γαῖα

AI based microwave transmission network synthesis and optimization

The first complete, truly microwave network design know‑how with global multi‑objective optimization at the network level.

Methodology

Evolutionary design for microwave networks — from Chaos to Gaia.

From Chaos to Gaia

In Greek mythology, Gaia was the primordial Mother Earth goddess who brought structure and stability to the world that emerged from Chaos. Inspired by this narrative, our approach follows a similar transformation: from an unorganized network state with undefined structure to optimized network architecture, where topology, node positions, frequency resources, and capacity are designed to meet given requirements.

From link‑level tools to network‑level intelligence

Traditionally, in microwave network design, computational tools have focused primarily on individual links rather than on the network as a whole. Consequently, key decisions regarding network formation have relied heavily on the designer’s expertise and practical experience. As a paradigm shift in design philosophy, we have reformulated the design task as a multi-objective optimization problem, thereby elevating computational assistance from the link level to the network level. This approach enables us to treat the network as an integrated system rather than merely as a collection of independent connections.

TRADITIONALGAIA APPROACH
Link-by-link designComplete network design
Manual topology constructionAutomated topology construction
Link-optimal frequency allocationGlobal frequency allocation
Post-design cost evaluationCost embedded in optimization
Single solutionPareto-optimal solution set
Engineer-dependent qualityObjective, verifiable metrics

Evolutionary Network Design Process

At the core of this methodology lies an evolutionary algorithm inspired by the principles of natural evolution. Starting from an initial population of candidate solutions, the algorithm iteratively evolves network configurations through successive generations, selecting and refining them according to predefined objectives.

01 — Input Data and Existing Network Definition
Import current network information, technical constraints, equipment parameters, geographic data, and predefined design requirements.
02 — Generation of Initial Network Population
Produce a diverse set of candidate network designs based on available resources, constraints, and predefined design rules.
03 — Performance Evaluation and Multi-Objective Assessment, Pareto Sorting and Population Selection
Evaluate each candidate network using defined fitness indicators, including cost, path length, and other quality metrics, rank candidate solutions according to multi-objective performance and retain the most promising designs while eliminating dominated solutions.
04 — Iterative Optimization Process
Repeat the generation, evaluation, and selection cycles until no further meaningful improvement can be expected.
05 — Pareto Front Generation
Produce a final set of Pareto-optimal network designs representing different trade-offs between competing objectives.
06 — Subjective Engineering Selection
Select the preferred solution based on strategic priorities, operational requirements, and professional judgment.

Services

Comprehensive capabilities for microwave network design and optimization.

What Gaia does differently

Every aspect of the design process — from topology selection to lifecycle cost optimization — is addressed within a unified optimization framework, representing a fundamentally different approach from conventional network planning. While the resulting design can be readily verified using standard tools, the true challenge is identifying the optimal combination of network elements under multiple constraints.

Natural Handling of Conflicting Objectives
Applies domain-specific multi-objective optimization algorithms designed to resolve conflicting network design objectives and identify optimal trade-offs across topology, capacity, frequency allocation, dimensioning, and cost-performance criteria.
Elevated Step in the Process
Extends computational intelligence beyond link-level analysis to comprehensive network-level design optimization, generating solutions aligned with predefined technical requirements and business objectives.
Objective Performance Metrics
Evaluates network solutions using multiple quality indicators, including total cost, path length, efficiency, capacity, and reliability metrics.
Global Integrated Cost Optimization
Treats cost as an integral optimization objective, enabling global trade-off analysis between technical performance and economic efficiency while supporting any customer-specific pricing model.
Future‑Proof Planning
Simulates long-term demand scenarios based on the current network state, enabling lifecycle-based planning and evolution analysis over a 10–12 year operational horizon.
Data-Driven Decision Making
Enables business decisions based on full engineering-grade network designs.
Pareto-Optimized Network Design Solutions
Generates a comprehensive set of Pareto-optimal network designs, enabling engineers to select the most suitable solution based on their priorities, trade-offs, and professional judgment.

Every cost component, in scope from day one

Cost evaluation is seamlessly integrated into the network design process itself. Unlike traditional approaches — where costs are assessed post‑design — our methodology embeds both OPEX and CAPEX, along with ROI, directly into the optimization framework.

Towers & Base Stations
Construction and rental fees of towers and base stations.
Equipment Procurement
Procurement costs of equipment for existing and newly deployed microwave links.
Software & Hardware Interventions
Retuning, polarization switching, redirectioning and reinstallation costs.
Dismantling & Reutilization
Dismantling and reutilization expenses.
Antenna Site Rental
Fees depending on antenna size and location.
Monthly OPEX
Monthly operational costs of microwave links.
Structural Maintenance
Structural reinforcements and periodic mandatory tower maintenance costs.
Other Costs
Additional non‑categorized costs.

Under the hood

The optimization framework integrates elements of classical microwave network design, graph theory, and numerical optimization techniques, enriched by a set of very advanced computational concepts.

Network Topologies
Supports tree, ring (inner or outer), and mixed topologies of arbitrary complexity.
Link Configurations
Supports a wide range of link configurations, including 1+0, 2+0 CP/XP, 4+0 CP/XP, and dual-band deployments.
Integration of Existing Networks
Existing network infrastructure can be seamlessly incorporated into the optimization process.
Interference Environment
Accounts for arbitrarily sized external networks and complex interference environments to ensure accurate performance evaluation.
Frequency Planning
Utilizes an exact and highly efficient global-scope algorithm for optimal polarization and frequency allocation.
Channel Frequency Management
Supports selectable channel frequencies, with optimization solutions based exclusively on available channel resources.
Equipment Sub-band Management
Equipment sub-bands are handled according to a configurable priority order.
Outage Calculations
Provides advanced outage analysis, including differential fading effects and outage correlation modeling.
Cost Framework
Supports flexible economic evaluation frameworks, including OPEX, CAPEX, ROI, and customer-specific pricing policies.
Calculation Standards
Based on the latest industry standards, including ITU-R P.530-19 and ITU-R P.452-18.
Coordinate System Support
WGS84-native raster and vector database support.
Network Scale
Designed to support large-scale networks with more than 1,000 connections.
Platform
Fully in-house developed software platform, enabling integration of customer-specific requirements and tailored optimization functionalities.

Contact

Let's talk about your network challenge.

Get in touch

Whether you're planning a new network, optimizing an existing one, conducting a feasibility study, or evaluating procurement decisions, we’re here to provide the expertise and answers you need.

Email
wsc@wsc.hu
Address
WS Engineering Kft.
Diófa u. 9.
2330 Dunaharaszti, Hungary
WORLDWIDE AVAILABILITY

GAIA supports global WGS84-based raster and vector online databases. We are also capable of directly generating geographic databases for any location through AI-based monocular height estimation, enabling global coverage and adaptation to projects anywhere on the globe.

Start a conversation
We respond within one business day.
Your information is used solely to respond to your inquiry. WS Engineering Ltd. does not share contact details with third parties.
✓   Message sent. We'll be in touch within one business day.