IN the past I've suggested that one of the essential problems facing the US armed forces in Iraq, and more broadly all modern armies in a world of messy little wars, is best thought of in economic terms. In any firm, there is a theoretical optimum combination of capital and labour used to process one unit of land(in the economic sense - that is, including all natural non-human resources) that maximises productivity and hence profitability. You can draw a graph showing the possible combinations for any given version of the problem, and it will be a convex curve about the origin. The optimal combination is the furthest point on the curve from the origin. That's the theory.
When you turn up at a row at a petrol station with a 155mm self-propelled howitzer and crew, it should be pretty obvious that you are too far towards the capital end of the scale, if we assume that material is equivalent to capital in the destruction industry. You have a factory crammed with stuff until it's too dangerous to walk through the door and too expensive to turn any of it on. Now, I've just been reading up on cognitive-radio systems, and this made me think of the problem. Cognitive radio means that a radio device implemented in software can intelligently change its behaviour in response to the radio environment. That's a minimal definition. Think seeking out the part of the radio spectrum that is least busy, or setting different channels different priorities. Another use of the concept, the one I'm professionally interested in, is so-called overlay networking, where a communication device chooses between the different forms of radio access in its area depending on some criterion or other. For example, your phone call might go over a free Wi-Fi network if there's one around, over the cheapest 3G data channel, failing that over the cheapest 3G voice channel, failing that over good old GSM - that's assuming you set it to "find cheapest" rather than "find most reliable" or "find fastest" or "find the Crazy Frog".
A huge amount of research in this field, though, is the work of the military. Practically everyone involved seems to work for DARPA, whether directly or at one remove. And some of their plans for smart radio are of stupefying ambition. Check this presentation from General Dynamics out. (Beware techy link, pdf.) If you stagger on past the first few radiohead slides, you'll get to their aspirations for cognitive radio. Apparently it's going to be waveform aware, spectrally aware, network aware, geographically aware, locally-available services aware, user needs and requirements aware, speaks and understands your language, situation aware, security policy aware, adapts and optimises, learns usage patterns of user, applies model-based reasoning to current context. And they're also talking biometrics, not to mention GPS. Now, think of that list of requirements - what other military field system would be capable of all those things? (Forget the first three radiohead/bellhead items)
Yes - that's right. That would be a competent soldier adequately prepared for their task.
Now, have a read of this from Jamie Kenny on the death of an American blogger in Basra. Please forgive me my slightly callous comment, by the way. But, having read his blog, which seems to be marked by a strange Anglophobia and a serious failure to understand that political Shi'ism is very much a locally-available service in Basra, I can't help wondering what he expected to find. And - going back to the language discussion between Irregular Analyses, Intel Dump and me - is the sheer astonishing scope of what is being proposed for mere radios an admission of defeat?