Hot Posts

7/recent/ticker-posts

The Complete Wholesale VoIP Infrastructure Guide: Building a Scalable Voice Network


A wholesale VoIP business can generate millions of minutes without having the right infrastructure to manage them profitably. The difference between a network that simply carries calls and one that supports sustainable growth lies in how switching, routing, billing, monitoring, carrier management and automation work together.

For operators building or expanding a wholesale voice business, wholesale VoIP infrastructure is the foundation behind every commercial transaction. It determines how traffic enters the network, how routes are selected, how carriers are managed, how customers are billed and how quickly teams can respond when something changes. DeNovoLab's Class 4 Fusion is positioned specifically around this integrated model, combining switching, routing, billing, monitoring, reporting, portals and automation within one Class 4 platform for termination and origination traffic. (denovolab.com)

This guide explains the major components of wholesale VoIP infrastructure and how they fit together. It also compares DeNovoLab's approach with platforms such as PortaSwitch from PortaOne and ProSBC from TelcoBridges so operators can understand the architectural differences before selecting a technology strategy.

Understanding Wholesale VoIP Infrastructure

What makes wholesale VoIP different?

Wholesale VoIP is fundamentally a traffic exchange business.

An operator purchases voice termination from carriers then routes customer traffic through those suppliers while managing the commercial difference between customer revenue and carrier cost.

A simplified workflow looks like this:

Customer traffic → Class 4 switching → routing decision → carrier termination → CDR generation → billing and reporting

Every stage has commercial consequences.

If routing selects an expensive carrier when a suitable lower-cost route is available then margin can decline.

If billing does not accurately rate traffic then revenue can leak.

If monitoring fails to identify degraded carrier performance then call quality can suffer.

If fraud controls are weak then abnormal traffic can create unexpected termination costs.

This is why wholesale VoIP infrastructure should be viewed as a connected operating environment rather than a collection of independent telecom tools.

The core components

A complete wholesale VoIP infrastructure typically needs:

  • Class 4 switching

  • Intelligent routing

  • Carrier and trunk management

  • Billing and rating

  • CDR management

  • Monitoring

  • Fraud protection

  • Capacity controls

  • Reporting

  • Customer portals

  • Vendor management

  • Automation

  • Backup and archival

  • API integration

DeNovoLab's Class 4 Fusion currently combines these functions within one platform. Its published product information specifically lists switching, routing, billing, monitoring, rate generation, CDR and PCAP backup, client and vendor portals and automated operational workflows. (denovolab.com)

Class 4 Switching: The Core of the Wholesale Voice Network

What does a Class 4 switch actually do?

A Class 4 switch primarily handles traffic between carriers, customers and other voice network endpoints.

Unlike a Class 5 environment that is typically associated with subscriber-oriented services, a Class 4 platform is designed around larger-scale traffic exchange and routing.

The switch needs to understand incoming SIP traffic then apply the appropriate rules before sending the call toward an outbound destination.

That makes switching the foundation upon which the rest of the wholesale VoIP infrastructure operates.

Capacity is more than a single number

Two important measurements are:

  • Calls Per Second or CPS: This measures how quickly calls can be established.

  • Concurrent sessions: This indicates how many active calls can exist simultaneously.

A platform might experience a high call-establishment rate while maintaining relatively short sessions. Another deployment may have lower CPS but much longer calls and therefore higher concurrent session requirements.

DeNovoLab currently publishes 42k CPS for Class 4 Fusion and describes its switch core as designed for high-volume carrier traffic. These are vendor-published capabilities rather than a guarantee for every deployment since actual performance depends on infrastructure configuration traffic patterns routing complexity and other variables. (denovolab.com)

Capacity planning should include headroom

Imagine a network normally operating at 60% of its practical capacity.

The remaining capacity provides room for traffic spikes, carrier failover or temporary operational changes.

Running constantly near maximum utilization is risky.

It is similar to a highway operating at full capacity. One accident can create congestion across the entire network.

A properly planned wholesale VoIP infrastructure should therefore consider:

  • Normal traffic

  • Peak traffic

  • Growth forecasts

  • Failover traffic

  • Carrier concentration

  • Customer concentration

  • Processing overhead

The objective is not merely to handle today's traffic.

It is to create enough capacity for tomorrow's traffic without turning every growth milestone into an emergency infrastructure project.

Routing: Turning Network Traffic Into Commercial Decisions

Why routing matters so much

Routing determines which carrier handles a call.

That means routing directly influences:

  • Termination cost

  • Call quality

  • Carrier utilization

  • Failover

  • Customer experience

  • Gross margin

A basic wholesale operation might use Top Down routing.

A more advanced network can use Least Cost Routing or LCR, percentage routing, round robin, QoS routing, capacity routing and other policies.

DeNovoLab's Class 4 Fusion supports LCR, QoS, percentage, round robin, top-down, priority, time-based and capacity constraint routing. It also supports routing and blocking based on factors such as ANI and DNIS. (denovolab.com)

Lowest cost does not always equal best route

Consider an illustrative example.

Carrier A costs $0.0040 per minute.

Carrier B costs $0.0045 per minute.

At first glance Carrier A looks better.

Now suppose Carrier A consistently performs worse for a particular destination.

Carrier B may provide stronger completion and stability.

The additional $0.0005 per minute is a commercial trade-off rather than an automatic disadvantage.

At 10 million minutes the difference would be:

10,000,000 × $0.0005 = $5,000

Whether that additional cost is worthwhile depends on the operator's customer requirements, route performance and margin strategy.

This is why sophisticated routing should evaluate more than price.

Dynamic routing adds another layer

DeNovoLab's routing documentation distinguishes static routing from dynamic routing. Its dynamic routing options include Least Cost Routing, Highest ASR Routing and Highest ACD Routing based on real-time traffic conditions and vendor cost for a specific destination. (cookbook.denovolab.com)

That creates a more adaptive routing model.

Instead of asking:

"Which carrier is cheapest?"

the system can be configured to ask:

"Which available carrier best fits the defined commercial and performance criteria right now?"

Carrier and Trunk Management

Wholesale VoIP depends on carrier relationships

A carrier may provide:

  • International termination

  • Domestic termination

  • DID origination

  • Toll-free traffic

  • Specialized destinations

  • Backup routes

Each supplier can have different rates, technical requirements and performance characteristics.

As the number of carriers increases, manually managing these relationships becomes increasingly difficult.

The infrastructure therefore needs a structured way to represent carriers and trunks.

Trunk groups provide operational structure

A trunk group can represent a logical collection of routes or network connections.

It can be used with routing rules and failover policies to determine how traffic should move through the network.

DeNovoLab currently lists trunk groups and failover among its Class 4 routing capabilities. (denovolab.com)

This becomes particularly important when a business works with multiple suppliers for the same destination.

Example: Carrier failure

Suppose Carrier A normally handles a major portion of traffic to a destination.

At some point its service becomes unavailable.

A resilient wholesale VoIP infrastructure should have alternative routes already defined.

Instead of manually changing hundreds of routes the platform can use configured failover logic to redirect traffic according to the operator's policy.

The exact behavior depends on configuration and deployment.

The broader principle is straightforward:

Carrier failure should be a routing event rather than a business catastrophe.

Billing and Rating: Turning Calls Into Revenue

Every call needs a financial interpretation

Wholesale VoIP does not end when a call disconnects.

The network needs to determine:

  • Who made the call?

  • Where did it terminate?

  • Which carrier handled it?

  • How long did it last?

  • What did the carrier charge?

  • What should the customer pay?

That is the role of rating and billing.

DeNovoLab's Class 4 Fusion includes rate decks, invoices, balances, credit limits and customer and vendor billing. It also supports A-Z traffic, US jurisdiction workflows and origination traffic. (denovolab.com)

Small pricing differences become significant at scale

Consider an operator processing 100 million minutes per month.

A billing discrepancy of just $0.0001 per minute would represent:

100,000,000 × $0.0001 = $10,000

This is an illustrative calculation rather than a prediction.

It demonstrates why billing accuracy matters in wholesale voice.

A fraction of a cent may appear insignificant on one call.

Across hundreds of millions of minutes it becomes a material business variable.

Customer and vendor billing must connect

Wholesale operators have two sides to the commercial equation.

They collect from customers.

They pay vendors.

The infrastructure should therefore help operators understand both sides.

DeNovoLab's current platform includes customer billing as well as vendor workflows covering supplier rates, CDRs, invoices and traffic settlement. (denovolab.com)

This can give finance and network teams a more connected view of route economics.

CDRs and Data Management

Why CDRs matter

A Call Detail Record contains information about a call transaction.

Depending on the system and configuration it can include information such as:

  • Calling number

  • Called number

  • Start time

  • Duration

  • Route

  • Carrier

  • Cost

  • Customer

  • Disconnect reason

CDRs support billing, reporting, troubleshooting, customer service and financial reconciliation.

DeNovoLab's Class 4 architecture includes CDR generation and storage as part of the broader switching and billing workflow. Its documentation describes a high-volume CDR data architecture with local and cloud storage options. (denovolab.com)

Data volume grows quickly

Suppose an operator processes 10 million calls per month.

That means potentially millions of individual records requiring storage and retrieval.

At 100 million calls the challenge becomes substantially larger.

This is why CDR infrastructure should be planned as part of the initial architecture rather than treated as an afterthought.

Historical data becomes a business asset

CDRs can help answer questions such as:

  • Which customers are generating the most traffic?

  • Which destinations are becoming more profitable?

  • Which carriers are producing higher failure rates?

  • How has traffic changed over time?

  • Where is unusual activity occurring?

That makes CDRs more than billing records.

They become a source of operational intelligence.

Monitoring and Fraud Protection

Visibility is essential at scale

Large wholesale networks can generate enormous amounts of traffic.

Operators need to understand what is happening across:

  • Customers

  • Vendors

  • Trunks

  • Destinations

  • ANIs

  • DNIS

  • CPS

  • ASR

  • ACD

DeNovoLab's monitoring platform describes monitoring from carrier and trunk level down to ANI and DNIS granularity. It also supports automated responses to low ASR or ACD and suspicious traffic. (denovolab.com)

Fraud is both a security and financial problem

Imagine an account that normally generates 50,000 minutes per day.

It suddenly generates 500,000 minutes toward expensive destinations.

The increase does not automatically prove fraud.

But it is an anomaly that deserves immediate investigation.

If the traffic is fraudulent then the operator may still face termination costs.

That means fraud protection is directly connected to financial risk.

DeNovoLab currently lists fraud blocking and traffic limits as part of Class 4 Fusion's security controls. (denovolab.com)

Automation makes monitoring more practical

Monitoring thousands of calls manually is impossible.

The objective is therefore to configure meaningful rules.

For example:

  • If ASR falls below a defined threshold → trigger a defined response.

  • If suspicious traffic is detected → block or investigate according to policy.

  • If a trunk reaches its configured capacity → apply the appropriate routing response.

DeNovoLab describes Class 4 Fusion as configurable to automatically act on network exceptions including traffic loops, low ASR or ACD and suspicious traffic. (denovolab.com)

Automation: The Difference Between Scaling Traffic and Scaling Operations

More traffic should not mean proportionally more manual work

A wholesale VoIP business may eventually manage:

  • Hundreds of carriers

  • Thousands of destinations

  • Large rate decks

  • Many customer accounts

  • Millions of CDRs

  • Multiple billing models

If every process remains manual then operational overhead can grow quickly.

DeNovoLab currently promotes automated rate generation, route testing, rate import, rate delivery, reporting, invoicing and fraud controls. (denovolab.com)

Example: Rate management

Imagine an operator works with 50 carriers.

Each carrier can provide thousands of destination rates.

A manual workflow might involve downloading files, validating them, importing them, applying margins and generating customer rates.

The repetitive workload can become substantial.

Automation turns that into a defined workflow.

The goal is not to remove people.

It is to allow people to focus on exceptions and commercial decisions rather than repetitive administration.

Automation creates operating leverage

If a task takes 20 hours every week then it consumes approximately 1,040 hours per year.

If automation eliminates 75% of the repetitive work then approximately 780 hours could potentially be redirected to higher-value activities.

Again this is an illustrative calculation.

The actual result depends on the operator's workflow.

The principle remains:

A scalable telecom platform should increase processing capacity faster than administrative complexity.

Portals and APIs Complete the Infrastructure

Customers increasingly expect self-service

Wholesale customers need visibility into their business.

They may want to access:

  • Usage

  • CDRs

  • Invoices

  • Balances

  • Payments

  • Reports

DeNovoLab's current platform provides client, vendor and agent portals. Its client portal supports invoices, payments and usage reports while the vendor portal provides access to supplier rates, CDRs, invoices and traffic settlement workflows. (denovolab.com)

This can reduce repetitive support requests while improving partner transparency.

APIs connect the telecom stack

No platform exists in complete isolation.

Operators may need integrations with:

  • CRM platforms

  • Payment gateways

  • Fraud systems

  • Number databases

  • Provisioning systems

  • Internal applications

  • Analytics platforms

DeNovoLab currently describes API integration for CDR, reporting and switch control and promotes API-driven workflows as part of Class 4 Fusion. (denovolab.com)

The API layer therefore becomes part of infrastructure strategy rather than merely a developer convenience.

Deployment and Infrastructure Architecture

Cloud versus self-hosted

Wholesale VoIP operators can have very different infrastructure requirements.

Some prefer direct control over servers and networking.

Others want cloud deployment.

DeNovoLab currently supports server and VM deployments as well as AWS and Google Cloud Marketplace options. Its deployment documentation also describes RPM installation and installation from source. (denovolab.com)

The right choice depends on:

  • Traffic volume

  • Security requirements

  • Internal engineering resources

  • Geographic distribution

  • Redundancy requirements

  • Cost model

  • Regulatory considerations

Distributed architecture can support resilience

A single switching instance creates a single operational dependency.

A distributed architecture can allow multiple switching and routing instances to operate within one facility or across separate facilities.

DeNovoLab's published architecture describes this distributed model and positions it as a way to increase redundancy and support larger-scale operations. (denovolab.com)

The architecture should ultimately reflect the operator's actual availability requirements.

Comparing Major Wholesale VoIP Infrastructure Approaches

DeNovoLab Class 4 Fusion

DeNovoLab Class 4 Fusion takes an integrated Class 4 approach.

Its current platform combines:

  • Switching

  • Routing

  • Billing

  • Monitoring

  • Reporting

  • Rate generation

  • CDR and PCAP backup

  • Client portals

  • Vendor portals

  • Automation

  • Fraud controls

It supports termination and origination traffic and currently offers a 500-port Community Edition for live-traffic evaluation. Its published commercial model lists additional port capacity at $0.50 per port per month and usage at $0.0003 per minute. (denovolab.com)

The key proposition is consolidation.

For an operator that wants core wholesale operations within one Class 4 environment this can reduce the need to assemble multiple disconnected systems.

PortaOne PortaSwitch

PortaOne PortaSwitch documentation describes PortaSwitch as a unified platform for telecom service providers, wholesale carriers, ISPs, MVNOs and NGN operators.

Its architecture combines PortaBilling for real-time converged billing and service provisioning with PortaSIP as a Class 4 and Class 5 SIP softswitch. PortaBilling also controls real-time routing and billing functions. (PortaOne Documentation)

PortaSwitch supports broader service-provider scenarios beyond pure wholesale voice. Its documentation also describes wholesale traffic exchange between isolated virtual environments which can allow one provider to host another operator on its infrastructure. (PortaOne Documentation)

This makes it particularly relevant for operators seeking a broader converged telecom business platform.

TelcoBridges ProSBC

TelcoBridges ProSBC takes a different approach centered on the Session Border Controller.

Its current product information states that ProSBC supports up to 1,000 call attempts per second and 60,000 concurrent sessions with high-availability clustering. It also supports deployment across VMware, KVM/Proxmox, AWS, Azure and bare-metal environments. (TelcoBridges)

The platform emphasizes network-edge security, IP-to-IP adaptation, routing, troubleshooting and voice-network interoperability. Its current documentation also describes up to 350,000 endpoint registrations and up to 1,024 Network Access Points in a single instance. (TelcoBridges)

This makes ProSBC particularly relevant when the primary requirement is a carrier-grade SBC rather than an integrated wholesale billing and business-management platform.

The strategic difference

The three platforms represent different infrastructure philosophies.

DeNovoLab Class 4 Fusion: integrated wholesale Class 4 operations with switching, routing, billing, monitoring and automation.

PortaSwitch: broader converged telecom platform combining Class 4/Class 5 switching with billing and service provisioning.

TelcoBridges ProSBC: carrier-grade SBC architecture focused on voice-network security, routing, interoperability and scalable session handling.

There is no universal winner.

The appropriate architecture depends on whether the operator prioritizes an integrated Class 4 business platform, a broader converged telecom environment or a specialized carrier-grade voice edge.

How to Design a Wholesale VoIP Infrastructure Strategy

Start with traffic requirements

Determine:

  • Current CPS

  • Peak CPS

  • Concurrent sessions

  • Monthly minutes

  • Destination distribution

  • Customer growth

  • Carrier growth

Do not design solely around today's average traffic.

Define the routing strategy

Decide whether your network requires:

  • LCR

  • QoS routing

  • Percentage routing

  • Priority routing

  • Time-based routing

  • Capacity routing

  • Failover

  • Dynamic routing

Build billing around actual commercial models

Determine whether customers require:

  • Prepaid billing

  • Postpaid billing

  • Per-minute rates

  • ANI-based rating

  • DNIS-based rating

  • Jurisdictional rating

  • DID billing

  • Port billing

DeNovoLab's current Class 4 Fusion documentation describes support for multiple rating approaches including ANI, DNIS, fixed and variable rates and US jurisdiction workflows. (denovolab.com)

Plan monitoring before launch

Define what constitutes abnormal behavior.

Establish thresholds for:

  • ASR

  • ACD

  • CPS

  • Capacity

  • Fraud indicators

  • Route performance

Automate repeatable processes

Identify the workflows that consume employee time without requiring strategic judgment.

Those are the first candidates for automation.

Design for failure

Test:

  • Carrier failure

  • Trunk failure

  • Server failure

  • Network interruption

  • Traffic spikes

  • Fraud events

A system should not be considered carrier-grade simply because it performs well during normal conditions.

Conclusion: Building the Right Foundation for Wholesale VoIP Growth

A complete wholesale VoIP infrastructure is much more than a Class 4 switch.

It is an interconnected operating environment where traffic processing, routing, billing, carrier management, monitoring, security, reporting and automation work together.

The switching layer provides the traffic engine.

Routing determines how calls move.

Billing turns network activity into revenue.

CDRs provide the underlying transaction data.

Monitoring reveals network behavior.

Fraud controls protect financial and operational resources.

Portals improve partner visibility.

APIs connect the platform to the wider business.

Automation allows the operation to grow without allowing repetitive work to grow at the same rate.

DeNovoLab's Class 4 Fusion is explicitly positioned around this integrated model. Its current platform brings switching, routing, billing, monitoring, reporting, portals, rate generation and automation together and supports A-Z termination, US traffic and origination workflows. (denovolab.com)

PortaOne's PortaSwitch provides a broader converged telecom architecture with PortaBilling and PortaSIP supporting billing, service provisioning, Class 4 and Class 5 switching and real-time routing. (PortaOne Documentation) TelcoBridges ProSBC provides a carrier-grade SBC alternative with scalable session capacity, network-edge security, routing and high-availability capabilities. (TelcoBridges)

The most important lesson is that infrastructure decisions should be made around the whole operating model.

A switch that handles traffic but requires multiple disconnected systems for billing, routing and monitoring can create operational complexity.

A platform that integrates those functions can provide a more connected foundation.

For operators evaluating their next infrastructure investment the question should therefore move beyond:

"How many calls can this platform handle?"

The better question is:

"Can this infrastructure help us manage more traffic, more customers, more carriers and more complexity without creating the same growth in operational overhead?"

That is the real benchmark for a modern wholesale VoIP infrastructure strategy.

Ready to build a more connected wholesale voice operation?

Explore DeNovoLab Class 4 Fusion to evaluate an integrated platform for switching, routing, billing, monitoring, automation and carrier operations.

Post a Comment

0 Comments