GenioSim

GenioSim is a simulation platform for evaluating resource management and orchestration algorithms in PON-enabled edge computing infrastructures.

The simulator was developed within the GENIO (Edge cloud platform enabled by a new intelligent OLT for GPON networks), an industrial R&D initiative exploring new orchestration architectures for edge workloads deployed over Passive Optical Networks (PON).

GenioSim is implemented as an extension of the widely used PureEdgeSim simulator.

Overview

GenioSim extends the PureEdgeSim framework to support the simulation of distributed edge infrastructures built on PON access networks.

The simulator introduces support for:

• hierarchical PON topologies including OLTs and ONTs
• heterogeneous network links (fiber, MAN, WAN)
• hybrid virtualization environments combining containers and virtual machines
• hierarchical orchestration models separating container placement and task offloading
• configurable workloads representing realistic edge applications

These capabilities enable the study of orchestration strategies and resource allocation policies in complex PON-enabled edge computing environments.

Installation

Clone the repository:

git clone https://github.com/dessertlab/GenioSim.git
cd GenioSim

Build the simulator:

mvn clean package

After compilation, the executable simulator will be generated in the target/ directory.

Running the Simulator

General configuration files used by the simulator are located in the PureEdgeSim/settings/ folder. Experiment-specific configuration files are located in PureEdgeSim/examples.ProgettoGenio/ProgettoGenio_settings/. You can customize the scenario to run by modifying the following files:

• applications_NameScenario.xml
• cloud_NameScenario.xml
• edge_datacenters_NameScenario.xml
• users.xml
• simulation_parameters_NameScenario.properties

Simulations are executed through the main class:

PureEdgeSim/examples.ProgettoGenio/ProgettoGenio.java

You can run the simulator either from the command line or through an IDE.

Running from the command line

java -cp target/geniosim.jar examples.ProgettoGenio.ProgettoGenio

Running from an IDE

1. Open the repository in IntelliJ IDEA or VSCode
2. Navigate to PureEdgeSim/examples.ProgettoGenio/ProgettoGenio.java
3. Run the main() method.
4. Simulation logs will appear in the console, and output results will be written to: /Output

Extension of PureEdgeSim

PureEdgeSim was selected as the base simulator due to its flexibility in modeling cloud and edge computing environments. GenioSim extends it with several new capabilities:

Network Modeling

• multiple link types (optical fiber, MAN, WAN)
• configurable latency and bandwidth parameters
• hierarchical PON infrastructure

PON Topology Support

GenioSim models realistic PON deployments including:

• Optical Line Terminals (OLTs)
• Optical Network Terminals (ONTs)
• edge devices connected through the access network

Hybrid Virtualization

The simulator supports hybrid environments where:

• infrastructure services run inside virtual machines
• application workloads run inside containers

Workload Modeling

Applications are modeled as streams of tasks with configurable parameters such as:

• maximum latency requirement
• task length (Millions of Instructions)
• task generation rate
• request size
• response size

These features allow modeling heterogeneous edge workloads and orchestration strategies.

Simulated Application scenarios

Based on the application domains of the GENIO project, GenioSim supports the following scenarios:

• Smart City – Connected smart lighting nodes deployed under the same OLT.
• E-Health – Multiparametric medical monitoring devices generating continuous health data.
• Smart Building – Video surveillance cameras performing analytics tasks.
• Sports Streaming – Users consuming live multimedia streams.
• Video Gaming – Interactive cloud-gaming sessions with strict latency constraints.
• Mixed Scenario – A multi-tenant configuration combining all applications into a single workload with 298 concurrent users, representing realistic heterogeneous demand.

Each scenario can be evaluated under different infrastructure configurations and CPU architectures.

Baseline Architecture for Comparison

To evaluate the benefits of GENIO architectures, GenioSim also supports a baseline configuration representing traditional edge computing deployments.

In this baseline:

• OLTs remain access aggregation points
• edge applications are deployed on remote servers
• OLTs reach edge servers through WAN links

This configuration allows comparing GENIO's architecture against conventional edge infrastructures.

Evaulation Metrics

GenioSim evaluates system performance through the following metrics:

• Task Latency - Time between the submission of a user request and completion of its processing.
• Task Success Rate (TSR) - Percentage of tasks completed within the simulation time and respecting the application SLA.

These metrics are collected for all simulated scenarios across different orchestration strategies.

Hierarchical Orchestration Architecture

GenioSim models a two-level orchestration architecture consistent with the GENIO system design.

Level 1 — Container Placement

The cloud orchestrator decides where application containers should be deployed across available nodes. This decision balances:

• resource availability
• infrastructure topology
• placement policies

Level 2 — Task Offloading

Edge orchestrators determine which container instance should execute each incoming task. This layer corresponds to the OLT level of the infrastructure.

DNS-Based Load Balancing

In GENIO deployments, multiple geographically distributed OLTs serve edge devices. GenioSim models this behavior through DNS-based load balancing. Each OLT has an associated DNS entity responsible for:

• mapping ONTs to OLTs
• balancing incoming workload requests

DNS nodes do not perform computation; they act purely as routing and load-balancing components.

How to Replicate an Experiment

All experiments described in the paper can be reproduced using the configuration files provided in:

PureEdgeSim/examples.ProgettoGenio/ProgettoGenio_settings/

Each scenario contains a dedicated folder with configuration files. Key files include:

applications_<scenario>.xml
cloud_<scenario>.xml
edge_datacenters_<scenario>.xml
users.xml
simulation_parameters_<scenario>.properties

These files define, respectively:

• application properties
• infrastructure configuration (cloud and edge datacenters nodes)
• user workload
• simulation parameters

To reproduce a specific experiment:

1. Select the desired scenario folder.
2. Open ProgettoGenio.java.
3. Set the variable exampleMode = "<ScenarioName>".
4. Select the desired placement and offloading algorithms.
5. Run the simulation.