Nokia has taken Apple to court over the iPhone's use of 10 core patents covering basic wireless technology. Nokia says Apple has been using the technologies without paying for them: it is undoubtedly true that Apple has been using them, as you cannot make a GSM phone without these. But the question of who pays whom and how much is very involved. In fact, it defines the shape of the industry.
While patent disputes have been a part of the business of radio since Marconi and his infamous Four Sevens patent of 1900, GSM and its progenitors have brought new levels of complexity and confusion to what is intended to be a mutually advantageous system of law.
There is an enormous irony at the heart of intellectual property. By giving to ideas some of the legal protection that is given to physical things, society recognises that inventiveness deserves reward, and rewards require ownership. At the same time, IP law is constructed so secrecy is not required, recognising that ideas have most worth when they are shared.
When dealing with simple things such as a better paperclip or a novel, the commercialisation of the rights is a straightforward bargain between the creator and the consumer — albeit with plenty of opportunities for middle men. But when the area of innovation is a mature, complex and hugely interconnected set of ideas, the power struggles can be gargantuan.
So it is with GSM, the global digital mobile telephone standard. We are blinded to the complexity of mobile phones by their ubiquity, low cost and simple appearance. You can buy a handset for £20 and, at a press of a button or two, it makes a phone call.
But what is actually happening during that phone call fills tens of thousands of pages of specification: frequency agile UHF radios constantly scan for available frequencies and negotiate connections; time division multiplexing weaves multiple calls together on a single channel; voice is encoded, encrypted, modulated, then demodulated, decrypted and turned back into audio, all in a channel barely wide enough to hold a 1930s Morse code signal.
That is before you consider the base stations linked together by atomic clocks to nanosecond accuracy, co-operating with themselves and their competition to create networks that share space and time without interfering with each other.
GSM is the culmination of decades of invention by hundreds of companies. It is also a market supporting four billion handsets and climbing, a global network of consumers who are avid for the next big thing and the next world-changing service.
Any handset maker who wants to join in has to license those thousands of pages of intellectual property, owned by immensely rich companies who depend utterly for those riches on making the most of their inventions.
There are ways and means for this to happen, if not simply, then manageably. The standards body gets the agreement of the rights owners that, in exchange for being part of the standard, they put their IP into a pool which will be marketed by an independent agency. Anyone wanting to use that standard buys the package of rights from the agency, which distributes the proceeds back to the rights owners.
This did not happen with GSM. Instead, the major IP owners — including Qualcomm, Nokia, Ericsson and the less well-known Interdigital — have cross-licensing agreements that let them use each other's IP.
Those outside this group must make their own arrangements with each individual party, which can include cross-licensing or a straightforward deal. But as the ETSI industry database of IP lists 4,455 claims of IP for GSM alone, that is a task of horrendous complexity — one where the big players with cross-licensing agreements have an enormous advantage.
They are not slow to use that advantage. It is very difficult to say how much of the cost of a GSM phone is IP, because of the extreme variability in the deals and because many agreements will be predicated on the end price of the handset. However, a good rule of thumb for a mid-range feature phone is that it costs the manufacturer around 35 percent of the manufactured cost in rights alone. Sometimes those costs are absorbed by the chipmakers, who put the costs on top of the price of the silicon; sometimes, however, the handset maker has to pay again.
The same is true for 3G, where, despite many attempts to create an agency-moderated pool of IP, some IP owners have found it to their advantage to destroy such attempts and keep control of their deals. And when it is impossible to bypass the patents they hold, it is impossible to break their stranglehold except through protracted, expensive and dangerous legal battles on their own turf.
It is against this backdrop that the Apple-Nokia dispute will play out. Apple is not a member of the circle of IP owners for GSM, 3G, et al, and so must license things separately. However, what an IP owner considers reasonable terms often sits badly with the supplicant, and Apple is not known for meekness or passivity in doing deals.
While little is known of the details of the dispute, it is fair to say Apple is not giving Nokia what Nokia expects, and while Nokia's expectations are normally met by other companies, other companies do not have upwards of $30bn (£18bn) in the bank and Steve Jobs at the helm. Also, Apple can be aggressive with IP — remember it went ahead and used the iPhone name while Cisco still claimed it for its own.
Whether Apple is being unreasonable in refusing to pay what Nokia expects, or whether Nokia is being unreasonable in its demands, will be for the court to decide — assuming that the case is not settled quietly before it gets there, which remains the most likely outcome.
There are plenty of other questions that need attention. Is Apple in similar negotiations with the other GSM/3G rights owners, and if so what state are they in? If Apple loses, will the others take advantage? Does Apple have handset IP that Nokia wants, and is this a sticking point? These are very high stakes in a game with enormous implications: it is the irony of IP law that such matters are invariably kept as secret as possible.