotten. An average BER value of 1.8e-11 and Q-Factor value of 13.3 was gotten.