[Paper Review] Multi-Interface PLC / Wireless Network Simulation

Yvonne Anne Pignolet, Ilias Rinis+2 more|arXiv (Cornell University)|Sep 27, 2012

Power Line Communications and Noise8 references3 citations

TL;DR

This paper extends the Cooja network simulator for Contiki OS to support multi-interface communication, enabling concurrent simulation of wireless and medium-voltage (MV) power line communication (PLC) in a single network. It introduces a distance-dependent PLC medium model based on exponential signal attenuation and bit error rate, allowing accurate modeling of packet success rates in MV PLC networks, which is validated in a Smart Grid scenario with Ring Main Units equipped with dual PLC couplers.

ABSTRACT

Many communication networks consist of legacy and new devices using heterogeneous technologies, such as copper wire, optical fiber, wireless and power line communication (PLC). Most network simulators, however, have been designed to work well with a single underlying link layer technology. Furthermore, there are hardly any suitable models for network simulators of PLC. In this paper we present extensions of the Contiki OS network simulator Cooja: A device may support multiple interfaces accessing multiple PLC segments or wireless channels and a simple PLC medium model is introduced describing packet loss probability as a function of distance. We test our approach to simulate a Smart Grid scenario of Ring Main Units equipped with PLC devices.

Motivation & Objective

• To address the lack of multi-interface and realistic PLC network simulation in existing tools.
• To enable simulation of heterogeneous communication technologies—especially PLC and wireless—within a single network infrastructure.
• To support pre-deployment testing of Smart Grid networks using realistic PLC channel models.
• To model dual-interface PLC devices (e.g., Ring Main Units with two couplers) that can switch between network segments.

Proposed method

• Extends Cooja simulator to support multiple communication media simultaneously, with each interface linked to a distinct medium.
• Implements a MV PLC medium model based on exponential signal attenuation with distance, using a path-loss model.
• Uses the Directed Graph Radio Medium (DGRM) as a backend, with success probabilities computed from distance-dependent SNR and BER.
• Models PLC communication as a complete graph (clique) for each connected powerline segment, with path distance computed as the shortest path between source and destination.
• Introduces a PLC configurator in Cooja to define powerline topology and automatically generate DGRM-compatible graphs.
• Applies interference thresholds to simulate collisions during simultaneous transmissions above a success rate threshold (e.g., 5%).

Experimental results

Research questions

• RQ1 How can a realistic, distance-dependent PLC channel model be integrated into a network simulator for multi-interface communication?
• RQ2 What is the impact of path distance on packet success rate in medium-voltage power line communication?
• RQ3 How can multi-interface devices (e.g., with dual PLC couplers) be effectively modeled in a single simulation environment?
• RQ4 Can the extended Cooja simulator accurately simulate Smart Grid scenarios involving both wireless and PLC networks?
• RQ5 How does the concurrent use of multiple communication interfaces affect network reliability and throughput in lossy environments?

Key findings

• The MV PLC model achieves a 50-byte packet success rate of 80% at 2 km distance, corresponding to a bit error rate (BER) of 5.6 × 10⁻⁴.
• The model predicts a maximum theoretical success rate of 1 − 10⁻⁶, corresponding to an SNR of 15.3 dB, representing the best-case scenario.
• Packet success rate decreases exponentially with distance due to path loss, with BER increasing as distance increases.
• The simulator successfully models dual-interface PLC devices in a Smart Grid topology, allowing communication across open switches or breakers.
• The model enables simulation of ring main unit networks with two PLC couplers per device, maintaining connectivity across open rings.
• The implementation supports concurrent operation of multiple media (e.g., wireless and PLC) with independent medium models and interference handling.