https://urgentcomm.com/wp-content/themes/ucm_child/assets/images/logo/footer-new-logo.png
  • Home
  • News
  • Multimedia
    • Back
    • Multimedia
    • Video
    • Podcasts
    • Galleries
    • IWCE’s Video Showcase
    • IWCE 2022 Winter Showcase
    • IWCE 2023 Pre-event Guide
  • Commentary
    • Back
    • Commentary
    • Urgent Matters
    • View From The Top
    • All Things IWCE
    • Legal Matters
  • Resources
    • Back
    • Resources
    • Webinars
    • White Papers
    • Reprints & Reuse
  • IWCE
    • Back
    • IWCE
    • Conference
    • Special Events
    • Exhibitor Listings
    • Premier Partners
    • Floor Plan
    • Exhibiting Information
    • Register for IWCE
  • About Us
    • Back
    • About Us
    • Contact Us
    • Advertise
    • Terms of Service
    • Privacy Statement
    • Cookie Policy
  • Related Sites
    • Back
    • American City & County
    • IWCE
    • Light Reading
    • IOT World Today
    • Mission Critical Technologies
    • TU-Auto
  • In the field
    • Back
    • In the field
    • Broadband Push-to-X
    • Internet of Things
    • Project 25
    • Public-Safety Broadband/FirstNet
    • Virtual/Augmented Reality
    • Land Mobile Radio
    • Long Term Evolution (LTE)
    • Applications
    • Drones/Robots
    • IoT/Smart X
    • Software
    • Subscriber Devices
    • Video
  • Call Center/Command
    • Back
    • Call Center/Command
    • Artificial Intelligence
    • NG911
    • Alerting Systems
    • Analytics
    • Dispatch/Call-taking
    • Incident Command/Situational Awareness
    • Tracking, Monitoring & Control
  • Network Tech
    • Back
    • Network Tech
    • Interoperability
    • LMR 100
    • LMR 200
    • Backhaul
    • Deployables
    • Power
    • Tower & Site
    • Wireless Networks
    • Coverage/Interference
    • Security
    • System Design
    • System Installation
    • System Operation
    • Test & Measurement
  • Operations
    • Back
    • Operations
    • Critical Infrastructure
    • Enterprise
    • Federal Government/Military
    • Public Safety
    • State & Local Government
    • Training
  • Regulations
    • Back
    • Regulations
    • Narrowbanding
    • T-Band
    • Rebanding
    • TV White Spaces
    • None
    • Funding
    • Policy
    • Regional Coordination
    • Standards
  • Organizations
    • Back
    • Organizations
    • AASHTO
    • APCO
    • DHS
    • DMR Association
    • ETA
    • EWA
    • FCC
    • IWCE
    • NASEMSO
    • NATE
    • NXDN Forum
    • NENA
    • NIST/PSCR
    • NPSTC
    • NTIA/FirstNet
    • P25 TIG
    • TETRA + CCA
    • UTC
Urgent Communications
  • NEWSLETTER
  • Home
  • News
  • Multimedia
    • Back
    • Video
    • Podcasts
    • Omdia Crit Comms Circle Podcast
    • Galleries
    • IWCE’s Video Showcase
    • IWCE 2023 Pre-event Guide
    • IWCE 2022 Winter Showcase
  • Commentary
    • Back
    • All Things IWCE
    • Urgent Matters
    • View From The Top
    • Legal Matters
  • Resources
    • Back
    • Webinars
    • White Papers
    • Reprints & Reuse
    • UC eZines
    • Sponsored content
  • IWCE
    • Back
    • Conference
    • Why Attend
    • Exhibitor Listing
    • Floor Plan
    • Exhibiting Information
    • Join the Event Mailing List
  • About Us
    • Back
    • About Us
    • Contact Us
    • Advertise
    • Cookie Policy
    • Terms of Service
    • Privacy Statement
  • Related Sites
    • Back
    • American City & County
    • IWCE
    • Light Reading
    • IOT World Today
    • TU-Auto
  • newsletter
  • In the field
    • Back
    • Internet of Things
    • Broadband Push-to-X
    • Project 25
    • Public-Safety Broadband/FirstNet
    • Virtual/Augmented Reality
    • Land Mobile Radio
    • Long Term Evolution (LTE)
    • Applications
    • Drones/Robots
    • IoT/Smart X
    • Software
    • Subscriber Devices
    • Video
  • Call Center/Command
    • Back
    • Artificial Intelligence
    • NG911
    • Alerting Systems
    • Analytics
    • Dispatch/Call-taking
    • Incident Command/Situational Awareness
    • Tracking, Monitoring & Control
  • Network Tech
    • Back
    • Cybersecurity
    • Interoperability
    • LMR 100
    • LMR 200
    • Backhaul
    • Deployables
    • Power
    • Tower & Site
    • Wireless Networks
    • Coverage/Interference
    • Security
    • System Design
    • System Installation
    • System Operation
    • Test & Measurement
  • Operations
    • Back
    • Critical Infrastructure
    • Enterprise
    • Federal Government/Military
    • Public Safety
    • State & Local Government
    • Training
  • Regulations
    • Back
    • Narrowbanding
    • T-Band
    • Rebanding
    • TV White Spaces
    • None
    • Funding
    • Policy
    • Regional Coordination
    • Standards
  • Organizations
    • Back
    • AASHTO
    • APCO
    • DHS
    • DMR Association
    • ETA
    • EWA
    • FCC
    • IWCE
    • NASEMSO
    • NATE
    • NXDN Forum
    • NENA
    • NIST/PSCR
    • NPSTC
    • NTIA/FirstNet
    • P25 TIG
    • TETRA + CCA
    • UTC
acc.com

content


Ready, set  calculate!

Ready, set calculate!

If you're involved in the design or support of land mobile radio systems, you've probably heard of the three Cs: coverage, capacity and cost. Designers
  • Written by Urgent Communications Administrator
  • 1st April 2001

If you’re involved in the design or support of land mobile radio systems, you’ve probably heard of the “three Cs”: coverage, capacity and cost.

Designers typically try to meet coverage requirements by strategically locating sites, optimally specifying and configuring antenna systems, and controlling transmit power and other technical parameters.

System capacity, or the ability of the system to carry user messages, involves quite a different approach.

Of course, improving system coverage or capacity (or both) almost always increases system cost.

Whether you’re evaluating a potential new trunked radio system, or operating an existing system, it’s important to understand the basics of trunked system capacity.

In effectively forecasting requirements, your goal should be to provide your sites with an adequate number of channels to support peak user load. This will avoid user complaints about “not getting through” or “excessive queuing” and also allow for future expansion of your wireless network. An overall understanding of traffic planning and capacity modeling can also help identify factors that contribute to capacity problems on an existing system.

Basic concepts

Land-mobile radio systems rely on the concept of trunking to accommodate a large number of users with a small number of radio channels. It’s similar to how a small number of telephone trunk lines connecting telephone central office switches can be shared by a large number of individual users.

System control logic differentiates various trunked system architectures. Two types of trunking include: dedicated control channel (centralized trunking) and subaudible signaling control (scan-based or decentralized trunking). Com-Net Ericsson’s EDACS and Motorola’s Smartnet systems use dedicated control channels. Systems using E.F. Johnson LTR technology use subaudible signaling.

In the dedicated control-channel architecture, when the user presses the push-to-talk button, his radio requests a traffic channel (or working channel) by sending a message on a control channel for the serving site. If all available traffic channels for this site are in use, the user is usually placed in a queue until a channel becomes available. In analyzing trunked system capacity, this queuing process is usually the limiting factor.

Trunking theory definitions

Trunking theory fundamentals are applied to capacity analysis using statistical tools created by A. K. Erlang, a 19th-century Danish mathematician. Today, the measure of traffic intensity bears his name.

One Erlang is defined as the amount of traffic intensity carried by a channel that is fully occupied. For example, a channel that is completely occupied for 15 minutes during an hour timeframe carries 0.25 Erlangs of traffic.

The grade of service is a common system benchmark that provides a measure of a system’s ability to provide a user access to a trunked system during the busiest hour. The term busy hour refers to an arbitrary period but is important because it provides a worst-case result. The actual “busy hour” depends on your particular use patterns over time. For example, an urban SMR operator may specify 5 p.m. to 6 p.m. on Fridays, while the busy hour for a natural gas company using a trunked LMR system may coincide with a 3:30 p.m.-4:30 p.m. shift change on the first Monday of the month.

Table 1. Performance parameters.
Symbol Units Description
GOS %, s Grade of service
t s Desired queue delay
H s Average PTT duration
f PTTs/s Average PTTs per second
U Number of users
C Number of traffic channels

Specifically, for trunked architectures that support queuing, the GOS represents the probability that any given call experiences a delay greater than a specified queue time. A typical system GOS requirement might be stated as follows: “We desire that no more than 2% of our user’s call requests, or PTTs, be delayed in a queue for longer than three seconds.”

Applying the theory

To create capacity estimates that ultimately determine the number of required channels to support a specified GOS, several performance characteristics need to be provided, as shown in Table 1 at the left.

When designing a new system, these parameters come from:

  • estimates provided by mobile users or dispatchers.
  • engineering studies or experiments to gather over-the-air statistics from an existing system.
  • published industry studies.

The first step in applying trunking theory is to calculate the offered load per user — that is, how much traffic, in Erlangs, a given talk group or individual user service generates. Using the parameters defined in Table 1, the offered load per user, Au, is expressed as:

Au = f × H

As an example, assume a user’s average PTT duration, H, is five seconds, and each user, on average, PTTs 10 times during the busy hour. Note the extra term below that converts PTTs/hr to PTTs/s:

In this example, the user offers 0.0139 Erlangs of traffic to the system. Remember that one Erlang represents a fully occupied channel during busy hour, so this user, if he had his way, would require 1.39% of one channel’s capacity during a one-hour timeframe.

For a system that contains U users, the total offered load, A, is the number of users multiplied by the offered load per user:

A = U × Au

Assume the example system is a single-site trunked system with 100 users during the busy hour, so the total offered load, A, is simply 1.39 Erlangs. Based on the definition of an Erlang, one could easily (and incorrectly) assume that at least two traffic channels would support the offered load.

But recall that one Erlang is a completely filled channel; it’s unrealistic to have competing users completely occupy a channel. Doing so would require every user to PTT every instant a channel becomes available.

To determine a more realistic channel requirement to support the offered load, an Erlang formula or table is typically used for trunking systems.

The Erlang-C model

In the Erlang-C trunking model, or blocked calls delayed, a queue is provided to hold calls that are blocked. When a channel resource is not immediately available, the PTT request may be delayed until one becomes available. The GOS for this model represents the likelihood that a call would be delayed after waiting a specific time in the queue.

To determine the GOS, two calculations are needed. The first is the probability that a call will be delayed, which is derived from the following Erlang-C formula:

The second formula provides the probability that, once the call is in a queue, the delay will be t seconds:

The GOS is merely the product of the Erlang-C probability and the conditional probability, QDelay, of exceeding a delay of t seconds:

GOS = ErC × QDelay

Computer programs can readily compute these formulas. To help make capacity analysis easy, I have implemented these calculations into a Microsoft Excel spreadsheet that you can request by email from MRT (email: [email protected]). It makes use of Excel’s built-in Poisson function, the underlying basis for the Erlang models.

It should be noted that an Erlang-B model also exists to calculate the probability that a blocked call would be dropped. This formula should be used when trunking systems do not use a queue.

Bringing it all together

In the preceding example, a set of 100 users presented a total offered load of 1.39 Erlangs. Assuming the GOS requirement restricts a maximum of 2% of calls to be delayed in a queue for longer than three seconds, the simple calculator portion of the spreadsheet can be used to specify the input parameters (H = 5s and t = 3s) as well as U and f:

Users Avg dur.
(s)
PTTs/user Offered Load
(A Erlangs)
100 5 10 1.389

The spreadsheet calculates the offered load and the corresponding GOS for a range of systems having one to 10 channels. The results, shown below, show that, to meet a 2% GOS requirement, four traffic channels are required:

GOS Channels
175.4% 1
39.4% 2
7.6% 3
1.2% 4
0.2% 5
0.0% 6
0.0% 7
0.0% 8
0.0% 9
0.0% 10

Assumptions

The preceding example, as capacity planning does in general, makes several assumptions. The Erlang models were designed with the assumption that call arrivals will follow what’s known as a Poisson distribution. A random process characterizes this type of distribution and requires that the times between successive events (or PTTs) be exponentially distributed.

Another key requirement in modeling traffic behavior is to determine an average call transmission length. Statistics gathered from an FCC study (under contract RC10056) support five seconds as a reasonable average transmission length. Your actual average may vary depending on user needs and the actual mix of voice, data and interconnect calls.

One point to remember about traffic analysis, as in many disciplines, is that the results you obtain are estimates of the assumptions that you’ve provided. The “answer” can only be as good as your assumptions. The more detailed your investigations are to acquire the input parameters, the more accurate your results will be.

Modeling a wide-area system

Estimating capacity for a wide-area trunked system requires special consideration. Several complex, and debatable, factors contribute to user load.

Your initial task should be to determine the quantity of users (whether mobile, portable or fixed) and their anticipated operational service area. For many systems, this can become subjective.

For each system talk group, estimate the number of users, and determine a reasonable estimate for the average call duration, H, and the anticipated number of PTTs per user, f, during the busy hour. The wide-area spreadsheet calculator includes an example, shown in Table 2 on page 48, that lists six such talk group sets for electric and gas divisions in three counties (Oakland, Macomb, and Wayne). The spreadsheet calculates the offered load, A, for each talk group set.

The example analysis is for a fictitious five-site trunked system and includes the Pontiac, Northville, Macomb, Armada, and Howell sites. Under each site in the spreadsheet is a cell to distribute the offered load, as a percentage, from a given source to each individual site.

Talk groups deserve special consideration when analyzing capacity. This is because, unlike in cellular, users of trunked system talk groups can load a site with traffic even if they’re just “logged-in” and listening.

The spreadsheet distribution percentages, then, reflect the probability that at least one user will log into that particular site and bring with it the corresponding traffic load of all talk group users for the entire busy hour. Be certain to take overlapping coverage into consideration; neighboring sites will almost always receive a portion of each other’s traffic.

Other types of calls that typically load a trunked system include data, individual and private calls, and telephone interconnect calls. The example in Table 2 includes private call users, interconnect users and data users, who are assumed to be equally distributed throughout the system.

Table 2. Estimating wide-area system GOS
Talkgroup or
traffic source
Users Avg.
duration
(s)
PTTs/user Erlangs
(A)
Distribution of traffic among sites (%) Offered load (Erlangs)
Pontiac Northville Macomb Armada Howell Pontiac Northville Macomb Armada Howell
Oakland Gas 40 6 12 0.800 100 90 50 45 70 0.800 0.720 0.400 0.360 0.560
Macomb Gas 20 6 12 0.400 50 50 100 50 40 0.200 0.200 0.400 0.200 0.160
Wayne Gas 25 6 12 0.500 70 100 50 30 80 0.350 0.500 0.250 0.150 0.400
Oakland Elect. 20 5 8 0.222 100 90 50 45 70 0.222 0.200 0.111 0.100 0.156
Macomb Elect. 10 5 8 0.111 50 50 100 50 40 0.056 0.056 0.111 0.056 0.044
Wayne Electric 15 5 8 0.167 70 100 50 30 80 0.117 0.167 0.083 0.050 0.133
Private Calls 10 6 8 0.133 20 20 20 20 20 0.027 0.027 0.027 0.027 0.027
Interconnect 5 25 6 0.208 20 20 20 20 20 0.042 0.042 0.042 0.042 0.042
Data Calls 100 3 20 1.667 20 20 20 20 20 0.333 0.333 0.333 0.333 0.333
Total (Erlangs): 2.146 2.244 1.757 1.317 1.855

Finally, the spreadsheet sums the offered load for each site to determine the site’s total offered load. Table 3, as shown on page 50, then presents a GOS matrix, calculated from the ErC and QDelay equations, for the maximum acceptable queue delay, t, of three seconds.

From the table, we can conclude that to meet or exceed a 2% GOS, there should be six traffic channels at Pontiac and Northville, five at Macomb and Howell, and four at the Armada site.

Other considerations

Exploring the specifics of your trunked technology (EDACS, Smartnet, MPT-1327 or LTR) may help you to understand how the spreadsheet calculator can be customized for the services and traffic sources of your particular system.

Some areas that require further work to provide a more thorough analysis may include:

  • factoring-in call setup time (typically between 250ms and 400ms).
  • considering hang time (if used).
  • providing data calls to be specified in terms of individual application messages or bytes.
  • addressing application-specific packet and network-layer data overhead.

This analysis does not account for two main aspects of capacity. One is that, in dedicated control channel systems, the carrier-sense multiple-access scheme for the control channel itself can limit capacity. The second factor involves the blocking probabilities of central switching and network equipment. Usually, however, unless you’re dealing with a large system, user load itself vs. the available channels is the major concern.

Table 3. GOS Matrix
Channels Grade of service
Site
(Total Erlangs)
Pontiac
(2.146)
Northville
(2.244)
Macomb
(1.757)
Armada
(1.317)
Howell
(1.855)
1 426.9% 473.3% 276.8% 159.3% 309.8%
2 121.3% 137.3% 71.0% 34.7% 81.8%
3 30.9% 35.9% 16.0% 6.4% 19.0%
4 7.0% 8.3% 3.1% 1.0% 3.9%
5 1.4% 1.7% 0.5% 0.1% 0.7%
6 0.2% 0.3% 0.1% 0.0% 0.1%
7 0.0% 0.1% 0.0% 0.0% 0.0%
8 0.0% 0.0% 0.0% 0.0% 0.0%
9 0.0% 0.0% 0.0% 0.0% 0.0%
10 0.0% 0.0% 0.0% 0.0% 0.0%

Gutowski is senior RF engineer at Consumers Energy Company in Jackson, MI. He can be reached by email at: [email protected]

Tags: content

Most Recent


  • Ready, set  calculate!
    Newscan: Securing the Internet of Things is quite a challenge
    Also: EWA requests dismissal of 900 MHz applications; TIA names tech and policy priorities for 2014; IJIS Institute names Shumate Award winner; App makes bus waits more tolerable; a Blackberry comeback may be in the offing.
  • Ready, set  calculate!
    Newscan: FCC certifies Carlson Wireless's white-space radio
    Also: Congress looks to revamp telecom law; Obama to place some restraints on surveillance; IEEE to study spectrum-occupancy sensing for white-spaces broadband; Major Swedish transport operator opts for Sepura TETRA radios; RFMD to partner on $70 million next-generation power grid project; NENA opens registratiuon for "911 Goes to Washington."
  • Ready, set  calculate!
    Newscan: A look at the critical job of 911 dispatchers
    Also: NYC launches website for tracking 911 response times; Oregon implements 911 on pre-paid cell phones; LightSquared wants to keep spectrum assets; Harris receives multiple government orders; FCC extends rebanding financial reconciliation deadline; Zetron gear at core of communications system upgrade; Ritron debuts wireless access control system; EWA seeks policy review of VHF vehicular repeater system deployments.
  • Ready, set  calculate!
    Newscan: Average peak data rates of 144 MB/s average realized in tests with CAT 4 LTE device
    Also: Verizon, T-Mobile to swap unused spectrum to improve coverage; Internet giants oppose surveillance--but only when the government does it; FCC Chairman says incentive auction will be delayed until middle of 2015; FCC chair announces staff appointments; Alcatel-Lucent names Tim Krause as chief marketing officer; New Jersey county deploys TriTech CAD system; Toronto airport deploys 26-position Zetron console system;

Leave a comment Cancel reply

To leave a comment login with your Urgent Comms account:

Log in with your Urgent Comms account

Or alternatively provide your name, email address below:

Your email address will not be published. Required fields are marked *

Related Content

  • RugGear: Contributing to the future of mission-critical broadband communication review and market vision
  • Photo gallery: 2014 Communications Marketing Conference (CMC) in Tucson
  • Ready, set  calculate!
    Top 5 Stories - Week of Sept. 22
  • Ready, set  calculate!
    RCA plans to expand this year's Technical Symposium

Commentary


Updated: How ‘sidelink’ peer-to-peer communications can enhance public-safety operations

  • 1
27th February 2023

NG911 needed to secure our communities and nation

24th February 2023

How 5G is making cities safer, smarter, and more efficient

26th January 2023
view all

Events


UC Ezines


IWCE 2019 Wrap Up

13th May 2019
view all

Twitter


UrgentComm

State and local leaders can alleviate the burden on public-safety personnel by tackling three workforce trends dlvr.it/SlBH89

20th March 2023
UrgentComm

6G is shaping up to disappoint, and the industry can blame itself dlvr.it/Sl918J

20th March 2023
UrgentComm

Change is coming to the network detection and response (NDR) market dlvr.it/Sl4cts

18th March 2023
UrgentComm

Cloud services: A cloudy forecast for state and local governments dlvr.it/Sl0m8J

16th March 2023
UrgentComm

Wassell describes FirstNet Authority CEO post as ‘dream job’ dlvr.it/SkygsL

16th March 2023

Newsletter

Sign up for UrgentComm’s newsletters to receive regular news and information updates about Communications and Technology.

Expert Commentary

Learn from experts about the latest technology in automation, machine-learning, big data and cybersecurity.

Business Media

Find the latest videos and media from the market leaders.

Media Kit and Advertising

Want to reach our digital and print audiences? Learn more here.

DISCOVER MORE FROM INFORMA TECH

  • American City & County
  • IWCE
  • Light Reading
  • IOT World Today
  • Mission Critical Technologies
  • TU-Auto

WORKING WITH US

  • About Us
  • Contact Us
  • Events
  • Careers

FOLLOW Urgent Comms ON SOCIAL

  • Privacy
  • CCPA: “Do Not Sell My Data”
  • Cookie Policy
  • Terms
Copyright © 2023 Informa PLC. Informa PLC is registered in England and Wales with company number 8860726 whose registered and Head office is 5 Howick Place, London, SW1P 1WG.