Capacity and congestion analysis for mobile operators
Mobile network congestion: diagnose it by application
Mobile network congestion happens when demand on a cell, a backhaul link or a core node exceeds its capacity at a given time, and subscribers see slower data and failed sessions. Your RAN, transport and core counters show where and when it happens; they do not show which applications are driving it. Application-level classification adds that answer, so capacity and policy decisions rest on evidence. Interkey Intelligent DPI classifies traffic by application in real time, including encrypted traffic, without decrypting it.
A Saudi product, built by Interkey
For RAN, capacity planning and network operations engineers at mobile operators. Reviewed November 2026.
- Counters say where and when; classification says what.Use both.
- Diagnosis does not touch the service path:an out-of-band deployment classifies and reports without sitting in the forwarding path.
- Three possible decisions:add capacity, apply an approved policy, or take no action because the load is expected.
- Governance stays with you:the output is metadata, policies act per application, subscriber group or cell, and the platform can run entirely in the Kingdom.
Where network counters stop and classification starts
| Question | Answered by |
|---|---|
| Which cells, links or nodes are congested, and when? | Your RAN, transport and core performance counters |
| Which applications fill the congested periods? | Application classification |
| Is one class growing, or is all traffic growing evenly? | Application-level trends over time |
| Did a policy change have the intended effect? | Policy execution feedback |
| What do we report to management or the regulator? | Both sources, combined in your reporting |
Patterns to tell apart
Congestion that looks the same in the counters can have different causes, and each calls for a different response:
- Steady busy-hour growth.One or more application classes grow week on week until demand passes capacity. This is a capacity-planning signal.
- Short peaks tied to places or events.Crowds at a venue, or one large download reaching many devices at once, create peaks that fade. Building permanent capacity for them may not be the right answer.
- One class dominating the busy hour.A single application class, such as video streaming, takes most of a cell's busy-hour capacity. This is where plan design and policy decisions come in.
- Hidden application mix.Tunnelled or evasive traffic disguises what it is, so simple reports attribute the load to the wrong class.
- A bottleneck away from the radio.A backhaul or core link reaches capacity while the cells look normal.
A diagnosis workflow
1. Locate.Start from your counters: the cells, links or nodes that are congested, and the hours when it happens.
2. Classify.Compare the application mix during the congested periods with the same points at normal load.
3. Trend.Check whether a class is growing steadily, peaking briefly, or appeared recently.
4. Decide.Choose between adding capacity, applying an approved policy, or no action.
5. Verify.After the change, check policy execution feedback and the application trends to see whether it worked.
Interkey Intelligent DPI supplies the application view for steps 2, 3 and 5. Its Live View gives a network-wide picture with drill-down to application and flow level, trend analysis at network and application level, and real-time policy execution feedback, with latency measured in seconds. See Live View on the features page.
Decisions the application view supports
| Decision | What the application view adds |
|---|---|
| Where to add capacity | Which application classes drive busy-hour growth at each point, so capacity plans follow real demand |
| Whether to apply a policy | Policies can act per application, subscriber group or cell, in support of your plans, fair-use terms and QoS rules. Which actions a policy takes is set by your policy and applicable regulation |
| How to design plans and tiers | Which application classes your plans and tiers need to account for |
| What to report | A reporting layer for capacity and regulatory teams, with traffic described by application |
A policy acts on traffic only in an inline deployment. Run it in report-only mode first, so you see what it would have done before it changes anything.
Out-of-band is enough to diagnose
For diagnosis, the platform observes traffic out-of-band, without sitting in the forwarding path, so nothing in the service path changes while you investigate. An inline deployment is needed only when a policy must act on traffic, and operators usually add it as a second phase after classification has been validated on their own traffic. See Inline vs out-of-band DPI: choosing the deployment.
The platform is designed for traffic intake at terabit-per-second scale across fixed and mobile networks. Throughput for your network is scoped per deployment and should be confirmed with a load test at your busiest-hour peak.
Encrypted traffic and data governance
A large share of mobile traffic is encrypted. Interkey Intelligent DPI classifies encrypted and tunnelled flows from characteristics that remain observable (handshake structure, packet sizing and timing, and endpoint behaviour) without decrypting them. See Network traffic classification engine for telecom operators.
The output is metadata about applications, traffic volumes and triggered policies. Retention and access are set in the deployment design and the contract, following your data-governance rules. The platform runs on your infrastructure and can operate entirely in the Kingdom.
Questions about mobile network congestion
What causes mobile network congestion?
Does DPI reduce congestion by itself?
Do we need an inline deployment to diagnose congestion?
Does it work on 5G networks?
What data does the platform keep?
Related pages
- Network traffic classification engine
- Inline vs out-of-band DPI
- Network traffic management for telecom operators
- Interkey Intelligent DPI features
- Deep packet inspection in Saudi Arabia
- Video optimization for telecom operators
- Quality of experience monitoring
- Bandwidth management for ISPs
- DPI for mobile operators
- DPI RFP requirements
Proof of concept in three phases
Written acceptance criteria.Agree which rules are tested, on which traffic and subscriber groups, how results are measured and what counts as a pass.
Out-of-band classification against verified traffic.Measure false positives as carefully as detection for every class the rules depend on.
Report-only policy, then the inline decision.Each rule shows what it would have done; enforcement follows only after classification has proven itself.
Discuss congestion analysis on your network
Tell us where congestion shows up (cells, regions, backhaul or core links), which counters your team uses today, and your timeline. Our team in Riyadh replies in Arabic or English.
- Our engineers read your request
- We agree the scope of a technical session
- You see the product on a use case you choose