GOS Calculator
GOS vs. Number of Servers
What is GOS Calculation?
GOS calculation, or Grade of Service calculation, is a fundamental concept in telecommunications engineering and queuing theory. It quantifies the probability that a new call or service request will be blocked or delayed due to the unavailability of resources. In simpler terms, it's a measure of how well a system can handle the traffic placed upon it. A lower GOS value indicates better service quality, meaning fewer calls are rejected or encounter busy signals.
This metric is crucial for network planners, call center managers, and anyone involved in designing systems where a finite number of resources must serve a variable demand. It helps in striking a balance between providing adequate service quality and minimizing infrastructure costs.
Who Should Use a GOS Calculator?
- Telecommunications Engineers: For network dimensioning, trunk group sizing, and capacity planning.
- Call Center Managers: To determine the optimal number of agents or lines needed to meet service level targets.
- IT Professionals: For sizing servers, network bandwidth, or other shared resources.
- Students and Researchers: Studying queuing theory, probability, or operations research.
Common Misunderstandings in GOS Calculation
One common misunderstanding is confusing GOS with "Quality of Service" (QoS). While related, QoS is a broader term encompassing various aspects like latency, jitter, and packet loss, whereas GOS specifically focuses on the probability of blocking or delay. Another frequent error involves unit confusion, particularly with traffic intensity in Erlangs. Erlangs represent the total traffic volume, not the number of calls, and must be correctly calculated based on call arrival rates and holding times.
GOS Calculation Formula and Explanation
The most widely used formula for GOS calculation in systems where blocked calls are cleared (lost) is the Erlang B formula. This formula assumes an infinite calling population, random call arrivals (Poisson process), and exponential call holding times. It's ideal for sizing telephone trunk groups or similar systems where calls are simply dropped if no resources are available.
The Erlang B formula is given by:
Pb = (AN / N!) / ∑k=0N (Ak / k!)
Where:
- Pb = Blocking Probability (Grade of Service)
- A = Traffic Intensity in Erlangs
- N = Number of Servers/Trunks
- N! = Factorial of N (N * (N-1) * ... * 1)
The numerator represents the probability of all N servers being busy, while the denominator is the sum of probabilities for k servers being busy, from 0 to N.
Variables Table for GOS Calculation
| Variable | Meaning | Unit | Typical Range |
|---|---|---|---|
| A | Traffic Intensity | Erlangs | 0.1 to 100+ Erlangs |
| N | Number of Servers/Trunks | Unitless (count) | 1 to 1000+ |
| Pb (GOS) | Blocking Probability (Grade of Service) | % or Decimal (0-1) | 0.001% to 100% |
Practical Examples of GOS Calculation
Example 1: Sizing a Small Trunk Group
A small business expects a total of 3 Erlangs of traffic during its busy hour. They want to achieve a Grade of Service of 1% (0.01 blocking probability). How many telephone lines (trunks) do they need?
- Inputs: Traffic Intensity (A) = 3 Erlangs
- Target GOS: 1%
- Calculation: Using the Erlang B formula, we would iterate through different values of N until Pb is approximately 0.01.
- Result: With 3 Erlangs, to achieve approximately 1% GOS, you would need 7 servers/trunks. (A GOS of ~0.0102 or 1.02% for N=7, and ~0.0048 or 0.48% for N=8).
- Interpretation: If the business installs 7 lines, about 1 out of every 100 calls during the busy hour will get a busy signal.
Example 2: Call Center Agent Staffing
A call center receives 15 Erlangs of traffic during its peak hour. They currently have 20 agents available. What is the Grade of Service they are providing, and what is the blocking probability?
- Inputs: Traffic Intensity (A) = 15 Erlangs, Number of Servers (N) = 20
- Calculation: Plugging these values into the Erlang B formula.
- Result: The GOS is approximately 1.84% (or a blocking probability of 0.0184).
- Interpretation: This means that during the peak hour, roughly 1.84% of incoming calls will be blocked because all 20 agents are busy. If this is too high, the call center may need to increase the number of agents or consider an Erlang C calculation if calls are queued instead of blocked.
How to Use This GOS Calculation Calculator
Our GOS calculator is designed for ease of use, providing instant results for your Grade of Service calculations.
- Input Traffic Intensity (A): Enter the total traffic volume in Erlangs. If you don't know this, calculate it by multiplying your average call arrival rate (calls per hour) by your average call holding time (hours) or (calls per second) by (seconds). For example, 30 calls/hour with 3-minute average holding time = 30 * (3/60) = 1.5 Erlangs.
- Input Number of Servers/Trunks (N): Enter the total number of available resources, such as telephone lines, network trunks, or customer service agents. Ensure this is a whole number.
- Click "Calculate GOS": The calculator will instantly display the Grade of Service as a percentage, along with intermediate values.
- Interpret Results: The primary result is the GOS (blocking probability) as a percentage. A GOS of 1% means 1 out of 100 calls will be blocked.
- Use the Chart and Table: The dynamic chart visually represents how GOS changes with the number of servers, while the table provides specific values, helping you quickly identify optimal resource levels.
- Copy Results: Use the "Copy Results" button to quickly grab the calculated values for your reports or documentation.
Key Factors That Affect GOS Calculation
Several critical factors influence the Grade of Service in any system. Understanding these helps in better capacity planning and service delivery.
- Traffic Intensity (A): This is the most direct factor. Higher traffic intensity (more calls, longer calls) for a given number of servers will inevitably lead to a higher GOS (worse service). Traffic is measured in Erlangs, a unit representing the average number of simultaneous busy channels.
- Number of Servers/Trunks (N): Increasing the number of available resources directly improves GOS (reduces blocking probability). However, there are diminishing returns, meaning each additional server provides less improvement as N gets larger.
- Call Arrival Pattern: The Erlang B formula assumes a Poisson arrival process (random arrivals). If call arrivals are highly clustered or predictable, the actual GOS might deviate from the calculated value.
- Call Holding Time Distribution: Erlang B assumes exponential call holding times. Deviations from this (e.g., very short or very long calls consistently) can affect the accuracy of the GOS calculation.
- Lost Calls Cleared (LCC) vs. Lost Calls Delayed (LCD): The Erlang B formula specifically applies to LCC systems, where blocked calls are immediately dropped. If calls are queued (LCD systems, like many call centers), the Erlang C formula would be more appropriate.
- System Overload: During periods of extreme overload where traffic intensity significantly exceeds capacity, the GOS can rapidly deteriorate, leading to a high percentage of blocked calls.
- Target Service Level: Organizations often set a target GOS (e.g., 1% blocking). This target then drives the required network dimensioning or staffing levels.
Frequently Asked Questions about GOS Calculation
Q: What is a good GOS (Grade of Service) value?
A: A "good" GOS depends on the application. For traditional telephony, a GOS of 0.01 (1% blocking) is often considered acceptable for busy hour traffic. For critical services, it might be 0.001 (0.1%), while for less critical applications, a higher GOS (e.g., 5%) might be tolerated. It's a balance between cost and service quality.
Q: How do Erlangs relate to GOS calculation?
A: Erlangs are the unit of traffic intensity, representing the total amount of traffic offered to a system. The GOS calculation (Erlang B) uses this traffic intensity, along with the number of available servers, to determine the probability of a call being blocked. Without traffic intensity in Erlangs, GOS cannot be calculated using Erlang B.
Q: Can I use this calculator for call centers?
A: This calculator uses the Erlang B formula, which assumes blocked calls are lost (cleared). If your call center queues calls instead of blocking them, you should use an Erlang C calculator, which accounts for queuing and calculates agent occupancy and average wait times.
Q: What happens if my inputs are outside the typical range?
A: The calculator will still perform the GOS calculation. However, extremely high traffic intensities or numbers of servers might lead to very small or very large intermediate numbers, potentially affecting floating-point precision in JavaScript. For practical purposes, ensure your inputs reflect realistic scenarios.
Q: Why is the GOS calculation important for voice network optimization?
A: GOS calculation is crucial for optimizing voice networks because it allows engineers to provision the correct number of trunks or channels. Over-provisioning wastes resources, while under-provisioning leads to high blocking rates, frustrated users, and lost business. Optimizing GOS ensures efficient resource utilization while maintaining acceptable service quality.
Q: Does GOS account for network latency or jitter?
A: No, GOS (specifically Erlang B) only accounts for the probability of a call being blocked due to resource unavailability. Metrics like latency, jitter, and packet loss are part of a broader Quality of Service (QoS) framework and require different measurement and calculation methods.
Q: How does the "Number of Servers" relate to "Trunks"?
A: In the context of GOS calculation, "servers" and "trunks" are often used interchangeably. A "server" refers to any resource that can handle a call or service request. In telecommunications, these are typically "trunks" or "channels" – the individual lines or paths available for communication.
Q: Why is GOS expressed as a percentage?
A: While the Erlang B formula outputs a decimal probability (e.g., 0.01), it is often converted to a percentage (1%) for easier understanding and communication. It represents the percentage of calls that are expected to be blocked.
Related Tools and Internal Resources
Explore more tools and articles to deepen your understanding of telecommunications engineering and capacity planning:
- Erlang B Calculator: Specifically designed for systems where blocked calls are lost.
- Erlang C Calculator: Essential for call centers and systems where blocked calls are queued.
- Traffic Intensity Explained: Learn how to accurately calculate traffic in Erlangs.
- Telecom Network Design Guide: A comprehensive guide to designing robust communication networks.
- Understanding QoS: Dive deeper into Quality of Service metrics beyond just GOS.
- Voice Network Optimization: Strategies and tools for improving voice communication efficiency.