Spectrum trading—the new allocation sensation

Telecom Asia, Nov, 2004 by John C. Tanner

The Telecom Regulatory Authority of India raised some hackles in late September when it proposed to fix a CDMA : spectrum shortage by allocating new spectrum in the 1850- to 1910-MHz band paired with 1930- to 1990-MHz for 2G services--which, unfortunately, overlaps with the international global spectrum band (1885-2025 MHz and 2110-2200 MHz).

GSM operators in India were livid. So was the GSM Association, whose chairman Craig Ehrlich blasted the proposal at 3GSM Asia in Singapore, saying the proposal would deny the Indian people the "benefit of unrestricted access to global roaming in the 3G world."

Technically he's right, if a bit melodramatic. Technological isolation on the mobility front hasn't hurt Korea or Japan in terms of 3G development (with the caveat that those are developed nations; India is not). The roaming issue is also debatable, at least in the near term, as India is arguably years away from ever seeing a UMTS launch.

Complicating things is the fact that wireless broadband isn't just for cellular anymore. Everything from Bluetooth to WiFi and WiMAX is becoming part of the wireless broadband landscape alongside UMTS and EVDO. India's IT and communications minister, Dayanidhi Maran, knows this, and said in May that he wants to see India leapfrog 3G to "4G" because 3G "has not proved cost-effective."

The CDMA Development Group is enjoying all this--partly because they've pushed for the allocation plan, reportedly to give Indian CDMA operators improved roaming options, and partly because it highlights GSM's difficulties in maintaining economies of scale when it's forced to work across different frequencies. The WRC 2000 conference in Istanbul identified other frequency bands for IMT-2000, and UMTS is being outfitted to operate at 850 MHz and 1850/1900 MHz to accommodate GSM operators in the US. But this means market fragmentation for GSM equipment manufacturers, which means lower volumes and thus higher costs for operators.

No wonder the GSMA is concerned.

But even discounting GSM's scale issues, the TRAI proposal is pretty bad policy--unless uneven playing fields and politics is one's idea of good policy.

Trade your spectrum

Interestingly, the India row is the latest example of regulatory decisions on spectrum making waves, from Hong Kong's OFTA proposing to create a fifth 3G license out of expiring 800-MHz licenses, to Japan's telecoms ministry planning to reallocate 800-MHz frequencies to NTT DoCoMo and KDDI.

The common theme here is shifting resources (spectrum) to support advanced mobile services (3G), which is getting more complicated as new technologies emerge requiring new or additional spectrum.

Other new spectrally-efficient technologies like digital broadcasting, cognitive radios and smart antennas can help operators alleviate the scarcity problem. But for many, the bottom line is that the old regulatory model just isn't hacking it. Some in the industry are calling for a new, flexible approach to allocation--namely spectrum trading, a concept under which carriers can buy, sell, lease and trade their spectrum however they see fit as a secondary mechanism for spectrum assignments.

This isn't a new idea. New Zealand has been doing it since 1989. Australia and Singapore also allow it to varying degrees. But the spectrum trading model is generating new interest as efficiency and scarcity issues become more prominent.

It's a tall order, to be sure, and there's no one way to go about it. PA Consulting's Julian Lake pointed out at 3GSM that regulators face a number of challenges and decisions regarding issues such as transparency, whether to allow traded spectrum to be used for different technologies, and making it all gel with existing regulations for things like anti-competitive behavior and fulfillment of license obligations.

There's also the question of how much of a difference spectrum trading might make. Marloes van Caspel of Analysys recently noted that while New Zealand is a prime example of the feasibility of implementing spectrum trading, telecoms players there have hardly ever used it in the past 15 years.

But with the coming wireless broadband wave, who knows? New Zealand's wireless market may turn out to be better prepared for the coming migration to next-gen ubiquitous connectivity than markets still governed by old-school spectrum policies.