Offered load¶

Measure the average service work presented to a queue or telecommunications resource before blocking, loss, or carrying effects, conventionally as arrival rate times mean holding time.

Version: v1 · 2026-08-30 · History
Domain-specific #: 2414
Origin domain: queueing theory
Subdomain: traffic intensity before admission
Aliases: Offered traffic, Traffic offered

Core Idea¶

Offered load is the mean amount of service work presented to a resource during a reference interval before blocking, dropping, abandonment, or finite-capacity admission removes any of it. Under a stationary arrival model with rate \(\lambda\) and mean holding or service time \(E[S]\), the dimensionless traffic intensity is \(A=\lambda E[S]\), expressed in erlangs in teletraffic contexts. It answers how much average simultaneous occupancy the arrivals would require if all were admitted, not how much the system actually carries.^[1]

Arrival events bring service requirements. Multiplying their mean rate by mean holding time converts events per time and time per event into expected concurrent work. In a loss system, some requests are blocked, so carried load is smaller; in a delay system, work can wait, making queue length and response depend additionally on service capacity, variability, discipline, and utilization. Offered load can be estimated from demand records or arrival and holding observations, but repeated attempts and measurement boundaries must be defined consistently.^[2]

The product \(\lambda E[S]\) is not a universal statement that every queue has a Poisson arrival process, nor is it itself a blocking probability. For one server it coincides numerically with utilization only when stable work is admitted and definitions align; for multiple servers it represents traffic offered across the group and can exceed one erlang without implying any single server exceeds full occupancy. Carried load, throughput, queue length, demand, and capacity are different quantities linked only under stated models.^[3]

Structural Signature¶

Arrival stream. Requests enter or attempt to enter at a defined mean rate.
Service requirement. Each request would occupy resource capacity for a holding or service duration.
Reference resource. A server, circuit group, channel, or queueing station defines where load is presented.
Observation interval. Stationarity and aggregation are assessed over a declared period.
Offered-work calculation. Arrival rate times mean duration yields dimensionless expected concurrent work.
Admission boundary. Blocking, loss, delay, and retry rules separate offered from carried traffic.
Capacity comparison. Servers or channels contextualize, but do not define, the offered quantity.
Estimation uncertainty. Sampling, burstiness, duration variation, and retries qualify the reported value.

What It Is Not¶

Not carried load. Carried load includes only work actually admitted or served.
Not throughput. Throughput counts completed work per time rather than concurrent service demand.
Not utilization in every system. Equality requires particular stable single-resource and admission conditions.
Not blocking probability. Blocking is an outcome predicted from offered load plus capacity and a traffic model.
Not queue length. Waiting work depends on service configuration, variability, and discipline.
Not economic demand. Teletraffic attempts and holding time define a performance quantity, not a price-quantity schedule.

Scope of Application¶

The abstraction is literal wherever practitioners can identify the same constitutive roles, apply the same boundary tests, and obtain the same kind of output. The following habitats are uses of Offered load itself, not metaphors based only on resemblance.

Circuit-group dimensioning. Relating attempted call traffic to capacity and target blocking under a declared model.
Contact and service systems. Estimating work presented before abandonment or admission effects.
Computer performance. Characterizing task work offered to a server or resource pool.
Capacity comparison. Separating rising demand from changes in service time or carried throughput.
Model calibration. Estimating arrival and holding components rather than fitting occupancy alone.
Traffic accounting. Distinguishing initial attempts, repeated attempts, blocked work, and completed service.

Clarity¶

A clear account of Offered load must preserve the recognition invariant stated in the Core Idea rather than rely on the title alone. Define one arrival, one service occupation, the resource group, and the interval. Report arrival rate and holding-time basis separately before their product. State how blocked attempts, retries, abandonment, and multi-resource service are counted. Do not infer blocking, delay, or utilization without the required capacity and stochastic assumptions. These declarations are not editorial extras: each changes what observations count, which transformations are licensed, and what conclusion can be drawn. A reader should be able to reconstruct the input, the operative rule, the output, and at least one defeater from the account without consulting an implementation or guessing an unstated convention.

Manages Complexity¶

Offered load manages complexity by replacing a diffuse field of observations or possible operations with a bounded role structure: arrival stream supplies requests enter or attempt to enter at a defined mean rate.; service requirement supplies each request would occupy resource capacity for a holding or service duration.; reference resource supplies a server, circuit group, channel, or queueing station defines where load is presented.; observation interval supplies stationarity and aggregation are assessed over a declared period.; offered-work calculation supplies arrival rate times mean duration yields dimensionless expected concurrent work.. The compression is useful because it localizes disagreement. One can ask whether the input was properly formed, whether a constitutive relation held, whether an alternative explanation defeats the inference, or whether the output was overinterpreted. The same compression can mislead when its discarded detail is exactly what the decision requires. A reference-grade use therefore reports both the invariant retained and the information intentionally lost.

Abstract Reasoning¶

Specify the queueing boundary and distinguish attempts from admitted work.
Estimate a mean arrival rate over a regime where the average is meaningful.
Estimate mean holding or service time on the matching population and basis.
Multiply the two quantities and verify that time units cancel.
Compare offered and carried quantities to locate blocking, loss, or abandonment.
Introduce capacity and variability only through an explicit queueing model.
Report uncertainty and sensitivity when burstiness or nonstationarity undermines the mean.
Test the candidate interpretation against the nearest named confusable rather than accepting a shared surface feature.
State the conclusion at the same scope as the source conditions, and retain uncertainty or nonuniqueness where the construct does not remove it.

Knowledge Transfer¶

The strict upward abstraction is Queueing. Offered load instantiates Queueing because it compresses an arrival stream and service-time requirement into the mean work presented at a queueing resource before admission outcomes. Within traffic intensity before admission, the full mechanism transfers literally when the same roles and boundary tests recur. Beyond that domain, only the parent-level skeleton should travel. Reusing the label Offered load after removing its constitutive vocabulary would hide a change of mechanism behind an analogy. The honest transfer rule is therefore two-stage: recognize the domain-specific pattern first, then lift only the parent relation that remains invariant under a substrate change.

Examples¶

Canonical¶

A circuit group receives an average of 120 call attempts per hour, each requiring an average of three minutes if admitted. The offered load is six erlangs because 120 per hour times 0.05 hour equals six. This does not say six calls are always active or that blocking is zero; those conclusions require capacity and traffic-distribution assumptions.

Mapped back: input and conventions → constitutive role test → bounded output → explicit interpretation and defeater check.

Applied / In Practice¶

A service reports four erlangs carried and claims demand is four erlangs. If blocked attempts and abandoned requests were omitted, the figure is not offered load. Reconstructing attempted arrivals and their intended holding-time basis may show six erlangs offered. The two-erlang gap is an admission outcome, not a unit conversion error.

Mapped back: field observation or problem → candidate recognition → confusable and limit checks → appropriately scoped conclusion.

Structural Tensions¶

T1: Mean work versus burstiness. Equal average load can produce different delay under different arrival variability. Diagnostic: What variability information is needed for the performance claim?
T2: Attempted versus repeated demand. Retries can be new load or consequences of previous blocking. Diagnostic: Does the counting rule identify reattempts?
T3: Offered versus carried. Admission makes observed occupancy smaller than presented work. Diagnostic: Where in the measurement chain are blocked jobs recorded?
T4: Dimensionless unit versus physical capacity. An erlang measures average concurrent work, not a specific circuit count. Diagnostic: How many service units and what sharing model are assumed?
T5: Stationary summary versus time variation. A daily mean can hide overload intervals. Diagnostic: Is the reference interval homogeneous enough for the intended decision?
T6: Autonomous quantity versus Queueing. Queueing supplies arrivals and service; offered load fixes their pre-admission work product. Diagnostic: Would the quantity remain meaningful without a queueing resource and service duration?

Structural–Framed Character¶

Offered load is formal and metrological: its dimensional relation is simple, while observation boundaries and stochastic interpretation require engineering judgment. The five framing criteria point in a consistent direction. Evaluative weight is limited to whether the defining conditions are met, not whether the outcome is desirable. Human practice matters to the extent that experts choose conventions, instruments, or reporting thresholds, but those choices do not make every verdict arbitrary. Institutional history explains the name and standard use; it does not replace the recognition rule. The operative vocabulary travels within the home field and closely adjacent subfields, while transfer farther away requires translation to the parent prime. Thus recognition remains disciplined even where interpretation is defeasible.

Structural Core vs. Domain Accent¶

What is skeletal. Offered load instantiates Queueing because it compresses an arrival stream and service-time requirement into the mean work presented at a queueing resource before admission outcomes. This is the part that can be expressed without the candidate's specialist nouns.

What is domain-bound. The domain accent is attempted traffic, arrival rate, holding time, erlang units, admission, blocking, carried traffic, circuits or servers, and teletraffic accounting. Remove those elements and the result is no longer Offered load; it is only the parent relation or a loose analogy.

Why this does not clear the prime bar. The name does not recur with unchanged diagnostics across three independent domains. What transfers is already represented by prime:queueing. The candidate remains autonomous because its in-domain recognition rule, failure modes, and consequences are stable, but its vocabulary and interventions do not float free of the home substrate.

Offered load instantiates Queueing because it compresses an arrival stream and service-time requirement into the mean work presented at a queueing resource before admission outcomes.

The prospective workspace queue contains one strict upward edge to prime:queueing. No live DAG mutation is authorized.

Relationships to Other Abstractions¶

Current abstraction Offered load Domain-specific

Parents (1) — more general patterns this builds on

Offered load is a kind of Queueing Prime

Offered load instantiates Queueing because it compresses an arrival stream and service-time requirement into the mean work presented at a queueing resource before admission outcomes.

Hierarchy paths (2) — routes to 2 parentless roots

Offered load → Queueing → Allocation → Scarcity → Constraint

Show alternative path (1)

Neighborhood in Abstraction Space¶

Offered load sits in a sparse region of the domain-specific corpus (92^nd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Carried load. The average service occupancy successfully accepted by the resource.
Traffic intensity. Sometimes a synonym, but in queueing formulas may denote offered load normalized by server count.
Utilization. The fraction of capacity busy, bounded per resource and model-dependent.
Throughput. The completion rate, with dimensions of jobs per time.
Queue length. The number waiting or in system, dependent on more than mean offered work.
Demand. A broad or economic request relation not fixed by mean service duration.

References¶

[1] International Telecommunication Union. (1993). ITU-T Recommendation E.600: Terms and Definitions of Traffic Engineering. https://www.itu.int/rec/T-REC-E.600-199303-I registry ↩

[2] Cooper, Robert B. (1981). Introduction to Queueing Theory, 2^nd ed. North-Holland. ISBN 978-0-444-00379-9. registry ↩

[3] Kleinrock, Leonard. (1975). Queueing Systems, Volume 1: Theory. Wiley. ISBN 978-0-471-49110-1. registry ↩