Measuring Trunk Capacity: Where the Quoted Channel Count and the Reality Diverge
A carrier quotes twenty channels. A spreadsheet assumes a codec rate. Nobody accounts for packetisation or transport overhead, and the first busy Monday produces calls that sound broken while every document in the project claims the link was sized correctly. The gap is not incompetence. It is three different numbers being used to mean the same thing.
Four Numbers That Are Regularly Confused
A channel count is a commercial unit. A codec rate is a payload figure. A packet rate describes how often the codec places samples on the wire. A link capacity is a property of the transport. Mixing them is how a trunk sized for one thing ends up carrying another, and the arithmetic error usually lands in the same direction, because the codec rate is the number everyone has written down.
The other confusion is between sessions offered and media carried. A platform's channel counter may show twenty active sessions while the media path is already saturated, because a signalling session and a media stream are not the same thing and one can be much heavier than the other.
| Quantity | What it measures | Common mistake |
|---|---|---|
| Channel count | Commercial rights sold by the carrier | Read as a guaranteed concurrent call figure |
| Codec rate | Payload bits per second of audio | Used as the total on the wire |
| Effective rate per call | Measured bits including all overhead | Never measured at all |
| Concurrent sessions | Signalling state held by the platform | Assumed equal to media carried |
| Interface throughput | Actual traffic on the media port | Sampled over a quiet period only |
Overhead Is Not a Rounding Error
Take the payload rate and add what the transport costs: media transport headers, the secure variant of those headers where it is used, UDP, IP, and Ethernet framing, plus the effect of any tunnel or encryption, plus the cost of keeping the line alive with silence. On a short packetisation interval the fixed per-packet cost dominates, so the effective rate per call sits noticeably above the codec figure. A long interval reduces header cost and increases both latency and jitter sensitivity.
Packet aggregation improves the ratio but adds delay and makes loss more damaging, which is a poor combination for speech. If a trunk is short of capacity, aggregation makes the shortfall smaller rather than solving it.
What Measured Means
Measurement has to come from the interface that actually carries the media, over a window that includes a genuine peak, with the routing and the services as they are on an ordinary day. Interface counters on the media-facing port are the primary source. Platform statistics give concurrent session counts but not bitrate. A capture on a span port gives packet counts and identifies the codec in use at the time, which is the fastest way to confirm that a trunk is running something other than what was assumed.
Sustained load and burst load are separate measurements and should be reported separately. A trunk absorbs a burst of busy signals happily while failing on sustained occupancy, because buffering and scheduling smooth the burst but do nothing for the average.
- State the codec and packetisation interval alongside every capacity figure
- Read bitrate from interface counters rather than from the codec specification
- Separate peak, sustained, and burst figures in the capacity report
- Measure the upstream direction, which is the one that is usually constrained
- Confirm what the shaping policy does to voice before blaming the carrier
- Capture a span port to identify the codec actually in use on the trunk
- Recalculate the capacity whenever a site changes its voice services
The Upstream Direction Is Usually the Constrained One
In a typical branch, the outbound direction to the carrier carries media for calls the branch places, while the inbound direction carries media for calls it receives. A site with heavy outbound calling has a different constraint from one with heavy inbound, and a symmetric contract hides the difference. Check the upstream capacity of the internet circuit, any shaping or bandwidth policy applied to it, and whether voice is marked and policed differently from everything else.
Policy-based shaping is where the surprise usually lives. A router that treats voice as best effort will behave perfectly until somebody uploads a large file or a backup starts, at which point the fault appears at the far end as rough audio and at the near end as nothing at all.
Redstone Approach
Redstone platforms expose concurrent session counts, per-trunk call detail records, and platform statistics that can be exported for correlation with interface counters, so a measured figure can be assembled from two independent sources rather than one. Trunk profiles make the codec and packetisation assumptions explicit configuration rather than inherited defaults, which is what allows a capacity figure to be recomputed honestly when a site adds a channel.
Redstone Systems, Inc. founded in Delaware, USA in December 2002, has been the ODM vendor for many well-known communications companies, serving the Southeast Asian market. In 2020, Redstone Systems will begin to return to the North America market with its self-developed brand.
Redstone has a complete product line of intelligent voice gateways, providing IP-PBXs, analog VoIP gateways (FXS/FXO), digital VoIP gateways (E1/T1), border appliances, and session boundary controllers (SBCs).
With advanced technology in digital signal processor (DSP), speech coding and speech processing, as well as efficient operational tools such as cloud remote management, auto provisioning, Redstone gateways are widely used in markets of enterprise communications, cloud communications, call centers, operators’ IMS/SIP trunks, bringing users friendly, efficient and reliable communication experience.
For more information about Redstone, You can follow us on Facebook, Linkedin, and Youtube to be the first to get the latest news.

Current Location :

