![]() |
![]() |
|
||||||
|
|
||||||||
|
January 6 - 7 This is the first CCNC conference, having been put together
in only 9 months. Yet, even from the first day, it has shown the value
of focusing on technology in a public forum and not getting mired in
PR hype. Now some highlights. It is infrequent that one hears the details of a home networking strategy from an individual in such a position as Yoshiaki Kushiki, Managing Director and Member of the board of Matsushita Electric Industrial Co. His presentation was entitled “Guiding Digital Network Technology into the Consumer World.” One of the most informative aspects was the video which showed the Panasonic vision on how, in 2008, a digital network could impact the home and family. This was a scenario about the finding and placement of a puppy. In it examples were shown of: Ultra-Thin Image Communicator One value of this video is that it showed that CE devices do not exist in isolation but are supported by complex and integrated services. It remains to be seen how Panasonic will either support the development of such an infrastructure or provide it. In the Ubiquitous Network Panasonic placed emphasis on
what consumers are concerned about: There is also a chasm in realizing this Ubiquitous Network.
In Panasonic’s view the PC is responsible for much of this. It
has three problems: Thus, there is a clear need for consumer devices which
avoid the limitations of the PC and which can bridge the chasm. The Panasonic
solution is Digitally Networked Consumer Electronics. The foundation
of this is the DNA which makes CE devices successful: Panasonic is already implementing this strategy in Japan.
They have released products which include: Wireless plays a key role in the strategy, supporting both 802.11g and 802.11a. A 400MHz radio is also used for appliance control. More interesting is the role that power line could play. Kushiki-son cited Home Plug 2.0 which will support a signaling rate of 190Mb/s and thruput of 100Mb/s. This would enable a mixed mode system which combined wireless and power line. Panasonic also sees the need for “Ultra Hi-Speed Wireless.” Two new wireless technologies are MIMO and UWB. He predicted that MIMO would enable 100Mb/s by 2005 and be applied to cellular systems by 2010. UWB would support 1 GB/s by 2010. Kushiki-son also described a role for agent technologies, including Interface Agent, Interpreter Agent and a CE Agent. It was expected that the CE network would automatically determine the user need and conditions for such agents. TNO TNO, an independent research organization in Netherlands,
described a research study on Residential Gateways. The focus was to
determine: The study examined many RG products. One product of the work was a RG classification model. The model was, in part based on the OSI network model. This work is important in that it looks at the RG issues from an equipment provider perspective and seeks to understand where the market is going. OSGi OSGi seems to have gone silent over the last few years but in a short presentation that was certainly not the impression given today. This is an alliance and not a standards body. The focus in on creating a solution infrastructure based on creating an OSGi Service Delivery Platform. Release 3 of this platform came out in April 2003. The intent of this platform is: To provide an open software standard for service providers, device manufactures, and others to remotely and dynamically deliver, integrate and manage a rich set of services to the building, Mobil, vehicle and other environments. UWB One of the more interesting presentations was given by David Leeper, Chief Technologist, Intel Capital, Ultrawideband Networking Operation. He cited a recent statement by FCC Commissioner, Michael Powell, in the recent NPRM ET 03-108. On a real-time basis, smart radios determine their location and environment, have the flexibility to select the best frequencies to use, know how to avoid interference with existing users, and can use vacant spectrum channels. A very interesting graph was shown of near real-time radio that operated in “Sculpted” spectrum. It asked the question – Foreshadowing Things to Come? Such a radio was stated to be possible with the combination of UWB, OFDM and Smart Radios. Wireless Mesh Networking for the Digital Home Presentations were given by Sony, MSR, Philips Research and Intel. Sony Shin Saito of Sony Japan R&D, Ubiquitous Technology Laboratory, asked the question – What’s the appropriate approach to CE-Wireless Mesh Networking? Sony’s mission statement is that now is the time to establish CE-Wireless Mesh Networking into all CE products. The advantages of mesh networking it that it enables a dynamic network with no user involvement, is decentralized and ad-hoc. From a user perspective the equipment must be “easy to get smart.” The conventional networking equipment style is not acceptable to the consumer. The need is for zero-configuration and automatic authentication and security key setting. A requirement for network association and device discovery of <1 second was stated. Bluetooth was cited and an unacceptable solution to this problem. Sony’s assessment of mesh networking has resulted in the creation of CW-Net (Consumer Wireless Network) Project This project in Sony has a very broad impact, especially in technology development areas. An example was cited of one research path included: UWB Mesh Microsoft Richard Draves gave a presentation on the work being done at MSR in mesh networking. This was in two parts. The first described the value of mesh networking in enabling robust community networks. A user interface example was shown of how communities could be enabled where there is a shared network. Examples were shown of both community created content and commercial content. Apparently under the code name Chen this provided an interesting example of the role that a PC could play on such a network. The second part of the presentation described an 802.11a mesh network set up in Building 113 to test mesh network performance. The only result shown was a plot of TCP Throughput vs. Path Length. It was surprising that hop length could actually be used to improve performance. Philips Research Jorg Habetha gave an interesting presentation on the Philips
view. This was broken into four parts: The rationale given for a Home Wireless mesh is that it can be implemented in homes, such as Europe, which are difficult to service with current wireless technology, customer installation is much easier that traditional AP based networks; it is possible to bridge between various types of wireless technology and even including powerline. One of the most interesting applications for mesh is in vehicle to vehicle communications. This is for safety and what is called “Wireless Local Danger Warning.” Such a network would allow cars to communicate autonomously on road dangers. It is claimed that the major automobile companies have projects underway. An effort is underway also in 802.11 for set up a working group. When the question was asked – why? The answer was given that: Such a system could avoid 17% of all accidents and save 46% of the micro-economic costs. Government programs seek to reduce road fatalities by 50% by 2010. A case was also made for a public meshed network. This would also include Mobil access for pedestrians and automobiles. The peer to peer networking was combined with public networking for a very interesting application called Opportunity Scout. This used a cute bull’s eye interface to show how mesh networking could scout out devices, people and content and show there proximity based on relevance to the individual. Impressive. Intel Lakshman Krishnamurthy of the Communications Technology Lab, Intel give a presentation on Standards based Mesh Networking for the Digital Home. One of the factors cited, now multiple times in this panel, is that mesh due to its ad hoc nature has a much higher ease of use for consumers. The talk focused on a test run in home which used multiple AP to test mesh network performance. It was claimed that single node throughput increase, multi-node throughput increased and that latency went up by only a factor of 2ms. These results were quite interesting but only a sample of what needs to be done to test the value and performance of home mesh networks. It is expected that shortly a working group will be formed in IEEE to set a standard for mesh networking. Motorola The keynote was given by Jed Johnson, Director of Systems Engineering, Motorola Broadband Communications Sector. Key points from the presentation include: Based on Motorola’s research the value of broadband
to the home is based on three factors: These three factors are at the center of all of the product strategy. Motorola is building a wireless handheld device that allows consumers to remain connected to AOL IM anywhere in the home. A central thrust of the Motorola business model is around the managed home gateway. The intent is to provide universal network connectivity, always-on applications, customizable platform and for the management of devices. The first product which fully reflects this direction is the MS1000 Managed Application Gateway. This platform uses Linux and OSGi 2.0 along with Java J2ME. The MS 1000 is targeted to any broadband provider and thus does not have a modem. When asked if this is a cable product Jed stated that the product was designed to be broadband access neutral. In fact, the first customer is not a cable company. In response questions Jed stated that the STB market is cost driven. The lowest cost, single function box, just the cable modem, is a hot seller to the operators but the feature rich STB are not selling. Thus Jed sees the high end STB like appliances, as going to retail, while direct distribution will be the lowest cost entry STBs. Jed was also asked about the status of middleware. It is too early to forecast but OSGi is getting traction after being stalled for a number of years. Linux The panel on Linux for CE was quite positive on the prospects. Of course, it was made up of Linux supporters but some of the data they presented was interesting. For example: It is expected that Linux use will grow by 60% on a compound basis in embedded applications from 2002 to 2006. The market share will be approximately equally distributed between commercial and in house developed embedded OS’s. Japan is showing a strong migration to Linux. This is directly impacting the use of IT Tron as the embedded OS of choice. Sony and Panasonic joined together to work with Linux. This was considered sufficiently successful that the CE Linux Forum was established. There have been 7 working groups set up in the forum. OSGi Strong statements were made about the potential role of OSGi: OSGi is a network OS for the delivery of services. Its key value is its ability to remotely and dynamically deploy provision, maintain and manage applications and services. OSGi was originally established to support the home. But
the platform has become valuable in many other applications. These include: The objective of OSGi is to provide an open, common architecture for service providers, software vendors, gateway operators and equipment vendors to develop, deploy and manage services in a coordinated fashion. Wave Comments It is much too early to evaluate the value of mesh networking. Yet, over and over the value of the mesh in terms of the user and ease of use continues to be cited. Here is a technology which is quite sophisticated and, if successful, that sophistication will get buried while at the same time there is a positive consumer benefit. Most surprising was the value of mesh for automobiles. If the promises are realized this could be a massive public gain. The requirements for software in home networking run deep. This conference has certainly made this clear. The fact that the PC will exist in the home is not the issue. The more important topic is Will Windows or some variant have a strong role in CE networks? An answer, based on just this conference, would be no. |
|||||||||||||||||
| Comments? |
||||||||||||||||||