Custom Search

Monday, June 30, 2008

Any Lessons Zigbee can Learn from Bluetooth (2)

Part 1 Part 3 Part 4

The third, in tight association with the previous two points, Bluetooth is able to get to the magic heaven of consumer electronics market, where, no matter it will be in use or not, Bluetooth is a default configuration. Let's see some numbers first. In 2007, 505 million handsets or 47% of total handsets sold worldwide equipped with Bluetooth. At the same time, only 69 million Bluetooth headsets shipped. Headset attach rate, which simply is the ratio of the above two volumes, could be an inaccurate but relative index to indicate how many Bluetooth on a cell phone is actually in use, because it is still the dominate application of Bluetooth. The attach rate of 2007 is 14%. (Because of this fact, I have challenged the success of Bluetooth is somewhat fictitious in my blog "Another Emperor's Clothes - Bluetooth")

Nevertheless, economically Bluetooth is a big success in terms of volume and revenue, which are the only things that those semiconductor companies and patent holders care about. As a side effect, Bluetooth ICs to some extend becomes commodity with huge volume and cheap price. Some companies may say they do not admire such market. Their grapes are really sour.

Lack of strong pull force from system developers and integrators, Zigbee has been driven unfortunately into the complex networking application, where it is hard to identify such a central equipment as cell phone that has many leading vendors in hot competition, and that devices of varies usages can be attached to.

The fourth, when the technology is in implementation, Bluetooth is much easier, simpler and more straight-forward. Bluetooth only supports peer to peer and to its utmost a star network with 7 slave nodes. Bluetooth can be expanded to support “scatter net”, but this kind of application has never got into the core. Zigbee is exactly on the contrary, peer to peer and star network becomes the niche, not in terms of the market needs but the willingness that this technology wants to address. Zigbee is devoted to bigger and more complex topologies consisting of at least 10s to 100s nodes. The result of this difference? Bluetooth is for consumers and Zigbee is for engineers, well-trained ones.

Setting up a Bluetooth network is as easy as making an earpiece talking to a phone. It may take you a bit long time, but comparing to Zigbee, it is still a piece of cake (still, recently in California, some Bestbuy stores charge $10 or more for pairing a headset to a phone). For Zigbee, there is a pivotal step you can not skip, which is called network planning or network infrastructure. Does this sound scary? Looks like setting up a cellular phone network. This is about the network topology, RF scanning, signaling trouble shooting. Companies of Zigbee usually provide hardware and software tools to help you to plan, analysis and maintain such as a network. Set up a Zigbee network requires ambition. (If Bestbuy stores, I do think they are capable, charge for setting a Zigbee network, I guess the price would be $10k or more)

In short, setting up a Zigbee network vs. Bluetooth network, the difference is a project vs. a task.

The cost of this difference is huge for Zigbee, as the complexity will lengthen the learning curve and lead time, add the cost for implementation and maintenance and increase the likelihood of failure.

Actually, at 802.15.4 layer, the widely used star topology is already supported. Zigbee just added some unnecessary complexity.

Part 1 Part 3 Part 4

No comments:

Post a Comment