Enter radio access delay
Use the 5G radio-segment delay on the same measurement basis as the other components.Add transport and core delay
Enter backhaul or fronthaul transport plus the processing or forwarding delay attributed to the 5G core.Add external and application delay
Include the path beyond the core and the application-processing contribution.Set the end-to-end target
Enter the maximum latency budget for the service or architecture being evaluated.Inspect headroom and concentration
Use the remaining milliseconds and largest component to see where optimization effort may have the greatest effect.
5G Small Cell Latency Budget Planner
The 5G Small Cell Latency Budget Planner adds radio, transport, core-network, external network, and application-processing delay to produce an end-to-end latency estimate. A target latency input then shows how much headroom remains and what percentage of the latency budget has been consumed, making the model useful for interactive and time-sensitive 5G service planning.
This planner is intended for architecture comparisons rather than protocol simulation. 5G latency can vary with scheduling, retransmissions, congestion, transport topology, core placement, application hosting, and whether measurements are one-way or round-trip. Keep every component on the same measurement basis and use values from the actual deployment path where possible. Jitter and packet loss should be assessed separately because they can affect user experience even when the average latency total meets its target.
5G small cell latency inputs
ms
ms
ms
ms
ms
ms
Result
—
estimated total latency
Headroom vs. target—
Target budget used—
Largest component—
Total latency = Radio access + Transport/backhaul + 5G core + External network + Application processingHeadroom = Target latency − Total latencyBudget used (%) = Total latency ÷ Target latency × 100
Where:
- Latency components — additive delay values in milliseconds
- Target latency — maximum desired end-to-end delay on the same one-way or round-trip basis
- Headroom — difference between the target and the summed latency
Assumptions: Each component is entered once and uses the same directionality. The model does not calculate percentile latency, jitter, queueing distributions, or packet-loss effects.
What the result means
The main result is the total of the entered delay components; the breakdown shows whether the target is met and which component is largest.
For strict service objectives, evaluate high-percentile latency and jitter in addition to an average or nominal budget.
Given:
- Radio access: 7 ms
- Transport/backhaul: 9 ms
- 5G core: 4 ms
- External network: 14 ms
- Application processing: 11 ms
- Target: 55 ms
Calculation:
Total = 7 + 9 + 4 + 14 + 11 = 45 ms
Headroom = 55 − 45 = 10 ms
Budget used = 45 ÷ 55 × 100 ≈ 81.8%
Result: 45 ms total latency
Interpretation: The modeled path is 10 ms below the target, with the external network contributing the largest single entered delay.
Why separate transport delay from core delay?
They often change independently. A closer edge core can reduce core or external-path delay, while transport topology and distance can still add their own latency.
What does it mean if budget used exceeds 100%?
The sum of the entered components is higher than the target. The headroom field will show the amount by which the target is exceeded.
Should radio retransmission delay be included?
If your radio-access value is measured and already reflects typical retransmissions, do not add it again. If you are building a component model, include retransmission allowance only once and document the assumption.
Can this planner evaluate URLLC performance?
It can organize a nominal latency budget, but URLLC design also depends on reliability, tail latency, scheduling, radio configuration, and end-to-end architecture. Those requirements need a more detailed engineering analysis.
How can I use the largest-component result?
Treat it as a prioritization clue. Confirm that the measurement boundaries are correct, then investigate whether architecture, routing, placement, or processing changes can reduce that dominant delay source.