Ericsson Submission to NZ Ministry for Economic Development © Ericsson AB 2007 Steve Inglis, Country Manager (Acting), Ericsson New Zealand 2 (13)
The Discussion Paper was released following the announcement by the Minister, the Honourable David Cunliffe, of an intention to auction the 2.3GHz and 2.5GHz bands “to make spectrum available for national and regional wireless services including WiMax”. The stated objective is to use wireless technologies to promote competition in order to achieve faster and cheaper broadband services.
The paper then proceeds to discuss a range of detailed topics (lot design, spectrum parks, etc.), but does not consider what general approach best serves New Zealand.
The assumptions we find inherent in the Discussion Paper are:
- New Zealand lacks sufficient coverage and competition for wireless services as a consequence of insufficient licensed spectrum
- WiMax is a technology that warrants Government assistance
- New Zealand can or should develop Spectrum Plans independent of decisions made by other authorities (especially regional groups such as CEPT (Europe), APT (Asia-Pac), ASMG (Arab States), ATU (Africa) and CITEL (Americas))
- The auctioning of RF bands in this way will not disadvantage New Zealand
These assumptions are unfortunately not justified in the Paper. Ericsson believes that these are extremely important assumptions that deserve to be clearly identified and analysed in light of current and future, local and global trends.
The Internet has had a significant impact on society, and today there are over one billion daily users worldwide. Research has shown that ready access to communications including broadband services has a direct impact on Gross National Product (GNP). Broadband services enhance our lives in many ways, including education, industry, research and development, healthcare, telecommuting and bridging the digital divide.
There are now more than 3 billion mobile subscribers globally, and forecasts indicate that mobile traffic generated by data services will exceed that of mobile voice traffic by 2010. By the year 2012, there are expected to be 1.8 billion broadband subscribers, with around two-thirds of these being mobile broadband subscribers.
Significant growth is forecast for mobile broadband services in the future, including TV and streaming video, downloads, laptops, and portable communication devices, to name a few (Figure 1).
Figure 1 - Information services, as well as voice, will drive demand for mobile broadband (Source: Ericsson)
According to international studies, the total amount of traffic per user is estimated to grow to 50 times that of today in many markets. This presents broadband service providers with both a challenge to provide access networks to support these massive traffic volumes, and also an opportunity to invest in a technology evolution that will deliver cost-effective broadband today and into the future.
Figure 2 – Explosion in mobile broadband subscriptions is forecast
As the date rates capable in mobile telecommunications systems is largely determined by the available bandwidth and the way technology uses that spectrum it is crucial that spectrum allocations are globally harmonized and provisions made for the forecast capacity growth and evolution of the respective technologies. In the initial deployments in many countries today, mobile broadband service offerings have been deployed using a single carrier frequency (eg with a 5MHz bandwidth for 3G/WCDMA systems) for both voice and mobile broadband. As mobile broadband traffic increases, operators will require additional carrier frequencies (blocks of 5-20MHz) and new spectrum bands to meet this demand (Figure 2).
With more than 80 per cent of the global mobile market, the GSM family has proven itself to be the dominant mobile technology. The rationale for operators migrating from cdma2000 to GSM/WCDMA has been mainly economic - it is the most competitive way to deliver voice and data services to address the future growth in mobile subscribers with cost-effective handsets.
Building on GSM/WCDMA, High Speed Packet Access (HSPA) is the evolution towards higher mobile broadband data rate in both the downlink and the uplink, as shown in Figure 3. HSPA enabled devices are widely available. Handheld terminals, PC cards, Fixed Wireless Terminals, mobile broadband modules for installation in PC notebooks and even a camera with inbuilt HSPA support have already been launched. This gives HSPA a unique economy-of-scale advantage that will drive network roll-out and mass market services and acceptance, ensured by device offerings at attractive prices. Evolved HSPA introduces simpler IP centric architecture for the mobile network and even more enhanced radio data rates.
Furthermore, Long Term Evolution (LTE) is part of the GSM/WCDMA/HSPA evolutionary path beyond 3G. LTE is an entirely new radio platform technology with standards currently under development by the Third Generation Partnership Project (3GPP) standardization body, with initial deployments planned in 2009/2010. LTE uses Orthogonal Frequency Division Multiple Access (OFDMA) on the downlink, which is well suited to achieve high peak data rates in high spectrum bandwidth (downlink peak data rates up to 100 Mbps with 20 MHz bandwidth and uplink peak data rates up to 50 Mbps with 20 MHz bandwidth).
Figure 3 – 3G/WCDMA Evolution Path
According to analysts Strategy Analytics, it is expected that HSPA technology will represent over 70 per cent of the mobile broadband market by 2010, as shown in Figure 4. This is achievable because HSPA is the evolution of an existing, mass-market technology (WCDMA), with a huge number of deployments already underway.
HSDPA (High Speed Downlink Packet Access) has been enabled on almost 75 per cent of all 3G/WCDMA networks globally. As published by the GSM Association (GSMA) for July 2007, 128 HSDPA networks have been commercially launched in more than 63 countries and a further 50 HSDPA networks are currently being deployed.
Figure 4 – HSPA will drive mobile broadband, representing over 70 per cent of market share by 2010
Currently, the only globally harmonised spectrum available for IMT-2000 technologies (such as 3G/HSPA) until at least 2015 is the 2.5GHz band.
It is therefore essential that this spectrum be preserved for meeting the demand for mainstream mobile broadband services and not unnecessarily fragmented or allocated to niche radio technologies.
WiMax is a promising niche technology that is now being deployed in unlicensed bands of spectrum (most notably the use of 5.8GHz in Australia’s OpEl network). While there are no globally harmonized standards for licensed frequencies for WiMax today, this is not an impediment to rollout should proponents wish to deploy the technology today in New Zealand.
However, the incompatibility between the TDD approach used by WiMax, and the paired-frequency FDD approach used by the mainstream mobile standards (in particular 3G/WCDMA) means that assigning WiMax frequencies within a band is likely to prohibit the use of mainstream mobile technologies, and so fragment the respective spectrum band. This will limit the opportunity of the mainstream technologies to evolve to their fastest speeds and so offer cost effective mobile broadband solutions for New Zealand.
According to the International Telecommunication Union, the total spectrum needed for both existing and future mobile cellular systems (including pre-IMT-2000, IMT-2000 and its enhancements, as well as IMT-Advanced) is between 1280 and 1720 MHz by the year 2020.
This implies that an additional 700-1200MHz of spectrum will be required by 2020. Multiple, competitive deployments of mobile broadband technologies in each country will increase spectrum requirements even further.
New Zealand has over recent years enjoyed more coverage by wireless broadband from more operators than most developed countries. Furthermore the introduction of wireless broadband by the mobile operators using the HSPA (high speed packet access) standards in mainstream 3G/WCDMA will add to the choices already available to New Zealanders.
If New Zealand were to adopt spectrum plans inconsistent with the globally harmonized plans, New Zealanders will suffer from:
- poor availability and high cost of network equipment and handsets
- the inability of users to roam (use the networks outside NZ)
- the inability of visitors to roam (gain access to the NZ networks)
- the resulting fragmentation of spectrum would also make it virtually impossible to revert to globally harmonized plans later and create significant economic damage
New Zealand represents a relatively small market for equipment manufacturers, and decisions taken here are unlikely to shape global spectrum harmonization. However, New Zealand can only enjoy the benefits of modern wireless networks by staying well within the mainstream of globally harmonized spectrum plans, in particular WRC-2000, and WRC-2007.
Vodafone NZ already utilises 2.1GHz spectrum for WCDMA/HSPA services, and Telecom NZ has also announced plans to migrate to a WCDMA/HSPA network. In order for Telecom NZ to migrate to WCDMA, additional spectrum will be required during the transition phase, further contributing to the demand for IMT-2000 spectrum. Both operators will require increased spectrum allocations to support the significant forecasted uptake of mobile broadband services in the coming years. The
Ericsson Submission to NZ Ministry for Economic Development © Ericsson AB 2007 Steve Inglis, Country Manager (Acting), Ericsson New Zealand 8 (13)
natural plan then for both operators is to be able to easily acquire IMT-2000 spectrum to facilitate the capacity growth in their networks. We believe it is likely that the two current NZ mobile operators, will require additional spectrum in the 2.5-2.69GHz band as the popularity of mobile broadband services increases over the coming few years.
It is therefore essential that sufficient 2.5GHz spectrum is made available to both Telecom and Vodafone to ensure they are able to meet this forecast increase in mobile broadband demand given their commitment to using 3G/WCDMA/HSPA/LTE technologies.
We recommend refocusing on the longer term objective of securing the economic benefits that flow from full use of modern wireless/mobile communications, and in particular the proven economic benefits of supporting, or at least not hindering, mainstream IMT-2000 technologies and standards such as 3G/WCDMA.
Our recommendations are directed at maximising the economic outcome for New Zealand, in particular we wish to point out that,
- The 2.5GHz band has been accepted as an IMT-2000 Extension band for the growth of IMT (mobile telephony) with HSPA and LTE. This must be sustained in line with International practice.
- The growth in wireless/mobile broadband will make the use of the 2.5GHz for mobile/wireless broadband necessary in a few short years. It is prudent to begin the process of freeing the 2.5GHz band in the way proposed by the Discussion Paper.
- The uses of the 2.3GHz band has not yet been determined, but several regional groups (specifically APT (Asia-Pac) , ATU (Africa), and potentially ASMG (Arab States)) are supporting its use for IMT as well.
- WiMax operators will be best-served by co-ordinating their spectrum requirements with those being used in our nearest neighbour and most common roaming partner, Australia
- Full spectrum opportunities exist for the future demands of the existing 3.6 million cellular subscribers as serviced by Telecom NZ and Vodafone today
Consequently New Zealand should postpone the auction of the 2.3GHz and 2.5 GHz bands until the international position has been resolved at the coming WRC-2007 conference.
New Zealand is best served by careful alignment with the main trends of mainstream telecommunications standards, and specifically with the spectrum plans agreed on at WRC-2000 and WRC-2007. In this way New Zealand consumers will have access to the inexpensive equipment developed for much larger countries, and (via roaming) have the ability to use this equipment in other countries, as well as allowing overseas visitors to use the networks in New Zealand.
Attempts to introduce spectrum plans that are inconsistent with those adopted by the major countries and regional groups would be detrimental to New Zealand.
In recent years we have seen the consequences of the declining global market share for CDMA/EV-DO mobile services. In its most visible form this has lead to fewer handset choices for consumers as vendors preferred to realise scale economies with the dominant GSM/WCDMA standard, driven also by the globally harmonized spectrum for this technology family.
The winning strategy for New Zealand is to follow mainstream international trends as soon as possible after they are standardised.
Since the launch of the first HSDPA network almost two years ago, there has been exponential growth of mobile broadband traffic across WCDMA/HSPA enabled networks globally.
The 2.50-2.69GHz band is of critical importance, as it is the only globally harmonized spectrum that will be available until at least 2015. It was identified as an IMT-2000 during WRC-2000, almost seven years ago.
For wide area coverage of Broadband Wireless Access (BWA) spectrum should be allocated as frequency division duplex (FDD) in all areas where it is possible to pair spectrum. Hence, time division duplex (TDD) should be allocated to spectrum that is not possible to pair.
The primary reason for this recommendation is that cost per square kilometre and cost per added broadband wireless subscriber is kept to a minimum.
Furthermore, Ericsson recommends allocating spectrum according to guidelines and standards identified by ITU as IMT-2000 standards (refer Figure 5), since these standards will safeguard the best possible economies of scale, both in terms of consumer electronics as well as network infrastructure.
Figure 5 – 3GPP IMT supported frequency bands
Hence, the recommended spectrum allocation for the 2.5 to 2.69 GHz band is:
- 2500 to 2570 MHz – uplink for IMT 2000 FDD technology
- 2570 to 2575 MHz – guard band
- 2575 to 2615 MHz – allocated for TDD technology, or additional downlink FDD
- 2615 to 2620 MHz – guard band
- 2620 to 2690 MHz – down link for IMT 2000 FDD technology
The optimum size of each allocated spectrum is a trade off between number of competing operators and the efficiency gains from a less fragmented spectrum. Allocation of 10MHz or greater per operator is recommended for mobile broadband applications.
Ericsson supports the recommendation for clearing out of the 2.5-2.69GHz band for use by IMT-2000 technology.
Use of the frequency band 2.3-2.4 GHz is not globally harmonised, however due to its proximity to the 2.5-2.69MHz band is extremely valuable for IMT-2000 technologies, as they are able to re-use the same device antennas and support similar radio propagation.
Ericsson recommends withholding auction of spectrum in this band until the candidate frequency band is adopted for use by IMT-2000 at WRC-2007, to avoid unnecessary allocation of valuable spectrum.
a) According to the International Telecommunication Union, the total spectrum needed for both existing and future mobile cellular systems (including pre-IMT-2000, IMT-2000 and its enhancements, as well as IMT-Advanced) is between 1280 and 1720 MHz by the year 2020. This implies that an additional 700-1200MHz of spectrum will be required by 2020.
b) No preference for MSP spectrum allocation, only that both FDD and TDD spectrum is allocated in accordance with ITU recommendations to avoid spectrum fragmentation.
c) No comment.
d) 2.5-2.69GHz band spectrum allocation should be allocated in accordance with ITU recommendations to avoid spectrum interference and fragmentation, ensuring that New Zealanders are able to benefit from the global scale of standardised CPE operating in globally harmonised spectrum.
e) No comment.
f) No comment.
None of the proposed options for the 2.5-2.69 GHz band are recommended, as none are in accordance with ITU recommendations.
- Option A supports TDD spectrum allocation only, thereby restricting use to only TDD radio access technologies. This may result in insufficient spectrum being available for mainstream mobile broadband technologies (such as HSPA/LTE) in the future, resulting in the use of “niche” technologies for mobile broadband services in New Zealand.
- Option B provides greater spectrum allocation flexibility than Option A, however may lead to extremely fine-grain fragmentation of spectrum and create unwanted interference. Furthermore, the ITU recommends an allocation of 10MHz or greater for mobile broadband services.
- Option C provides more ideal lot-size allocations than either option A or B, however is not in accordance with ITU recommendations.
We would like to propose a variant of Option C which would provide similar FDD/TDD flexibility but also is aligned with ITU recommendations:
TDD U/D link, including 5MHz guard-bands
(or FDD downlink)
As stated previously, use of the 2.3-2.4 GHz band is not globally harmonised, and we would advise that auction of this spectrum be withheld until the candidate band is adopted for use by IMT-2000 at WRC-2007 and recommendations provided for FDD and/or TDD allocation, to avoid unnecessary allocation of valuable spectrum.
In a free market, any party should be able to bid for spectrum. However, it is essential that the auction process ensures that spectrum is not “fragmented” into small spectrum blocks, which would greatly reduce the efficiency of running mobile broadband services in this valuable spectrum.
Based on Ericsson’s experience, it generally takes around six years from the time a standard is agreed to mass market launch of commercial services. Any special conditions associated with spectrum allocation should reflect this trend to ensure that owners are not disadvantaged.
Refer to Question 5.
Ericsson is a world-leading provider of telecommunications equipment and related services to mobile and fixed network operators globally. Over 1,000 networks in 140 countries utilise our network equipment and 40 percent of all mobile calls are made through our systems. We are one of the few companies worldwide that can offer end-to-end solutions for all major mobile communication standards.
While mobile phone networks were initially developed for telephony, the transition to digital mobile technology has made possible the use of these mobile networks for extensive wireless data connectivity. More recently the transition to GSM and then WCDMA has allowed the introduction of Mobile Internet access. Today modern 3G/WCDMA/HSPA networks are capable of providing 14Mbps connections to subscribers (an experience comparable to a good ADSL fixed broadband access). There is a clear roadmap to enhance these networks to provide 42Mbps (HSPA Evolved) in 2008 and 200Mbps (LTE) in 2009.
In Australia, Ericsson was the prime integrator of Telstra's NextG 3G/WCDMA network. The process of rolling out this network has been seen by the telecommunications industry worldwide as a "world's best practice" in the deployment of major networks.
Ericsson is also a major global supplier of other broadband technologies both wireless (WiMax 802.16d) and wireline (ADSL2+, VDSL2, GPON). In our region, Ericsson is a major supplier of fixed broadband equipment to Australian and New Zealand ISPs. Over 500,000 ADSL lines have been shipped to Australian and New Zealand ISPs.