Qualcomm

September 14th, 2007

Ministry of Economic Development
Radio Spectrum Policy and Planning
Energy and Communications Branch

Email: radiospectrum@med.govt.nz

Qualcomm Incorporated® would like to thank the Ministry of Economic Development (MED) for the opportunity to provide comments on Radio Frequency Auction: 2.3 and 2.5 GHz Bands.

As the MED may already be aware, Qualcomm is a leader and innovator in the development of digital wireless technologies including those based on Code Division Multiple Access (CDMA) and Orthogonal Frequency Division Multiplexing (OFDM). These solutions are available today for a number of communications applications, including mobile cellular, fixed wireless access, broadband wireless access, trunking and satellite services. Qualcomm is interested in the success of all the air interfaces that comprise the International Telecommunication Union’s (ITU) IMT-2000 recommendations that use CDMA technologies, including CDMA Multi-Carrier (CDMA2000), CDMA Direct Spread (WCDMA/HSPA) and CDMA Time Division Duplex (UTRA TDD & TD-SCDMA), which share a common technology base.

Qualcomm has also developed other technology solutions for wireless service providers such as MediaFLO™ ¹ , an end-to-end solution that enables broadcasting of high-quality video streams, audio channels, as well as other multimedia applications (video clips, IP datacasting applications, etc.) to mobile handsets. FLO (Forward-Link-Only) technology, a key component of the MediaFLO system, is a new mobile broadcast air interface – based on OFDM modulation – that is designed and optimized to increase capacity and coverage while reducing the cost of multimedia content delivery to mobile handsets.

Qualcomm supports the New Zealand Government’s efforts to facilitate the delivery of faster and cheaper broadband and to create a digital future for all New Zealanders using the power of ICT. QC also thanks the MED for its attempt to take into account the views of interested parties and propose band plan options to balance these interests and meet the needs of the New Zealand consumer. Qualcomm strongly supports the MED’s continued market-based approach to releasing spectrum and believes that auctions have proven to be an effective method for assigning spectrum in a fair and efficient manner.

Question (1): Considerations Relating to Lot Design

Do you agree with the considerations discussed in this chapter concerning lot design? Please provide any comments under the following headings:

a. Expected technical developments and likely demand.

3G technologies are currently providing advanced broadband data and voice services to more than 488 million subscribers worldwide.² As of September 4th, 2007, there were over 400 commercial 3G operators in 135 countries worldwide.³ HSDPA, part of 3GPP/UTRAN-FDD Release 5 WCDMA specifications, is a software based enhancement that significantly boosts the air interface capacity of WCDMA networks and delivers a 5-10 fold increase in downlink data speeds of standard WCDMA networks. As of September 13th 2007, there were 138 commercial HSDPA mobile broadband networks in operation in 66 countries and several more in various stages of deployment. ⁴ Alternatively, CDMA2000 1xEV-DO, part of 3GPP2 specifications, is a data-optimized evolution of the CDMA2000 standard. There are over 75 CDMA2000 1xEV-DO networks in commercial operation today including Telecom New Zealand’s EV-DO Revision A network.5

Importantly, laptops embedded with high-speed HSDPA and EV-DO capability as well as HSDPA and EV-DO multi-band data cards / USB modems are available in the market today and are able to meet the demand for portable (or nomadic) BWA services.⁶ Some studies forecast that the current 3G technologies and their enhancements will comprise 95 percent of the world’s mobile Broadband Wireless Access (BWA) subscriptions by the year 2010 whereas IEEE 802.16e, for example, will comprise just 2.5 percent.⁷ Importantly, both 3GPP and 3GPP2 are planning continuous enhancements for HSPA and EV-DO to deliver higher bandwidth and better economics. As 3G technologies continue to evolve (e.g., HSPA+, Long Term Evolution (LTE), Ultra Mobile Broadband (UMB)), they will require additional spectrum and larger channel bandwidths in order to provide additional capacity in the network. Access to the relatively wide bandwidth available from 2.5 – 2.69 GHz will be necessary in order to benefit from these evolved technological capabilities and to deploy richer applications and services.

To date, a handful of countries have recently released spectrum in the 2.3 – 2.4 GHz and/or 2.5 – 2.69 GHz bands. However, there remains uncertainty within the wireless industry on the near-term market readiness of infrastructure and user equipment that would be able to operate in these bands. There have been a limited number of commercial deployments and these deployments have not proven as successful as expected.

b. Size and location of the managed spectrum park or parks, including:

- whether two MSPs should be provided for in the 2.5 GHz band in order to allow for FDD

(including CMAR) uses.

Qualcomm has some reservations regarding the MSP approach and has provided its comments in response to Question 8 below. Given the difficulties already encountered in trying to implement a MSP type regulatory approach in other countries, we suggest the use of MSP’s be minimized and concentrated in the 2.3 GHz band. The 3G solutions planned for the 2.5 GHz band are tested and proven. It would be a better use of scarce spectrum resources to apply the MSP approach in the 2.3 GHz band where other technological solutions under consideration are more suitable with a MSP approach. To the extent that MSPs are provided for in the 2.5 GHz band, they should be located within the 2570 – 2620 MHz ‘center gap’.

Based on our understanding of CMAR services, this is a feeder technology, as opposed to a direct radio link to the customer/premises and it is used in remote areas of New Zealand. We believe CMAR services can be provided in spectrum assigned via management rights. The MED does not need to plan two MSPs in the 2.5 GHz band to accommodate CMAR FDD services. The spectrum requirements for CMAR could also be accommodated in the 2.3 GHz band. This position is supported by TNZ’s submission in response to the first consultation paper where it “Supports MSP with a minimum of 24 MHz, to use 2.3 band for servicing remote
rural communities using CMAR (unless MED can identify an alternative band for CMAR).”⁸

Limiting the MSP spectrum to the 2570 – 2620 MHz band and to within the 2.3 GHz band would also be consistent with the key decisions of Cabinet:

“provide for a managed spectrum park of at least 30 MHz (and potentially up to or exceeding 50 MHz) in the 2.3 GHz and/or 2.5 GHz bands to enable easy access for
regional and rural service providers, including Maori service providers;”

c. Acquisition limits.

Qualcomm supports the MED’s proposal to impose acquisition limits as a means to facilitate competition.

d. Technical issues such as:

- accommodating TDD and FDD technologies;

Qualcomm believes there are pros and cons to both FDD and TDD modes and the prospective spectrum licensee is in the best position to determine its business case and make technology decisions based on that business case. Within 3GPP, TDD standardization for WCDMA/HSPA is complete. There are two versions, TD-CDMA and TD-SCDMA, and both of these standards are published. TDD mode in LTE is one of the main features currently under discussion. Within 3GPP2, the UMB-TDD standard is expected to be finalized by the end of this year or early next year. In addition, FDD mode for IEEE 802.16d and 802.16e has been standardized.

Irrespective of technology, FDD has a link budget advantage over TDD, particularly in a coverage limited scenario (i.e., sub-urban and rural areas), and link budget impacts cell radius. Therefore, in a coverage limited scenario, a TDD deployment will require more base stations than the same technology deployed in FDD (See Attachment 2 for more detailed analysis). Given New Zealand’s demography, we believe the coverage advantages of FDD over TDD are particularly relevant.

Also irrespective of what technologies or services may be deployed, a common and harmonized band plan facilitates economies of scale which in turn brings benefits to consumers. For the 2.5 GHz band, there has been a great deal of work in various fora on developing and harmonizing a common band plan. For example, Europe (i.e., European Conference of Postal and Telecommunications Administrations) has decided upon a band plan whereby FDD operations would use the upper and lower 70 MHz of the 2.5 – 2.69 GHz band and the center 50 MHz gap from 2.57 – 2.62 GHz would be used for either TDD operations or external FDD downlinks.⁹ Also, the ITU has approved a recommendation which includes three different 2.5 GHz band channeling options, two of which are consistent with the European plan.¹⁰ Qualcomm believes there are major advantages to the European plan (i.e., 70 MHz FDD / 50 MHz TDD or external FDD downlink / 70 MHz FDD) in that it provides separate band segments for FDD and TDD operations, maintains the 120 MHz duplex separation specified by 3GPP and 3GPP2 (see Attachment 1), and is consistent with the plan expected to be used by many other countries around the world. Deviating from a common band plan would impact New Zealand consumers due to the need for New Zealand-specific handsets that would be difficult, if not impossible, to be used for roaming with other countries.

- guard-bands;

Qualcomm supports the MED’s proposal for management right holders to take responsibility for managing interference issues, and acquire sufficient spectrum to be able to do so. That said, Qualcomm does believe this is an area where further study and vetting is needed by the technical community. Some co-existence studies between 3G IMT-2000 networks and fixed deployments of IEEE 802.16 TDD in the 2.5 – 2.69 GHz band have been conducted in various fora, including ITU-R Working Party 8F, ETSI BRAN, and Ofcom UK. However, the conclusions of these studies have recently been called into question due to updated information on the IEEE 802.16e TDD emission masks that was submitted to ITU-R Working Party 8F by the WiMAX Forum on June 15th, 2007.¹¹ A key parameter in the co-existence studies is the value of the Adjacent Channel Leakage Ratio (ACLR) which is defined as the ratio of the transmitted power to the power measured in the adjacent radio frequency (RF) channel at the output of a receiver filter. This parameter is determined from the emission mask but the emission mask information provided by the WiMAX Forum does not yield the ACLR values that the WiMAX Forum have used as the basis of their coexistence studies. Queries raised at the recent ITU-R Working Party 8A meeting in June and the WP8F Special meeting in Seoul have not resolved this problem. The statements made by China and Germany at the conclusion of the Seoul WP8F meeting highlighted their concerns about this discrepancy.

- other.

e. Preferences regarding auction type.

Qualcomm urges the MED to use the same auction method for the 2.3 GHz and 2.5 GHz bands that MED has used for other prior spectrum auctions. In general we strongly believe that 2.3 GHz and 2.5 GHz management rights holders should be subject to the same regulatory regime and licensing requirements as that required of management rights holders in the 2.1 GHz, 1800 MHz, 900 MHz and 850 MHz frequency bands.

f. Other.

Question (2): Proposed Lot Design

Do you prefer Option A, B or C for the lot design? Please explain why. If you prefer Option C is your preference for C(i) or C(ii)? Why? If there is an alternative option you prefer please specify it and explain why it would be preferable.

Option A does not support the 120 MHz duplex separation required by 3GPP, 3GPP2 and two of the three frequency arrangements recommended by ITU-R Recommendation M.1036 for the 2.5 – 2.69 GHz frequency band. We agree with the MED that “this option provides little opportunity for the market to deploy FDD services.” Qualcomm, therefore, urges the MED to eliminate this Option.

Options B and C could potentially support the 120 MHz duplex separation required by 3GPP, 3GPP2 and two of the three ITU-R Recommendation M.1036 for the 2.5 – 2.69 GHz frequency arrangements. However, Options B and C have the potential to provide a maximum of 2 x 50 MHz (total) of spectrum for 3G FDD operations. Given the existing 3G subscriber base and projected uptake, it is likely this would not be sufficient to support the future expansion needs of 3G network deployment.

In general, we believe that all of the options which entail a flexible FDD/TDD block assignment scheme would not provide the expected benefits to either paired or un-paired operations. This is because it will be extremely difficult for an FDD handset to operate in differing band plans with one Transmit/Receive (Tx/Rx) chain while protecting the unpaired spectrum from Tx interference. For example, an FDD terminal operating in accordance with the European agreed ECC band plan¹² could have difficulty meeting the protection requirements of a TDD handset operating in accordance with a different band plan. The FDD terminal would have to use a set of duplexer filters different from the set used for the ECC band plan in order to protect the unpaired spectrum. Requirements on the TDD terminal Tx would also be needed to meet the FDD handset Rx band requirements. These requirements could be difficult to meet. Terminals with roaming capabilities between countries that have adopted differing band plans will require more complex implementation likely resulting in higher cost, higher power consumption, and shorter distance coverage due to higher filter insertion loss. This will impact New Zealand consumers as it may lead to New Zealand-specific handsets which will be difficult to be used to roam in other countries. Irrespective of what technologies or services may be deployed, a common and harmonized band plan facilitates economies of scale, which in turn brings benefits to consumers.

Of the Options presented, Qualcomm believes Option C(i) is the best for achieving the Government and MED’s objectives. Qualcomm’s views on MSPs and the amount of spectrum to be included in MSP are detailed in response to Questions 1 and 8. In general, we believe the amount of MSP spectrum should be reduced and concentrated in the 2.3 GHz band. To the extent that MSPs are provided for in the 2.5 GHz band, they should be located within the 2570 – 2620 MHz ‘center gap’. With respect to protection of passive services (i.e., Radio Astronomy and Space Research) above 2690 MHz, Qualcomm believes that it is possible, and a more efficient use of radiospectrum, to reduce the amount of guardband between passive services allocated on a primary basis above 2690 MHz and management rights spectrum below 2690 MHz.

Alternately, we therefore suggest a variant of Option C for the 2.5 GHz band that is consistent with our findings. The proposed band plan is as follows:
2500 – 2520 MHz (20 MHz)
2520 – 2540 MHz (20 MHz)
2540 – 2560 MHz (20 MHz)
2560 – 2590 MHz (30 MHz) MSP
2590 – 2620 MHz (30 MHz)
2620 – 2640 MHz (20 MHz)
2640 – 2660 MHz (20 MHz)
2660 – 2680 MHz (20 MHz)
2680 – 2690 MHz (10 MHz) MSP

The advantages to this proposal are:

• It is technology neutral,
• It maintains the 120 MHz duplex separation specified by 3GPP and 3GPP2 and the ITU-R,
• It provides two MSPs to transition existing users, accommodate local/regional networks and potentially CMAR FDD services,
• It has the potential to be consistent with the band plan expected to be used by many other countries around the world,
  

Because of the international roaming requirements and the need for mass market user equipment, we strongly believe MED should keep open the option for 3G to access the 2.5 GHz band.

Question (3): Māori Interests

What provision should be made regarding allocation of spectrum in the 2.3 GHz and 2.5 GHz bands for use by Māori (i.e. Māori service providers)? What provision should be made regarding allocation of spectrum in the 2.3 GHz and 2.5 GHz bands for use for Māori (i.e. Māori as service consumers, or in the interests of Māori language and culture)? What terms and conditions should apply to this spectrum?

Qualcomm generally supports preserving the interests of Maori language and culture and the need for special provisions to meet these objectives.

Question (4): Eligibility to Bid

Do you agree that no restrictions should be placed on the eligibility of parties to bid for lots?

We believe minimum pre-qualification requirements would be needed to ensure that prospective bidders have the technical and financial capability to build and operate a network. These eligibility requirements can be borrowed from that required at prior Radio Frequency Auctions in New Zealand. In general, we believe it is critical for the MED to ensure consistent licensing and regulatory frameworks for the various frequency bands that have been auctioned in New Zealand.

Question (5): Duration of Acquisition Limits

Do you consider the expiry of acquisition limits 1 year before the (December) use or lose date to be satisfactory? Do you prefer a different acquisition limit date? Please explain reasons for your view? Do you have any other comments on the time for acquisition limits to expire?

No comment

Question (6): Use or Lose Provisions

Do you prefer a date of December 2012, 2014, or 2016, for applying the use or lose test? If not what alternative date would you prefer to implement the Cabinet decisions on use or lose? Please explain. Do you prefer Option A or Option B or some other option regarding the test for ‘use’? If you do not agree with the proposals, what ‘use or lose’ provisions do you propose? Should financial consequences result from not using rights, either in addition or as an alternative to loss? How should the amount of any financial consequence be calculated?

Qualcomm generally supports use of lose provisions as a means to ensure efficient use of scare and valuable spectrum resources. We believe that 2.3 GHz and 2.5 GHz licensees should be subject to license requirements and obligations, including use or lose provisions, similar to that required of 3G licensees in the 800 MHz, 900MHz, 1.8 GHz and 2.1 GHz bands. We also believe that clear definition of the use or lose provisions should be specified well in advance of the auction. However, the use or lose dates proposed by the MED would give licensees a considerable amount of time to build out their network – possibly five, seven or nine years from the date of the auction.

While we agree there is uncertainty within the industry on the near-term market readiness of infrastructure and user equipment that can operate in the 2.3 and 2.5 GHz bands, these proposals to allow an extended period of time to build out the network are not in the best interests of the New Zealand consumers, are not consistent with the Government’s broadband objectives, and are excessive. Qualcomm believes 2010 or 2011 would be a more reasonable date for applying the use or lose test. By comparison, the Hong Kong OFTA recently proposed a two year time period for the 2.3 GHz band in Hong Kong. And, Singapore, back in 2004, proposed 18 months for the 2.5 GHz band and three years for the 2.3 GHz band. Use or lose provisions need to be put into place to prevent against spectrum warehousing and speculation, and ensure provision of broadband services to New Zealand consumers as envisioned by the Government.

Question (7): Settlement

Do you agree with the proposed settlement terms (30 days following completion of the auction)? If not, what other factors do you see as relevant?

Qualcomm supports the MED’s proposal.

Question (8): Managed Spectrum Parks: Proposed Allocation Procedures and Usage Rules

1. Do you agree with the suggested eligibility criteria for access to a MSP?

2. Do you agree with the suggested core technical and usage requirements

3. Of the three options outlined above, which is your preferred method for implementing a MSP? Why?

4. Are there better alternatives or variations on these implementation options?

5. What incentives for gaming arise under the various options, and what measures, if any, could be taken to minimise such incentives?

6. What fees or resource charges should be levied (particularly for your preferred option)? 

With regard to the Managed Spectrum Parks proposal, Qualcomm is generally of the view that spectrum licensees should have the exclusive right to transmit in the spectrum for which they hold licenses. Qualcomm takes this position not out of any ill-will toward unlicensed technologies. Qualcomm is developing 802.11n unlicensed technology, and Qualcomm believes that unlicensed technologies have an important role to play in local area, short range communications—such as inside buildings, in homes, and on campuses. Qualcomm also believes, however, that unlicensed technologies are simply not suited to wide area, longer range communications. Only licensed services operating on spectrum dedicated exclusively to licensed services can provide coverage over wide areas.

Qualcomm does not believe that a MSP licensee could start a wide area wireless business on the basis of a non-exclusive license. This is because no operator will spend the large amounts of money needed to build out a network without clear and undeniable protection from interference. Without close oversight, it is possible that the day after one licensee has turned on its network, another licensee for the same spectrum and geographic area could turn on its network and cause interference. For these reasons, Qualcomm does not believe the MSP is suitable for frequency bands where 3G or IMT-2000 may be deployed.

Qualcomm would also like to point out that this new regulatory framework has similar characteristics to the non-exclusive licensing provisions outlined in the Federal Communications Commission (FCC) Report & Order 05-56. Importantly, we would like to emphasize that the FCC’s decision in this proceeding has come under significant criticism from a number of companies in the wireless industry. The FCC proposed to rely on dynamic frequency selection and contention-based protocols yet there has not been any showing that these technologies will actually work in the 3650 – 3700 MHz band as envisioned. While adopted by the FCC for the 3650 – 3700 MHz band, this non-exclusive licensing concept has not yet been implemented because the decision is subject to a large number of petitions for reconsideration.

Question (9): Transition Plan for the 2.5 GHz Band

Do you have any comments on the proposed transition plan for existing licences in the 2.5 GHz band?

Qualcomm agrees there needs to be a clear transition plan for existing licenses in the 2.5 GHz bands. The MED is proposing to shift many of these existing users to a 2.5 GHz MSP for a transitional period. It may be too optimistic to accommodate all of these diverse technologies, including broadband, within a MSP and provide sufficient protection from interference. Managed park spectrum rights may be more suitable for some of these existing licenses. MED further proposes “to create a license within the auctioned management rights to allow existing camera mounted equipment to transmit at low power levels until December 2010.” We believe further technical study is needed to assess the potential for interference from the camera mounted equipment into the services of the management rights holders. It is understood that the camera mounted equipment would not be entitled to interference protection from management rights holders. On a related matter, Qualcomm has conducted its own extensive testing to determine the impact of allowing various unlicensed operations on licensed spectrum, and the tests have uniformly supported the same conclusion: unlicensed devices cannot operate in licensed spectrum without causing harmful interference to the licensed services.¹³

Conclusion

Qualcomm appreciates the opportunity to provide these comments to the MED. Qualcomm supports the New Zealand Government’s efforts to facilitate the delivery of faster and cheaper broadband and to create a digital future for all New Zealanders using the power of ICT. Should you have any questions or comments on this submission, please do not hesitate to contact me.

Respectfully,
Julie Garcia Welch
Director, Government Affairs
Southeast Asia & Pacific
cc: Mr. Robert Hart, Senior Director, Business Development, Qualcomm Australia and New Zealand

Footnotes

¹ http://www.Qualcomm.com/mediaflo/index.shtml.

² www.wirelessintelligence.com as of July 2007.

³ see notes 4 and 5.

⁴ GSMA Association, www.gsaworld.com as of Sept 2007.

⁵ www.3Gtoday.com as of Sept 2007.

⁶ See http://hspa.gsmworld.com/devices/default.asp and www.3Gtoday.com for updated figures on HSPA and EV-DO enabled embedded notebook computers, handsets, data cards, USB modems, equipment suppliers etc.

⁷ Strategy Analytics “Beyond 3G: Looking for True Mobile Broadband,” November 2006.

⁸ See Summary of Submissions, http://www.rsm.govt.nz/cms/pdflibrary/2.3_ghz_spectrum_auction_9_summary_submissions.pdf. As of 10 September 2004, the actual submissions have not been posted on the MED website.

⁹ ECC decisions are mandatory for member states.

¹⁰ ITU-R Recommendation M.1036, “Frequency arrangements for implementation of the terrestrial component of International Mobile Telecommunications-2000 (IMT-2000) in the bands 806-960 MHz, 1 710-2 025 MHz, 2 110-2 200 MHz and 2 500-2 690 MHz.”

¹¹ Documents 8F/1329 and 8F/1330 found on the ITU-R WP8F website at http://www.itu.int/md/R03-WP8F-C/en.

¹² See “ECC Decision of 18 March 2005 on harmonised utilisation of spectrum for IMT-2000/UMTS systems operating within the band 2500 – 2690 MHz.” This decision shows the frequency plan agreed upon within the European Conference of Postal and Telecommunications Administrations (CEPT) Electronic Communications Committee (ECC) to be 70 MHz FDD / 50 MHz TDD / 70 MHz FDD.

¹³ QUACOMM filed its test results with the US Federal Communications Commission in the proceeding over ultra wideband, Docket No. 98-153. See Ex Parte Filings dated March 5, 2001; January 11, 2002; February 1, 2002, Docket No. 98-153. See also Petition for Reconsideration filed June 17, 2002.

Attachment 1

Excerpts from 3GPP and 3GPP2 Specifications for the 2.5 – 2.69 GHz Band

3GPP TS 25.101 V7.6.0 (2006-12)

5.2 Frequency bands

a) UTRA/FDD is designed to operate in the following paired bands:

Table 5.0: UTRA FDD (WCDMA/HSPA) frequency bands

b) Deployment in other frequency bands is not precluded

5.3 TX-RX frequency separation

a) UTRA/FDD is designed to operate with the following TX-RX frequency separation

Table 5.0A: TX-RX frequency separation

b) UTRA/FDD can support both fixed and variable transmit to receive frequency separation.

c) The use of other transmit to receive frequency separations in existing or other frequency bands shall not be precluded.

3GPP2 CS0057-B

CS0057-B defines 120MHz duplex separation for CDMA2000 FDD in Band Class 13.

Table 2.1.14-1. Band Class 13 Block Frequency Correspondence

Attachment 2