Interkey Intelligent DPI

Traffic management for operators and ISPs

Network traffic management for telecom operators

Network traffic management at operator scale is a loop: identify what the traffic is, decide what should happen to it, apply the decision and check the result. Flow counts tell you how much traffic moved; deep packet inspection (DPI) tells you what it was. Interkey Intelligent DPI classifies traffic in real time at carrier scale, applies policies per application, subscriber group or cell, and shows the outcome in Live View with latency measured in seconds.

For capacity planning, policy and network operations teams at telecom operators and ISPs. Reviewed November 2026.

  • Classification comes first:without knowing what the traffic is, congestion is diagnosed by guesswork and policies cannot be verified.
  • Policy scope:per application, subscriber group or cell.
  • Policy actions:real-time deployment and adjustment, service-chain control, and blocking where your policy and applicable regulation require it.
  • Visibility:Live View from the network-wide view down to application and flow level, with policy execution feedback.

Why traffic management starts with classification

Most operators already know how much traffic crosses each link. What volume counters cannot say is which applications make up that traffic, so a congested cell or link gets the same answer as every other: more capacity, everywhere. With classification, capacity goes where a specific class of traffic is saturating a specific part of the network.

The same applies to commercial and regulatory rules. Fair-use rules, tiered plans and quality-of-service commitments all depend on identifying the traffic they refer to. Regulatory reporting assumes the operator can describe its own traffic.

Classification is also harder than it was. Encryption removes the payload signal, and many applications share the same common ports. Interkey Intelligent DPI classifies from flow behaviour (handshake structure, packet sizing and timing, and endpoint behaviour) rather than from port numbers or payload alone, including encrypted and tunnelled flows, without decrypting them.

The traffic management loop

The traffic management loop in Interkey Intelligent DPI
StepThe question it answersIn Interkey Intelligent DPI
1. IngestCan the platform take in our traffic?Designed for carrier-scale traffic across fixed and mobile networks, where volumes run from Tbps up to dozens of Tbps; throughput is sized for each deployment
2. ClassifyWhat is this traffic?An AI traffic model built into the engine core identifies the application of each flow, including encrypted and tunnelled flows
3. ActWhat should happen to it?Policies deployed and adjusted in real time, per application, subscriber group or cell; service-chain control; blocking where your policy and applicable regulation require it
4. ObserveIs the policy doing what it was approved to do?Live View: network-wide view with drill-down, trend analysis and real-time policy execution feedback
5. RefineWhat should change?Live View results feed back to refine models and policies

A refinement to a live policy is a change to a live network, so it should pass through the same approval as the original rule.

What classification supports, team by team

Traffic management jobs and what classification provides
JobWhat classification providesTeam
Capacity planningWhich applications drive load, where and when, from network- and application-level trendsCapacity planning
CongestionWhich class of traffic is saturating which part of the networkNetwork operations
Fair use and tiered plansIdentification of the traffic a rule refers to, with policies applied per subscriber groupPolicy and product
Quality-of-service commitmentsIdentification of the traffic a commitment refers toPolicy
Regulatory reportingA reporting layer for regulatory teamsRegulatory affairs and compliance
New servicesService chaining and flexible deployment of user-plane nodesService and network planning

Which enforcement actions your deployment uses, and how they connect to your policy and charging architecture and your 5G core, is defined with your team during scoping and documented in the integration plan.

Out-of-band first, enforcement second

Out-of-band and inline deployment
Out-of-bandInline
Receives a copy of traffic; classifies and reports without sitting in the forwarding pathSits in the forwarding path and can act on traffic
Gives capacity and reporting teams classification straight awayNeeded for policy enforcement
Where most operators startUsually a second phase, after classification has been validated

Both run on your own infrastructure, and the platform can operate entirely in the Kingdom. The architecture is cloud-based, with full-mesh interconnection between the control and user planes and multi-level disaster recovery.

Governance controls before go-live

Traffic management changes what subscribers experience, so the rules need an owner and an audit trail. These controls sit with the operator, not the platform:

  1. 1. A policy catalogue.For each rule: the application class, the scope (application, subscriber group or cell), the action, the commercial or regulatory basis, and an owner.

  2. 2. Report-only first.Every new rule runs in report-only mode, so you see what it would have done before it acts.

  3. 3. An approval route.Who signs off the move from report-only to enforcement, and on what evidence.

  4. 4. Data rules.The platform produces metadata about applications, traffic volumes and triggered policies; retention and access follow your data-governance rules and the contract.

  5. 5. Change control.Model and policy updates enter the live network through your change-control process.

Test it on your network in 3 stages

  1. 1. Written acceptance criteria.Agree what is tested, on which traffic, how it is measured and what counts as a pass.

  2. 2. Out-of-band classification against verified traffic.Measure false positives as carefully as detection: a false positive means a rule acts on the wrong traffic.

  3. 3. Report-only policy, then the inline decision.Enforcement follows only after classification has proven itself.

See also: How Interkey runs a DPI proof of concept

Questions traffic management teams ask

Is DPI the right tool for every network?

No. On an enterprise or campus network, flow records and firewall logs may already answer the question more cheaply. DPI earns its place on operator networks, where you need to know what the traffic is, at scale, to manage capacity, policy and reporting.

Can a policy be tested without affecting traffic?

Yes. In report-only mode a rule shows what it would have done without acting on traffic.

Do we need an inline deployment to get traffic reports?

No. Out-of-band deployment gives classification and reporting. Only policy enforcement needs an inline deployment.

Does it work when most traffic is encrypted?

It classifies encrypted and tunnelled flows from observable characteristics without decrypting them. Results depend on your subscriber mix, application mix and encryption share, so measure them on your own traffic.

How does it connect to our policy and charging system and 5G core?

Integration is part of the deployment design. You bring your policy architecture and enforcement points to scoping, and the integration plan documents what each side provides.

How is it priced?

On a quote basis, depending on deployment, capacity, throughput, network placement and the capabilities in scope. There is no published price list.

Discuss a DPI proof of concept

Tell us which traffic management job comes first (capacity, congestion, plans and fair use, or regulatory reporting), whether your network is fixed, mobile or both, and your timeline. Our team in Riyadh replies in Arabic or English.

By sending this form, you agree to our Privacy Notice.