Slashdot Mirror


New Chips Enable 2.4 GHz Sensor Networks

mindless4210 writes "Oki announced today that the world's first fully compliant IEEE 802.15.4 and ZigBee chip has been developed. The technology promises to start a new generation of wireless sensor networks, utilizing the unlicensed 2.4 GHz spectrum. The new IC integrates the digital circuit-based MAC and PHY with the analog circuit-based RF onto a single chip. The company also developed a kit which enables fast production of sensor networks which could control air conditioning, lighting, fire alarm systems, and many other applications. The low power consumption of the chip enables multi-year operation with only dry-cell batteries."

15 of 104 comments (clear)

  1. Not sure why this is a "first" ... by rskrishnan · · Score: 5, Informative

    Lots of companies have single-chip-2.4Ghz radio ICs. Like nvlsi.no or ChipCon. And NVLSI is better by a mile - quite literally a single chip 2.4GHz radio - make a sensor network in a few weeks .... if you find someone to help you solder the QFN pkg (goddamn SMD pkgs).Perhaps they bundle the ZigBee protocol within the IC ??

    1. Re:Not sure why this is a "first" ... by Dielectric · · Score: 4, Informative

      I think it's just the first production ZigBee chip, which, until now, has been total vaporware. Nordic and ChipCon and Cypress and others have had working 2.4GHz wireless chipsets for quite a while, but they're all proprietary, not a recognized IEEE standard.

      Hate the QFN? I agree, that would be pretty hard to solder down by hand. Check out the Cypress wireless chipset, you can get an SOIC which is actually pretty easy with a fine tip.

    2. Re:Not sure why this is a "first" ... by Dielectric · · Score: 2, Informative

      $250 for five boards is about right, actually. Your best bet on a budget is to make a friend at someplace like Lucent or Motorola for example.

      One more option would be to re-flow the boards at home in your toaster oven. No, really, I'm serious. You can use solder paste and an ordinary, inexpensive toaster oven to handle the reflow. I saw a well documented project that did this on the web somewhere, which should be easy enough to find with Google.

    3. Re:Not sure why this is a "first" ... by svirre · · Score: 2, Informative

      I think it's just the first production ZigBee chip, which, until now, has been total vaporware

      Actually we (Chipcon) launched a 802.15.4 compliant device in november last year. Datasheets and reference designs are availible here .

  2. So much for 802.11b/g by russotto · · Score: 5, Informative

    I'm glad my WiFi network is 802.11a. If this goes into wide use there's going to be so much noise in 2.4Ghz that 2.4Ghz wireless is going to be really hard to continue using. (Heck, it already is)

    As to the techno-babble, most 802.11a/b/g cards require a separate chip for the MAC (which handles the 802.11 level 2 protocol -- some chipsets do most of this on the host), the PHY (which handles the digital signal processing) and the RF section (which is black magic). Putting them all on one chip allows for smaller and hopefully less power-hungry devices.

    1. Re:So much for 802.11b/g by mindless4210 · · Score: 2, Informative

      I really don't think this will cause a great deal of interference in 2.4 GHz. It's for a sensor network that uses low data rates, plus the 802.15.4 standard is meant to create a PAN, so it's reach isn't likely to be that far.

      --
      Wireless News www.DailyWireless
  3. Re:Great... by Mondak · · Score: 2, Informative

    From what I understand, this is not really due to the spectrum itself, but the quality of components used when building the equipment such as your broadband modem. For example, a Microwave operates at 2ghz. If your hardware used cheap filters (like capacitors) when they contructed your device, these signals can interfere. If they spent the money up front on good components, they would not have this problem. It is really only a matter of a few cents when choosing components, but so many times - especially when dealing with the consumer market - there is price pressure on what the finished product has to cost. They shave in every area they can including the filters that make sure microwaves and cordless phones do not interfere. The same thing can happen at 5ghz as well if the hardware is not correctly specified.

  4. Re:This is news for nerds? by Anonymous Coward · · Score: 1, Informative
    I'd prefer a pile of naked men [drudgereport.com]

    I'd prefer a warning if your going to post a link like that. I thought I was going to hear what Drudge thought, not have to think about those pictures all night. Stop being an asshole. BTW, I don't look at the American war crime pictures either. I'll stick to porn.

  5. If you like this kind of stuff... by xplosiv · · Score: 4, Informative

    I suggest you check out CocoonTech.com, it's a site dedicated to Home Autmation, Theaters and Security, This article got posted there this morning before it appeared on slashdot ;) Z-wave, which is 'similar' to ZigBee is one of the newer but fast growing protocols in the HA world. There have been many times where manufactures have said they have developped an x-10 killer protocol, but so far ZigBee & Z-wave seem to be ones of the few which might actually become widely accepted (or in some cases is being deployed already). Home Automation for the masses!

  6. Re:How to monitor temperature? by xplosiv · · Score: 3, Informative

    Most people in the HA community use 1-wire sensors (you can get a few free samples from the Dallas Maxim site) as they are very cheap. You can interface them with a system using the serial port for a relatively low price too. Check out this thread on CocoonTech.com, it's discussing this technology, and all the interface available: 1-wire thread. As for your lights/appliances, I recommend good quality x10 switches (such as SmartLinc or Lightolier), or you can go with Z-Wave, which is similar to ZigBee, it's wireless, and each switch can act as an RF router to get the signal as far as possible. Pretty cool stuff. My entire house is automated this way, I can even start my car using voice, phone, web, email, you name it (right now it does it when I wake up in the morning). Doesn't have to be expensive either. If you do insist on still going with Ethernet, You would probably want to use a Global Cache device, which can process inputs, control outputs, IR and more all over IP.

  7. Re:Great... by jhouserizer · · Score: 2, Informative
    Yes, but...

    To my knowledge, there are A LOT MORE consumer devices that operatate at or near the 2.4GHz range as compared to devices that operate at or near 5GHz. Cordless phones are the perfect example. -- Thus, there is a lot more interference at the 2.4GHz range, and thus a lot harder for filters to actually do their job.

  8. Re:Unlicensed? by Anonymous Coward · · Score: 1, Informative

    It means that the band is set aside for use by unlicenced users.

    Actually the 2.4 band is shared. The licenced users are hams. They actually get priority on that spectrum but for the most part stay clear of that mess.

    Excellent breakdown of the spectrum:
    http://www.jneuhaus.com/fccindex/spectrum.html

    2.3 to 2.9 GHz:
    http://www.jneuhaus.com/fccindex/2_3_ghz.html

  9. Re:Split by drinkypoo · · Score: 2, Informative

    From the very same page: "Data rates of 250 kbps, 40 kbps, and 20 kbps". The lowest rate is five to ten times faster than a dialup modem, which is enough for very-low-bitrate video, and a decent stream of still images, or a fairly decent-quality mono audio stream when compressed (or about telephone quality without compression.)

    At its high end, it's nearly twice as fast as ISDN, which is commonly used for fairly decent-quality videoconferencing.

    Using a mesh network one could spread the traffic out between nodes and attain fairly high data rates in many situations, even with modems slower than these.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  10. Re:Great... by Beryllium+Sphere(tm) · · Score: 2, Informative

    Microwave ovens are at 2.4GHz, same as WiFi.

    This isn't coincidence. The reason the 2.4GHz band was available for innovation is that regular spectrum users considered it unusable due to the industrial equipment operating there. It's called the ISM band, for Industrial, Scientific and Medical.

    Microwave ovens are shielded for safety reasons but they start off ~20,000 times more powerful than a WiFi card. The noise is hard to filter because it's surprisingly broadband. The ultra-cheap power supplies drag the magnetron frequency up and down from resonance.

    In general you're absolutely right, though. Most consumer RF equipment is too cheaply built to tolerate strong signals outside the intended operating frequency.

  11. No news by Jeff+Kelly · · Score: 2, Informative

    First of all Zigbee is no protocol for wireless sensor networks. The curently supported network topology is to inflexible and the maximum number of devices per subnet (255) is not nearly big enough that Zigbee is of much use for ambitious installations. There are better solutions for sensor nets and zigbees priorities lie in differentz directions than that.

    Secondly neither 802.15.4 (The phy and mac layer) nor the Zigbee Protocol (network and app layers) are fully specified. The current specs lack important sections like flexible network topologies (currently only some kind of tree topology is supported), ad-hoc-networking, location of devivces and several other features which the Zigbee alliance proposed and announced but are not documented right now.

    In its current form zigbee is not much more than some kind of "low power usb". You can network up to 255 devices with several kilobit/s datarate, but that is about all you can do. All the features that make up good sensor networks are either not yet specified or not even part of the proposed zigbee spec.

    There are other and better suited projects for such applications and there are many other companies which have 2,4 GHz sulutions ready for at least to years now. Nordic VLSI for example with their nRF24E1 (www.nvlsi.no). Although in my opinion there are other frequency-bands better suited than 2,4 GHz. It might be licence-free but is much to crowded to be of good use. Because of that many companies also offer 868/422 MHz solutions.

    These solutions might lack the zigbee ready logo, but there are numerous other projects which are in my opinion better suited for sensor networks. The most popular being at the moment tinyos (http://webs.cs.berkeley.edu/tos/) an open source os for sensor networks with an extensive library of protocol modules to use fpr your own sensor application. You can even download circuit plans for your own custom design.

    Jeff