"6LoWPAN" is "Internet Protocol Version 6 over Low-Power Wireless Personal Area Networks." IPv6 allows a total of 2^128 unique addresses (compared to the mere 2^32 that were available in IPv4). Thatís about 3.4 x 10^38 available IPv6 addresses, compared to about 4 billion (4 x 10^9) IPv4 addresses.
Visionaries foresee a world where everything is connected to everything else by a giant IPv6 network. OK, fine.
Letís assume that some fraction of these devices will be connecting via wireless means in unlicensed spectrum, known in the U.S. as ďPart 15Ē, or ISM (Industrial, Scientific, and Medical) bands. The best-known ISM band is 2400-2483.5 MHz, available pretty much worldwide and used for WiFi (802.11b/g, etc.). In the U.S., another popular band is 902-928 MHz and there are similar sub-1 GHz allocations in other parts of the world (868 MHz in Europe, for example).
I worry that all sorts of equipment manufacturers will jump on the wireless-connectivity bandwagon in the Part 15 bands. Some applications will not be well-suited to these bands. Why, you ask?
Hereís a quote from the Part 15 rules.
ď15.5 General conditions of operation.
(b) Operation of an intentional, unintentional, or incidental radiator is subject to the conditions that no harmful interference is caused and that interference must be accepted that may be caused by the operation of an authorized radio station, by another intentional or unintentional radiator, by industrial, scientific and medical (ISM) equipment, or by an incidental radiator.Ē
That means that 6LoWPAN devices, like all other devices operating in the Part 15 bands, using a wide variety of protocols and signal formats, must not create interference to any other devices, so the transmitters must be squeaky clean. Protocol etiquette and good RF design practice can reduce if not solve this problem.
It also means that these Part 15 wireless IPv6 nodes must accept any interference from ANY OTHER RADIATOR. Doesnít matter if it is another 6LoWPAN device, a nearby authorized radio station, a nearby wireless device using another protocol in the same band, an unintentional radiator, or whatever.
My office is at one end of my house. I use Wi-Fi (802.11b) from the office to the wireless hub at the other end of the house. The kitchen is between the hub and the office, and the microwave oven leaks enough of its 1100 Watts of 2.4 GHz RF to clobber the connection. If someone uses the 2.4 GHz cordless phone from the kitchen in the family room where the hub is located, it clobbers the hub and I am off-line. ďClobberĒ is my term meaning ďcause sufficient interference to disruptĒ. Even if the interferer is off-channel, itís possible to cause problems by driving the receiverís first stages into nonlinearity or even compression. These interference cases are annoying, but I can live with them. And some day Iíll improve the frequency coordination in my house and buy some new devices. The 5.8 GHz family room cordless phone does not bother the hub, so when the kitchen phone finally dies, itíll be replaced by another 5.8 GHz phone.
I think some of the applications that will try to use Part 15 RF bands for IPv6 connections will fail because of interference issues. If the visionaries are correct, and we someday have several billionÖor trillion/quadrillion/quintillion wireless nodes trying to coexist, interference issues are inevitable. Itís not impossible to make these things work, itís just really hard, involving filtering, antenna patterns, and knowing the worst-case local RF environment. Consider the consequences of using unlicensed spectrum for your application very very carefully and plan accordingly.
Mmm.. I think this is not a real problem, because devices under 6LowPAN use a very low duty cycle, so they are using the channel only a few period, and do not collapse other devices. Wi-fi uses extensively the medium, but not the WSN.
"Itís not impossible to make these things work, itís just really hard..." Often it's not "hard" but "more expensive," meaning "more than someone wants to pay." I've peeked inside lots of RF boxes and asked myself, "How did this EVER get certified?"
An interesting side effect of this whole Part 15 thing, mentioned in the original article but usually NOT comprehended by the general public, is the "must NOT interfere with a licensed user" aspect. Someone who purchases a consumer Part 15 device that interferes with a licensed user of the same spectrum is required BY LAW to STOP the interference. This is often difficult to explain to a non-technically-savvy individual :-) It is really hard for them to understand that they can't use something they bought. It is difficult to get across the notion that the law protects the licensed user and not them with their (probably cheap) consumer-grade item...
My AT&T U-Verse gateway has ethernet jacks and wi-fi. I figured there was zero benefit to using wi-fi with my desktop PC three feet away, so I plugged in with a 2m Cat5 cable. Wrong! A thunderstorm blew in. I shut my PC down as a precaution, even turning the surge arrestor outlet strip off. A utility pole around the corner took a hit and my motherboard was toast. I bought a cheap USB 802.11g radio. As the EMC gurus say, only ground your system at one point.
I was thinking about this the other day. We used to use wireless communication to distribute signal broadly, as in a commercial radio station. Everyone seems to want to access this in a wired connection now, the internet.
Meanwhile, for simple point to point short range communication, we used to use a wired connection, as in from your CD player to your preamp. Everyone seems to want to use a wireless connection for this now, like from your keyboard to your computer. Interesting, ain't it?
I had a problem with interference from a cordless phone operating on 2.4Ghz. The Wifi router was set to auto channel, as there were no competing Wifis around that always meant the first channel. I set the Wifi router onto a fixed channel at the end of the channel list. Problem solved.
Now I am inclined to see how much disruption a leaky microwave oven can cause in some commercial areas! Of course, it certainly is correct to assert that wired connections are both more reliable and more secure, and now, we realize, they are more likely to be free of interference. The harsh reality is that for a huge realm of connections, wireless is a stupid choice, adding cost, complexity and additional power consumption , and providing only a marginal, at best, benefit. So here is the question again, that was asked a few years back, which is: "what happens when all of the channels are occupied?" How will the new toy work when all of the channel frequencies and all of the time-slots are already in use?
Of course, all of this is ignoring the question of what benefit is delivered by a toaster with an IP address. That is a separate issue altogether.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for todayís commercial processor giants such as Intel, ARM and Imagination Technologies.