What Don't I Understand About Meshes?

Here are some initial thoughts about mesh networks. Tell me what I am missing or what I have got wrong. In particular, I would be interested in learning what constitutes a "low-bit" network and what a 'high-bit" network is in the context of meshes. Also, who knows about packet radio? Tell me if my description in the "History" section is accurate. Finally, vote for your favorite mesh networking companies (you're not allowed to nominate your own).

1. HYPERBOLE VS. REALITY. The benefits of any truly disruptive technology are at first exaggerated, but their long-term effects surprise everyone. At the moment, mesh networks are experiencing such a misvaluation. Their promoters cannot describe them without hyperbole; but they will be the mechanism by which information becomes like electricity—invisible and ubiquitous.

2. HISTORY. Mesh networks are not very new: their conceptual lineage dates back to the 1970s and packet radio, a kind of digital data transmission used by amateur radio hackers. But investments by the Department of Defense in more reliable and intelligent networks in the 1990s created a renaissance in interest in meshes; and by 2000, a variety of academic institutions (including MIT’s Media Lab) and associated startups believed that mesh networks would overthrow traditional networking and permit a number of exciting new applications.

3. DEFINITIONS. To see why this is so requires a little effort for anyone unfamiliar with networking. At first acquaintance, a definition seems dry. Mesh networks (sometimes called Mobile Ad Hoc Networks or, even less winningly, MANETs) are local area networks all of whose nodes are mobile and communicate directly with each other through wireless connections. Any device with a radio connection can be part of a mesh (although the phrase is most commonly associated with high bit-rate networks like Wi-Fi). Meshes possess no fixed infrastructure; they have no central control. Meshes have no designated routers: instead, all nodes serve as routers for each other. Data packets are forwarded from node to node in a process that network technologists call “hopping.”

4. BENEFITS. Before dismissing meshes as being mainly of interest to specialists, consider their advantages over existing networks with their hubs and spokes. Meshes are self-healing: if any node fails, another will take its place. They are anonymous: nodes can come and go as they will. They are pervasive: a single node will never encounter dead spots without network coverage because there is always another nearby node. Meshes are cheap, efficient, and simple.

5. DISADVANTAGES. Mesh networks are still developing and therefore are not without faults. At the moment, a mesh will support only a dozen nodes and no more than 3 to 4 wireless hops: they are not very big. Low-bit rate mesh networking has a standard called ZigBee that is supported by around 100 companies, including Motorola, Mitsubishi, Phillips, and Samsung; but high-bit rate communications as yet have no such standard. Most notably, mobile nodes cannot communicate with other nodes “seamlessly” (to use the jargon of the trade): therefore, voice, video, and pervasive gaming are unhappy on meshes.

6. APPLICATIONS. But some of the technologies that meshes will make possible are already emerging. Meshes will allow municipalities to create cheap or free city Wi-Fi networks (dozens of cities are at work in this area; we will be writing about how Philadelphia is building its own mesh in TR’s November issue). Meshes have obvious advantages for police officers and soldiers who want networks that are unbreakable and “horizontal” (see “Instant Networks,” TR, June 2005, and “Communicating in Crisis,” TR, May 2004). Environmental scientists like meshes because they can be used to create low-bit networks of sensors that use little or no power; such sensors can offer researchers continuous environmental data from large geographical areas (see “Casting the Wireless Sensor Net,” TR, July/August 2003). But the most important application of mesh networks will be in what used to be called "pervasive computing": that is, embedding machine intelligence into ordinary human objects like clothes, consumer electronics, and buildings, and connecting those embedded systems into smart networks.

7. ECONOMICS. The economic impact of mesh networks will be profound. There will be 16.4 billion wireless devices by 2008, according to International Corporation. The Defense Advanced Research Projects Agency estimates that around 8 billion embedded processors are built every year. The only likely way such systems will be connected is through meshes. But their most shocking business implication is this: telecommunications companies do not own them. Meshes diminish the importance of enterprises and organizations who own and manage voice communications backbones.

8. BIOLOGY. But for me, the most intriguing characteristic of mesh networks is their cybernetic quality (see “Cybernought,” TR, June 2005): meshes are self-correcting mechanical systems that possess many of the qualities of biological systems like insect societies. In “AntHocNet: An Adaptive Nature-Inspired Algorithm for Routing in Mobile Ad-Hoc Networks,” a report published by the Dalle Molle Institute for Artificial Intelligence in Manno, Switzerland, Gianni Di Caro et al. describe how Ant colonies can converge to discover the shortest path from their nest to a source of food; just so, they say, mesh networks are adaptive and can display apparently intelligent behavior without awareness. Does this mean that intelligence is an emergent property that does not require consciousness? Write and tell me at jason.pontin@technologyreview.com.

Technorati tags: mesh networks

Do We "Get" the Social Impact of Tech?

Michael Zimmer, a doctoral candidate in the Department of Culture and Communication at New York University, has posted an interesting critique of our August 2005 issue. That issue was devoted to describing the social impact of some new technologies - but Zimmer believes we don't "get" the relationship of technology and society.

He writes,

"Technology Review just doesn't understand the complex relationship between technology and society. I've been a reader of MIT's flagship magazine for a few years now, and have had mixed feelings about its contents. At times, they've provided thoughtful insights into emerging technologies and trends, but too often, they seem to ignore many of the social impacts of the technologies they exault. In short, TR too often engages in technological utopianism without properly assessing the social, value and ethical implications of our emerging technologies."

My guard was immediately up - because it is grossly unfair to accuse Technology Review of utopianism. We are, if anything, aggressively skeptical about new technologies. We worry about whether novel technologies will work as advertised, and we fret about the unintended consequences of such technologies when they do work. We are not fools: we also know that all new technologies are human artifacts and are good and bad in so far as we make them so. As I wrote in a column last December,

"Scientific knowledge... is a kind of absolute good. No one can reasonably object to understanding cellular disease. But... technology... is morally neutral. Any technology must exist in a fallen world of methods and ends, about which men and women can disagree."

But Zimmer was only getting warmed up. In our leader to the August issue, we explained that emerging technologies (our subject matter) almost always appeared in an academic, government, or commercial setting - that is, inside institutions, and not within society at large - because emerging technologies are expensive and at first have a very narrow application. But this point annoys Mr. Zimmer. He writes,

To state that TR has kept clear of the social issues related to technology simply because it deals with expensive emerging technologies ignores the fact that all technologies at all stages of development interact with society and have social consequences.

I take issue with this last point. Of course, we know that technologies are created by social beings responding to social stimuli; and of course we know that all technologies, sooner or later, have social consequences. We have never been opposed to writing about technology and society. It was simply not our focus. Consult our mission statement: we like to say that:

"Technology Review describes emerging technologies and analyses their commercial, economic, social, and political impact."

Zimmer also misconstrues the purpose of the August issue's cover story by Wade Roush. There, we described those new techologies - predominantly associated with Web 2.0, wireless, and social computing - that were specifically designed to alter the way we connect with one another. Most particularly, we were interested in what we are calling continuous computing.

Zimmer doesn't like the focus of our cover story. He writes,

"They seem to think that the only technologies that have social implications are those "social technologies" such as wi-fi, cellphones, blogging, etc. And only now that such "fun" technologies have become ubiquitous will TR delve into the social impact of technology? Unbelievable."

It must be noted at once that our August issue on the social impact of new technologies also covered a variety of social subjects other than "social technologies": We also wrote about the uses of technology in the kingdom of Bhutan. We reported how the revolution in genomics might change human diet. And we even asked Dr. Sasha Shulgin, the "stepfather of ecstasy," to explain the social uses of psychedelics.

Finally, the August issue contained lots of fun stuff, too. It was our beach issue. Mr. Zimmer needs to lighten up. After all, in our August issue, we did.

Technorati tags: society and technology, social computing, continuous computing

Help Me Write About Mesh Networks

I've decided that my column for the October print edition of Technology Review will be about mesh networks. I will blog some of my initial thoughts soon - I will be trusting to you, my readers, to help me refine and develop my ideas. But first, help me out: tell me why mesh networks are so disruptive. Why do you think they're so cool? What do they do that was not possible before?

Technorati tags: mesh networks

The Death of Print? Not Yet.

I made some informal remarks at the IMAG Magazine Leadership Conference in Atlanta, Georgia recently; the theme of my speech was the growth and significance of electronic media. The purpose of my remarks was to educate an audience of print publishers about newer media technologies. I talked about vanilla-flavored Web publishing, but I also chatted about social computing and continuous computing. Like many who follow such technologies, I believe they have irreversibly changed how publishers do business. Sure, readers are beginning to consume media very differently: egocasting is increasingly popular. But more complicatedly, the electronic media undermine the justifications for print advertising. This is a problem because revenues from print advertising have been the foundation of publishing for more than a century.

In truth, many publishers are struggling to understand how their businesses will be structured in the future. The subject is being debated within every publishing company - and, in general, youth and blind optimism seem to be the best antidotes to panic and despair.

Predictably, perhaps, my speech became a story: Another 'Death to Print' Prediction—This Time, From Technology Mag Editor by Bill Mickey of Folio Magazine.

I won't accuse Mr. Mickey of misquoting me. I was possibly unclear. The tone of my speech was also a little "cheeky," as the Brits say: I was the last speaker of the conference, and I was trying to shake up the audience. I wanted folks to understand that the sea-change in media was real and imminent. But I feel I should correct the record. My views are more nuanced than Mr. Mickey suggested.

1. Print editorial has strong demand from readers. Indeed, readers still love to read longer-format, well-researched, and well-written stories in established brands like The New Yorker, Vanity Fair, and (dare I say it) Technology Review. Readers also like to read stories where the "form factor" of portable, colorful, printed pages is important: I am thinking of celebrity and fashion publishing. But, as is reasonably well known, readers - particularly younger readers - prefer to read their news and timely research online.

Here's what's less well understood: the last advantages of print's form factor will vanish with the advent of electronic ink, a technology developed at MIT's Media Lab. Electronic ink will produce a page of flexible "paper" that will be colorful and portable - but, like the screen of your laptop, it will also be a wireless, digital platform that will be able to access everything that has been indexed and archived. With the arrival of Google Print, that could be pretty much everything. How far away are electronic ink tablets? They're coming, and soon. The remaining problem is the powersource for electronic ink: it's still too heavy, hot, and expensive.

2. Print advertising, however, is more troubled. Readers still want to read magazines, at least for now. But outside of fashion and some consumer advertising, advertisers are more and more attracted to various forms of online advertising. I shan't tire the reader by enumerating how bad the market for print advertising is; do a Google search. The most radically discomforting of these new ads are the various online advertising auctions like Google AdSense and Google AdWords. These auctions have combined to make Google the largest single "media buyer" in the world (to use the jargon of the publishing industry). The advantages for advertisers from online advertising can be quickly summarized: online ads are efficient, they are accountable, they are targeted. In short, for most advertisers, online ads offer a superior return on investment - and ROI is what matters to rational media buyers.

The conclusion to be drawn from points 1 and 2 is that while there may still be demand for print publications from readers (at least until electronic ink is practical), publishers will have to find a much larger proportion of their revenues from online advertising and subscriptions.

Think about it this way: for 100 years print advertisers subsidized print magazines and newspapers. Readers got a very expensive product for almost nothing. That's going to change. Publishers will receive the larger portion of their advertising revenues from online. Print will be supported by subscriptions: if readers want a print product, they'll have to pay for it. To spell it out: all this will mean more online publishing, and fewer print publications.

And in the long term, the sole advantages of print will be nostalgia, or fanciness: some publishers will want printed publications because such things will be rare and expensive. When electronic ink becomes cheap and ubiquitous, print properties will (in the developed world at least) be as fabulous as phoenixes.

UPDATE: 7.13.2005. And old friend wrote to mildly upbraid me for failing to note the following:

"Print is still MUCH bigger than online. Online is growing faster but I can't think of a major integrated media business for which their print business isn't an order of magnitude larger than their online effort. Online media advertising is in the single digit billions and print advertising is in the tens of billions--probably more than an order of magnitude larger. I agree with the CONCEPT that online targeting is better, but print advertising is still huge. Also, you didn't note that USAGE trends are what's staggering. Online READERSHIP is growing by 10-20%, and blogs are fastest/biggest growing segment, and many newspaper circulation efforts are in decline--losing readers. So focusing on advertising decline is actually in error. Eyeballs are moving online FASTER than advertising is chasing those eyeballs--mostly because there is a transition cost as advertisers figure out how to set budgets and strategies against new forms of media. of course, the MAJOR area that is shifting in advertising is not display, but classified, which really IS moving online at a rapid pace, and is devastating to publishers. But you seem to be talking about display advertisers, and there the record is mixed--they are moving online more slowly than readers."

The old friend is completely right: Online advertising is still a fraction of print advertising; and readers are moving to electronic media more quickly than advertisers.

That said, the important point is that print advertising budgets of media buyers are stagnant or in decline; but online budgets are growing between 20 to 100 percent a year. This means that at most publishing companies, the online property is the only platform that enjoys vigorous revenue growth. The lag between the growth in online audience and the growth in online advertising, while real, I attribute to the same cause as my old friend: the transition costs for media buyers and advertisers.

Technorati tags: Media

Dog Experiments and Poetry

Sir William Empson (whose biography, Among the Mandarins, I have been reading, and about whom I have posted ) once wrote a vilanelle, Missing Dates, one of whose metaphors derives from an experiment that reminds me of the Safar Center's zombie dogs.

He wrote,

"They bled an old dog dry yet the exchange rills
Of young dog blood gave but a month's desires;
The waste remains, the waste remains and kills."

Sir Bill loved these sorts of metaphors derived from science (which is one of the reasons why I like him so). He was a tireless reader of scientific journals, and began his academic career as a Cambridge mathematician.

Explaining the dog metaphor, Empson wrote in his notes to his Collected Poems,

"It is true about the old dog, at least I saw it reported somewhere."

John Haffenden, Empson's biographer, tracked down the paper Empson was remembering. From Haffenden's edition of the Complete Poems

"The experiment to which WE refers... is the one carried out by Alexis Carrel and described by Lecomte de Nouy in Biological Time (London 1936): 'There was at the Rockerfeller Institute, before the war, a dog nearly eighteen years old. This poor animal never stirred from its corner and could hardly get up to eat... This animal was anaesthetized, put on the operating table, and treated as follows. Carrel bled him by the carotid artery and removed nearly two-thirds of his blood. This blood was collected aseptically and immediately centrifuged, so as to separate the red blood cells from the serum. The red cells were washed in Ringer solution, recentrifuged and mixed with fresh Ringer solution to re-establish the initial volume of blood. This was then reinjected into the dog... After several days, he had regained strength and appetite... Not only did he live, but once over the operative shock, he was a different dog. He ran and barked, a thing he had not done in years. His eyes were clear, his eyelids normal. His coat started to come in; he was gay, active, and most important of all, he was no longer indifferent to the charms of the other sex.'"

But as Empson remembered, the rejuvenation didn't take. It didn't last. The waste remained and killed.

Zombie Dogs

I read today that Pittsburgh's Safar Centre for Resuscitation Research has developed a technique in which a dog's veins are drained of blood and filled with ice-cold saline. Then scientists resucitated the dogs (who had been dead for three hours) by reinflating them with warm, living blood. The experimenters claimed the hounds suffered no brain damage, although I am not sure how they demonstrated this - I mean, it's not like Rover was busy solving quadratic equations or composing sonnet sequences the next day.

In stories of this sort, the word "dead" is often held between the pincers of inverted commas - but on this occasion we can dispense with that convention, surely. The dogs had no heart beat, nor any brain activity. They were dead-dead.

The Safar Centre plans to begin trials on humans next year. They say the first, obvious applications for the technique would be in surgery. Indeed, why not operate on dead people, then bring them back to life? But there are clearly other, more radical applications, too - many of which throw up all sorts of pseudo-philosophical questions.

This is another blow for the concept for the human soul, I fear. Where will that poor, pale thing reside for the three hours, eight hours, or twelve months that its housing freezes on a hospital slab?

Technorati tags: biomedicine, metaphysics