Network Deployment Optimizer

This project applies Mixed Integer Linear Programming (MILP) to optimize the deployment of wireless cellular networks. It explores and compares macro-only, micro-only, and mixed base station configurations over a 1500m × 1500m area using realistic radio propagation and throughput models.

Project Overview

• Optimization Objective: Minimize total deployment cost while ensuring full coverage and required throughput.
• Technology: MILP (via Python), Okumura-Hata path loss model, Shannon capacity theorem.
• Deployment Types: Macro, Micro, and Mixed base stations.

Mathematical Model

Sets

• C: Candidate Sites (50)
• T: Test Points (100)
• B: Base Station Types = {macro, micro}

Parameters

• cost[c,b]: Installation cost
• coverage[c,t,b]: Binary (0/1) signal coverage
• throughput[c,t,b]: Estimated Mbps throughput
• min_thr: Minimum throughput = 7 Mbps

Decision Variables

• install[c,b]: Whether to install BS type b at site c
• serve[c,t,b]: Whether BS at c serves test point t

Objective

Minimize total cost across selected deployments and assignments.

Simulation Results

Scenario	Total Cost	Macro BS	Micro BS	Deployed Sites
Macro-only	$9.24	4	0	[24, 26, 39, 42]
Micro-only	$7.32	0	12	[4, 6, 10, ..., 48]
Mixed	$6.77	1	9	Macro: [26], Micro: [1,...]

• 100% coverage in all scenarios
• All test points meet throughput requirement
• Mixed scenario achieves lowest cost with balanced performance

Visualizations

• Coverage maps for each scenario
• Throughput and cost comparison charts
• Network topologies for deployment strategies

How to Run

You can view and run the project on Google Colab:

Open Notebook