element14 Community
element14 Community
    Register Log In
  • Site
  • Search
  • Log In Register
  • Community Hub
    Community Hub
    • What's New on element14
    • Feedback and Support
    • Benefits of Membership
    • Personal Blogs
    • Members Area
    • Achievement Levels
  • Learn
    Learn
    • Ask an Expert
    • eBooks
    • element14 presents
    • Learning Center
    • Tech Spotlight
    • STEM Academy
    • Webinars, Training and Events
    • Learning Groups
  • Technologies
    Technologies
    • 3D Printing
    • FPGA
    • Industrial Automation
    • Internet of Things
    • Power & Energy
    • Sensors
    • Technology Groups
  • Challenges & Projects
    Challenges & Projects
    • Design Challenges
    • element14 presents Projects
    • Project14
    • Arduino Projects
    • Raspberry Pi Projects
    • Project Groups
  • Products
    Products
    • Arduino
    • Avnet & Tria Boards Community
    • Dev Tools
    • Manufacturers
    • Multicomp Pro
    • Product Groups
    • Raspberry Pi
    • RoadTests & Reviews
  • About Us
  • Store
    Store
    • Visit Your Store
    • Choose another store...
      • Europe
      •  Austria (German)
      •  Belgium (Dutch, French)
      •  Bulgaria (Bulgarian)
      •  Czech Republic (Czech)
      •  Denmark (Danish)
      •  Estonia (Estonian)
      •  Finland (Finnish)
      •  France (French)
      •  Germany (German)
      •  Hungary (Hungarian)
      •  Ireland
      •  Israel
      •  Italy (Italian)
      •  Latvia (Latvian)
      •  
      •  Lithuania (Lithuanian)
      •  Netherlands (Dutch)
      •  Norway (Norwegian)
      •  Poland (Polish)
      •  Portugal (Portuguese)
      •  Romania (Romanian)
      •  Russia (Russian)
      •  Slovakia (Slovak)
      •  Slovenia (Slovenian)
      •  Spain (Spanish)
      •  Sweden (Swedish)
      •  Switzerland(German, French)
      •  Turkey (Turkish)
      •  United Kingdom
      • Asia Pacific
      •  Australia
      •  China
      •  Hong Kong
      •  India
      • Japan
      •  Korea (Korean)
      •  Malaysia
      •  New Zealand
      •  Philippines
      •  Singapore
      •  Taiwan
      •  Thailand (Thai)
      • Vietnam
      • Americas
      •  Brazil (Portuguese)
      •  Canada
      •  Mexico (Spanish)
      •  United States
      Can't find the country/region you're looking for? Visit our export site or find a local distributor.
  • Translate
  • Profile
  • Settings
Wireless
  • Technologies
  • More
Wireless
Blog Industrial Wireless - Channel Access Latency
  • Blog
  • Forum
  • Documents
  • Polls
  • Quiz
  • Events
  • Mentions
  • Sub-Groups
  • Tags
  • More
  • Cancel
  • New
Join Wireless to participate - click to join for free!
  • Share
  • More
  • Cancel
Group Actions
  • Group RSS
  • More
  • Cancel
Engagement
  • Author Author: gervasi
  • Date Created: 19 Jul 2013 9:23 PM Date Created
  • Views 621 views
  • Likes 1 like
  • Comments 3 comments
  • wifi
  • industrial_wireless
  • latency
Related
Recommended

Industrial Wireless - Channel Access Latency

gervasi
gervasi
19 Jul 2013

In the past few weeks, I’ve written about challenges of industrial wireless (with a focus on Wi-Fi): multipath and interference.  This last post will be about the greatest challenge: channel access latency.

 

What is Fast?image

What does it mean when we ask how fast something is?  If we ask which is the faster way to move people, a bus or a motorcycle, we interpret the question to mean which one makes a trip faster.  The bus moves more people per unit time, but that’s not what we think of as “fast”.  When we ask how fast a data link is, though, usually the answer is in data per second.  To find out how long it takes for a piece of data to propagate through the network, we ask about “latency”. 

 

Low-Latency High Packet Rate

Many industrial networks need to send many small packets, maybe 60 bytes in size, very

frequently.  Such users define their networks not in Bps but in packets per second. 

 

Often these users deploy Wi-Fi networks.  Each RF packet takes less than 100us.  If the channel were shared perfectly, you could fit 10000 RF packets on the channel every second. 

 

Wi-Fi Channel Sharing

802.11 has an amazing scheme, called the distributed coordination function (DCF), to share the channel with Wi-Fi users, hoppers, and anyone else using the channel.  The DCF performs clear channel assessment (CCA) for a very brief period of time (between 16us and 63us, depending on the priority of the packet to be sent) before transmitting.  The transmitter waits for an ACK from the receiving station.  It must wait long enough for the receiver to verify the packet is good and send an ACK, plus 5us per mile of propagation delay. 

 

If the transmitter does not receive an ACK, it tries again at a random point during a contention window (CW).  The contention window on the first retry is short (default is 279us).  It doubles with each successive to a maximum of 9ms.  The frame is thrown out if no ACK is received after the retry limit. 

 

DCF in an Industrial Setting

The point of random backoff is to prevent two users from backing off the same amount of time.  As the number of users and packets goes up, however, contention happens more frequently.  Not only are more packets lost to collisions, but more often transmitters use larger CWs, wasting channel time. 

 

Users may configure radios on their network to use shorter contention windows to increase their chance of getting the channel.  They may also increase the CCA threshold, so their radios may transmit over weaker networks.  But these measures do nothing to share the channel more effectively among users in one’s own network.  Setting a shorter contention window will cause more collisions.

 

Data Rate Fallback

Some radios have fallback algorithms that select a lower data in response to frequent retries.  You typically need -64dBm of signal to receive a 20MHz WiFi signal at full data rate.  At the lowest data rate, you can have a reliable link down to -82dBm, almost 100 times less signal.  A good fallback algorithm uses the fastest rate that is reliable for the signal strength seen at the node it’s transmitting to.

 

Simplistic fallback algorithms seen in consumer equipment fall back to a lower data rate if it’s having to do many retries.  This is good if the retries are due to low signal strength.  If they’re due to collisions, however, a lower data rate makes the problem worse.  Lower data rate packets take longer to send, increasing the chance of collisions. 

 

Coordinated Channel Access

The original Wi-Fi standard provides a point coordination function (PCF) to allow one node to control which radios on the network can access the media.  The 802.11(n) spec provides for a hybrid coordination function (HCF) controlled channel access (HCCA).  In these schemes there is never an issue of nodes in the same network transmitting at the same time.  There is latency associated with waiting for a node’s turn to transmit, but this is not a waste of other nodes usually have traffic to pass during their turn. 

 

I would love to see PCF or HCCA implemented in a product.  I only know of it from the 802.11 specs. 

 

Conclusion

Achieving low latency in a network with a high packet rate is the biggest challenge to industrial wireless users.  The issues of multipath, interference, and channel contention covered in these recent posts are rare and can usually be mitigated by tweaking parameters.  Wi-Fi is way more than a protocol for connect to the Internet at coffee shops. 

 

Further Reading on Wi-Fi in General

Next Generation Wireless LANs - Spoonfeeds you the nuts and bolts of 802.11(n) and the earlier Wi-Fi standards.  The best book on Wi-Fi.

802.11 Wireless Networks: The Definitive Guide - Very dated because it was published before 802.11(n), but easy to read

The 802.11 Handbook: A Designer’s Companion - Also dated, but I’m including it because it prevents so much information so densely in an easy-to-follow way. 

  • Sign in to reply

Top Comments

  • gervasi
    gervasi over 12 years ago in reply to DAB +1
    @DAB: Yes. This is a subset of the problem you describe. You have plenty of bandwidth. You connect dozens of industrial devices all sending small packets, nothing near the max capacity. Once it reaches…
  • DAB
    DAB over 12 years ago in reply to gervasi

    Yes, a lot of the collision algorithms have not been updated in the last 30 years or so.

    They work, but they could be greatly improved with the newer technology avaliable today.

     

    I know that there has been some good network analysis done, but I am not sure that anyone has a new algorithm that works any better, but we can hope.

     

    DAB

    • Cancel
    • Vote Up 0 Vote Down
    • Sign in to reply
    • More
    • Cancel
  • gervasi
    gervasi over 12 years ago in reply to DAB

    @DAB:  Yes.  This is a subset of the problem you describe.  You have plenty of bandwidth.  You connect dozens of industrial devices all sending small packets, nothing near the max capacity.  Once it reaches a point where some collisions occur, the contention window increases, causing fewer collisions but more wasted channel time.  If the fallback algorithm is crappy, it dials back the data rate (thinking slow data rates are more robust) causing even more collisions b/c the packets take longer and there is less free time. 

     

    This is a frustrating problem b/c there's plenty of capacity.  It's like a room with dozens of people who only want to say a few words but they accidentally talk over each other.  Then they politely stop and wait to see if someone else needs to talk.  Then they talk slower to be hard.  All of this wastes time, and they can't get a few words out during a one-minute period. 

     

    There are ways a master can assign everyone a time slot, but that has its own issues and I've never seen it implemented on a high-throughput radio system. 

    • Cancel
    • Vote Up +1 Vote Down
    • Sign in to reply
    • More
    • Cancel
  • DAB
    DAB over 12 years ago

    There are many approaches to wireless networks about access, traffic levels and latency.

    The wireless provider does a level of usage analysis and then implements enough channel to reach the 99% or so, knowing that during peak access times, some people will lose their connections.

     

    I have seen some articles where they are considering the use of intelligent tracking algorithms that will follow the wireless device and preallocate the adjoining cells based upon position and direction of movement.  While these algorithms will improve service, it will also raise questions about the big bad government tracking individual devices.

     

    Face it, there are a lot of good reasons for the "system" to know where you are.  If they would setup a panic button on the wireless devices, the local emergency response teams could provide much faster service during an emergency.

     

    As for service, I think most people are satisfied for the most part.  If you know when peak times occur, you can choose to wait a while before trying to connect, assuming that you do not have an emergency.

     

    The wireless providers will always assess the number of devices to support verses availability.  If there is a swell of user demand, they may adjust their allocation numbers and install more channels, but that is only possible if they have the bandwidth to use.

     

    Thanks for the good series of blogs, they have been very useful to increase the awareness of wireless issues.

     

    DAB

    • Cancel
    • Vote Up 0 Vote Down
    • Sign in to reply
    • More
    • Cancel
element14 Community

element14 is the first online community specifically for engineers. Connect with your peers and get expert answers to your questions.

  • Members
  • Learn
  • Technologies
  • Challenges & Projects
  • Products
  • Store
  • About Us
  • Feedback & Support
  • FAQs
  • Terms of Use
  • Privacy Policy
  • Legal and Copyright Notices
  • Sitemap
  • Cookies

An Avnet Company © 2025 Premier Farnell Limited. All Rights Reserved.

Premier Farnell Ltd, registered in England and Wales (no 00876412), registered office: Farnell House, Forge Lane, Leeds LS12 2NE.

ICP 备案号 10220084.

Follow element14

  • X
  • Facebook
  • linkedin
  • YouTube