Friday, August 26, 2005

Embedded Systems – The New Center of Gravity

From NI Week 2005 it became clear that Embedded Systems Design programming is a key area to explore for virtual instrumentation. The center of gravity is shifting from personal computing to embedded computing due to the development of VLSI system level integration and system-on-a-chip technologies. Embedded systems have been around for sometime but were typically implemented in a monolithic manner with centralized control. The next generation will be implemented with distributed control working in a parallel manner in order to optimize cost, power, security, and performance.

Finding a way to make the systems more programmable is one challenge. Finding someone to do the programming for you is another. Check out the number of available jobs for embedded programming on the web at the Feedster Blog site. The Craigslist shows over 100 open jobs in the SF bay area alone for embedded programmers.

A great site for learning more about embedded programming is Embedded.com The Embedded Conference held in April, 2006, in San Francisco is another great venue for learning about the latest technologies.

If you are working with embedded systems, I would like to hear from you. You can reach me at hall.martin@ni.com.

Best regards,
Hall T. Martin

Friday, August 12, 2005

Mesh Networking – A New Round of Players

Mesh networks also called ad hoc networks communicate with each other through wireless connections. Information “hops” from one node to another without having to go through a router system. The benefits include

Self-healing – the network can route around a dead node.

Anonymous – nodes can join and leave the network at will.

Pervasive –self-healing with ability to grow the network through the addition of nodes provides for a robust network.

Lower cost -- an ad hoc network is cheaper than one with a fixed infrastructure.

Some of the early applications for mesh networking will be sensor networks, consumer electronics, and fixed wireless applications. It’s nice to see the consumer electronics world making use of this technology as that segment can drive volume and demand in ways that the industrial world cannot.

City-wide wireless networks are also making use of mesh networking technology. Tropos makes a system that uses mesh networking as the core architecture of it’s metro-wide WiFi solution which can be used to read utility meters, parking meters, and even monitor areas for crime.

The key advantage to mesh networks is that telecom companies don’t own the networks. Several new startup companies are taking advantage of this. PacketHop is a startup that turns a set of devices into a mobile ad hoc network using 802.11 technology. With the proliferating number of mobile devices using WiFi, PacketHop hopes to leverage that installed base into an autonomous mesh network which doesn’t require an infrastructure. Another company, Firetide also makes a play for mesh networking with its Hotspot software. Kiyon developed a single radio platform that also leverages the installed base of WiFi devices. For those requiring high reliability there is the Strix System which uses multi-radio, channel, and RF for high performance applications.

If you are interested in building your own mesh network system, there’s a number of Open Source options available. Check out this piece on the O’Reilly site for options and a short tutorial on how to use one of them.

The two issues holding back the development of mesh networking is first, the unreliability of wireless links. Battery power (or the lack thereof) brings down many nodes, while the spotty coverage of wireless signals makes other nodes unusable. Second, the lack of standards holds back progress. There are efforts under way by those who promote ZigBee, but in talking with many of the adherents, they find the proposed standard full of weaknesses and most only apply portions of the standard (the portions they like) to their products.

If you are working in mesh networks, I would like to hear from you. National Instruments will be holding their annual NI Week conference in Austin next week (August 15-18). During this conference several sessions will focus on wireless sensor networks including mesh networking. If you are interested in wireless sensor technologies, I recommend you join the session on Thursday, August 18, at 1:30.

Best regards,
Hall T. Martin

Friday, August 05, 2005

Open Source – Is the Future Open?

Open Source continues to grow in scale and acceptance. It basically means that software should be a shared resource. Wikipedia defines Open Source as freely available source code and documentation that is available to anyone. It goes on to list ten points that software must meet to be called Open Source.

There are several flavors of Open Source. The first is public domain which means anyone can do anything with it. The second is BSD which requires the user to acknowledge the original developer. The third is General Public License which requires a user to redistribute the source code along with any modified binary code. Then there is Shared Source which means you can look at the code but you can’t change it or recompile it. This is considered a step backwards in the effort to move toward an Open Source world.

Open Source is well established in the server business but its growth has been slower in other areas because even though the software may be free, the implementation is certainly not. Open source software comes with bugs and spotty feature sets that require additional work to make it useful for a particular application. Open source purveyors plan to mitigate this risk by creating a rating system that grades a piece of code on factors such as quality, security, documentation and more. For more details on the rating systems, you can find it at this site.

In the world of virtual instrumentation, LabVIEW VIs is one example of Open Source program code. You can download VIs from a variety of areas. I think some of these VIs could benefit from a rating system like the one listed above.

The main arguments against Open Source revolve around ownership of source code and the benefits to the developer of the code. A new concept called creative commons makes code available but only under certain conditions and with limited rights. Examples include letting others use your source code but only if they give you credit for your material. Another example is letting someone use your code but ensures they make no money from it. A real-world example of creative commons licensing is the book “Democratizing Innovations” by Eric Von Hippel who coined the term “Lead User”. You can pay $20 for the book version, or you can download it free at this web site under a creative commons license. Which do you think will generate more readers?

For a full range of resources on Open Source, including licenses, check out this site.

If you are working with Open Source technologies, I would like to hear from you. Please email me at hall.martin@ni.com.
Best regards,
Hall T. Martin