APG15-3/INF-06

/ ASIA-PACIFIC TELECOMMUNITY
The 3rdMeeting of the APT Conference Preparatory Group for WRC-15 (APG15-3) / Document
APG15-3/INF-06
09 – 13June 2014, Brisbane,Australia / 02 June2014

Telstra Corporation Ltd

Candidate bands for future IMT systems

Introduction

In regard to WRC-15 Agenda item 1.1, the ITU-R Joint Task Group 4-5-6-7 (JTG 4-5-6-7) has considereda range of studies related to possible identification of bandsto meet the future needs of IMT. Significantly, the final meeting of the JTG is scheduled for late-July 2014 in Geneva. In accordance with guidance from the Chairman of the JTG (JTG4567/588), and in accordance with the work plan (Annex 1 to JTG4567/584), no further technical sharing studies are expected to be submitted, and the primary focus at this last JTG meeting will be on summarizing the results of studies for purposes of finalizing the CPM text relating to various nominated bands.

While the list of suitable bands initially proposed for study by Working Party 5D (WP-5D) and othersincluded a significant number of frequency bands, subsequent studies and discussions on band characteristics, current usage, and realistic utility for mobile broadband systems have now trimmed this list down to just a few practical options, including:

  • 1350-1400 MHz – proposed asthe uplink part of a new 1.5 GHz FDD spectrum arrangement (along with downlink in the band 1427-1518 MHz, which is already allocated to the Mobile Service);
  • 2700-3100 MHz – involving rationalization of current radar usage,and band segmentation,to enable a new IMT FDD or TDD spectrum arrangement to be implemented; and
  • 3400-3800 MHz – a possible new TDD spectrum arrangement may facilitate a globally-harmonized spectrum resource for IMT small-cells

While the band 470-698 MHz has endured intensive study and debate at prior JTG meetings, it otherwise remainsin wide usage for television broadcasting in many countries that still consider it an important conduit for socio-political awareness, and community support and cohesion. As such, it may not secure sufficient global consensus at WRC-15.

Someother bands also remain on the list developed by JTG 4-5-6-7. But,at bestthey appear to have attracted only limitedsupport from administrations,and not been subject of specific technical sharing studies,for a range of reasons:

  • Various band segments between 1518/1525-1710 MHz are generally perceived to be rather fragmented, offeringonly relatively small bandwidthstomeet future IMT needs– andare currently well-used by other important radiofrequencyapplications;
  • Various band segments between 4400-6725 MHz may be of longer-term interest to IMT-Advanced systems, but at this stage have attracted little explicit support for resolving near-term coverage/capacity needs.

It is therefore likely that most of these other bands willbe removed from the JTG 4-5-6-7 list of suitable IMT bands during the JTG’s next and final meeting in July 2014 – while a possible new agenda item addressing future IMT bands above 6 GHz is expected to be proposed for the Conference following WRC-15.

Implications of Studies to date

Unsurprisingly, discussions to date on the IMT candidate bands have been highly contentious: there are no longer any easy options for resolving the additional spectrum needs of future mobile broadband systems. Nonetheless, most nations recognize the significantrole of mobile broadband systems in stimulating economic growth and development byimproving productivity in almost every commercial and industrial sector – along with strengthening social institutions and expanding community support. In most countries, there is also an increasing economic imperative to ensure that scarce spectrum resources are used to maximum efficiency – including not only efficiency of transmissions, but also avoiding spectrum being ‘sterilized’ or ‘denied’ due to poor receiver performance or by wasteful assignment and usage practices.

There are now fewer options available to administrations for meetingfuture mobile broadband spectrumdemand,andfor securing the flow-on benefits towardnational economic and social objectives. Thus, administrations will soon need to consider and decide on their preferred bands, to minimise the potential for unfavorable outcomes. But choosing preferred bands may involve some challenging trade-off considerations for manycountries. For example:

  • countries that foresee a heavy ongoing reliance on C-band satellite services will logicallyprefer thatmobile broadband needs are accommodated in other bands – such as the bands 1350/1427-1518 MHz and 2700-3100 MHz, and which may require systematic review and rationalization of current radar and other usage of these latterbands; while
  • other countries with extensive aeronautical and maritime radar systems could instead prefer to protect their S-band radar systems, and thus seek to identify the bands 1350/1427-1518 MHz and 3400-3800 MHz for future IMT.

Aside from the simple objective of protecting existing radiofrequency systems,an important consideration isthe implications of short-term decisions for longer-term nation-building opportunities – for example, a major influence on continuing growth of industrial investment and tourism is whether sufficiently fast wireless broadband services areavailable as-and-when necessary.

However, even if particular bands are identified by ITU-R for use by IMT, eachnation retainsits sovereign right to implement such band usage, or not – depending on local circumstances/needs and irrespective of the IMT identification. More specifically, existing frequency usage can continue unabated until such time (if at all) that an administration decides to permit IMT systems to be deployed in the relevant band. Thus, the key issue is whether these band options will be available for countries to considerin the futureif they need to do so.

Moreover, as in the case of previous band re-allocations, any change of usage of these bands will typicallyneedsome transition period (typically several years),subject tothe domestic situation and planningby individual national administrations.

demandimperative for mobile broadband

Importantly, the cost-effective response to emerging demand for additional mobile broadband capacity unequivocally relies on a combination of:

  • additional spectrum resources – more Hz.
  • technology improvements – improved bits/Hz; and
  • increasednetwork densification – increased bits/Hz/km2.

More specifically, none of these capacity improvement mechanisms arelikely to besufficient by themselves to meet forecast demand approaching the year 2020 - all three together are essential elements if countries are to meet their future wireless broadband capacity needs at lowest cost for domestic consumers.

Additionally, the imperative to find new bands goes beyond the simple need to cater for forward capacity demand. As new technologies are introduced they require access to new bands because the re-farming opportunities of existing bands is often limited –since current bands are occupied by legacy systems still in use, and the existing bandwidths are often too narrow for new broadband systems. For example most legacy 2G bands still can’t be immediately re-farmed for 3G or 4G systems without major restructuring – and this process needs sufficient time and planning effort, and can’t simply be implemented as an ‘overnight’ change.

Economic imperative for mobile broadband

A number of studies[1] have been conducted around the world in regard to the economic contribution or significance of mobile broadband systems.

For example, the Australian Communications and Media Authority (ACMA) has recently published a new report[2] thatshows the use of mobile broadband in Australia has enabled incremental growth in economic activity in 2013 of AU$33.8 billion – representing 2.28 percent of Australia’s total gross domestic product. Of this, AU$7.3 billion is attributable to the mobile communications sector itself, but AU$26.5 billion comes from time savings and efficiency improvements reported by other businesses and industries.

The leveraging effect of mobile broadband is clearly evidentin such studies - and similar experiences are found in many other countries. With the ongoing growth in wireless broadband usage and surging aggregate traffic volumes – along with the past experience of network operators underestimatingforward capacity demand – it isclearlycounterproductive to constrain future mobile broadband capacity by starving networks of vital spectrum.

It might also be noted that the significant revenue windfalls from past spectrum auctions has usually provided an attractive and significant fiscal boost for governments, which is often used to finance other social infrastructure and development.

RESULTS OF SPECTRUM SHARING STUDIES

At the outset, it was agreed that sharing studies submitted to the JTG were required to assume specific baseline systems characteristics and deployment scenarios, to facilitate comparison of the various results. Sensitivity analysis based on variations in certain characteristic values and deployment scenarios would then assist in identifying possible sharing conditions. However, theagreed approach does not appear to have been universally followed by all studies – and this has complicated the consideration of results, and possibly inhibited consensus. In particular, while some studies sought to identify the circumstances enabling satisfactory sharing between IMT systems and incumbent radiofrequency services, there were other studies submitted that seem deliberately constructed to prove incompatibility. Such contrasting approaches further contributed to heightened contention throughout the JTG proceedings to date.

While many of the spectrum sharing studies submitted to ITU-R concluded that co-channel sharing in the same geographic areas is not feasible, a number of studies highlighted that non-co-channel sharing is feasible– noting that:

  • in many countries, the relevant bands are only very lightly used (or not at all, in some cases) – and these countries are keen to realise economic benefit from the unused resource; and
  • the ITU-R already recognizes (Recommendation ITU-R SM.1132) that band segmentation is a legitimate form of sharing that offers benefits for bringing spectrum resources into more efficient usage; and
  • adoptinga sufficient combination of geographic separation and spectral guard-band can facilitate sharing without causing harmful interference – and resulting impact of incumbent systems performance was shown to be diminishingly insignificant.

Subsequent detailed review of those studies opposed to sharinghas revealed numerous shortcomings, including: assumed parameter values differing from agreed systems characteristics; omission of key parameters; numerical and methodology errors; inappropriate propagation models; and/or deployment scenarios that do not reflect realistic situations. Moreover, by inferring that unlikely events arising at extreme range are representative of the more general operational performance impact, these studies paint a highly unrealistic picture of actual sharing scenarios. If such worst-case scenarios were to stand unchallenged, then legacy and inefficient technologies could become entrenched – and administrations seeking to improve spectrum usage efficiency will be inhibited by incumbents (some even off-shore) or other minority interests.

Therefore, to present a fair and balanced view,Telstra suggests that the studies be summarised by way of the following concise and generalized draft CPM text:

Studies have shown that co-channel sharing in the same geographical area is not feasible in the band xxxx-yyyyMHz.

Some studies have suggested that non-co-channel sharing in the same geographical area may not be feasible, unless large separation distances are maintained between systems in the XXXXX Service and IMT systems.

But other studies have shown that non-co-channel sharing may be feasible, subject to a combination of geographical separation,inclusion of spectral guard-bands, and other mitigation techniques. These studies also suggest the use of band segmentation, as outlined in Recommendation ITU-R SM.1132, to facilitate satisfactory sharing of the band.

Where cross-border co-ordination is likely to be necessary:

  • A co-ordination procedure will need to be defined to provide a framework for discussions between the concerned administrations; and
  • The use of particular mitigation techniques may be subject to agreement between the concerned administrations.

Development of more definitive conclusions for incorporation intothe CPM Report could be very challenging - due to likelihood ofexcessively contentious debate at the forthcoming final meeting of JTG 4-5-6-7in late-July 2014, questioning the veracity and diversity of the various study assumptions and methodologies.

CONCLUSIONS AND RECOMMENDATIONS

The APT has already provided global leadership in developing and promulgating the APT-700 band structure (now 3GPP Band 28/44). This ground-breaking plan has been extensively adopted not only inRegion 3 but also bya growing number of countries in Regions 1 and 2. The opportunity for similar innovative leadership by Region 3 now arises in regard to higher bands.

Not surprisingly, the technical studies considered by JTG 4-5-6-7 make for contentious debate, and the sharing proposals are presenting new challenges –simply because the Table of Allocations is fully allocated to a wide range of Services. It is therefore clear that:

  • the easy solutions of the past are no longer available; and
  • administrationsmust be ready to engineer solutions in more complicated spectrum usage scenarios.

The primary motivation for embracing these challenges is to ensure more efficient use of the limited radio spectrum resource- consistent with the objective ofpromotingongoing national/global economic and social development.

Where co-existence with other Services can be achieved through frequency and distance separation (and other mitigation measures, if needed),the identification of additional bands for IMTretainsfuturespectrum usage options. Moreover, any such identification for IMT does not mandate the timing of spectrum change-of-use, nor does it insist on any change at all in every country. Butlocal planning flexibility is especially important in those countries where heavy reliance on wireless infrastructure is an important economic consideration.

Telstra,along with other regional wireless network operators,respectfully encourages the APT to favorably consider the bands1350-1518 MHz, 2700-3100 MHz and/or 3400-3800 MHz (or some reasonable portions thereof) ascandidate IMT bands – with a view to identifying around 500-600 MHz to meet forecast IMT spectrum needs around 2020 and beyond. APT support for some (or a portion) of these bands at the forthcoming JTG meeting in late-July 2014, and at the Conference Preparatory Meeting #2 (CPM15-2) in March next year, will help to ensure that the Asia-Pacific region remains the global engine for future world economic growth and prosperity.

Annexures attached……

ANNEXURES: STUDIES ALREADY SUBMITTED TO JTG 4-5-6-7

A1Sharing studies related to the theL-band

Both Telstra and GSMA have already submitteddetailed studies of sharing between IMT and generic RADAR systems defined by Recommendation ITU-R M.1463.

In particular, Telstra showed that the band 1350-1400 MHz could be identified for IMT uplink usage ONLY – that is, only IMT user-devices would transmit in this band – and the existing Mobile Service allocation in the band 1428-1518 MHz can be identified for IMT downlink signals (base-station transmit). In summary, the Telstra studies showed that sharing is feasible, subject to:

  • minimum geographic separation of 1.2 km; and
  • minimum guard-band between IMT channel edge and RADAR of 10 MHz.

The likelihood of interference to RADARs was shown to be vanishingly small: less than 0.01% of locations may exceed the threshold I/N = -6 dB for no more than 1 % of time:

Figure A1.1: Urban environment – 1350-1400 MHz – RADAR System 3

Figure A1.2: Rural environment – 1350-1400 MHz – RADAR System 3

A2Sharing studies related to the band 2700-2900 MHz

Similarly, Telstra and GSMA also undertook technical studies and made submissions to the JTG in relation to sharing between IMT and generic RADAR systems defined in ITU-R M.1464.

In this case, and based on previous studies of the potential for rationalization of RADAR systems usage of the band 2700-2900 MHz, especially in countries where it may be used inefficiently, Telstra has shown that that segmentation of the band per Recommendation ITU-R SM.1132 may offer significant opportunity to meet future IMT spectrum needs. Specifically, the Telstra studies again showed that sharing is feasible, subject to:

  • minimum geographic separation of 1.2-1.6 km; and
  • minimum guard-band between IMT channel edge and RADAR of 10 MHz.

Again, the likelihood of interference to RADARs was shown to be vanishingly small: less than 0.01% of locations may exceed the threshold I/N = -10 dB for no more than 1 % of time:

Figure A2.1: Urban environment – 2700-2900 MHz – RADAR System C

Figure A1.2: Rural environment – 1350-1400 MHz – RADAR System 3

A3Sharing in the band 2900-3100 MHz

In the case of sharing with maritime RADAR systems in the band 2900-3100 MHz, studies included both shore-based RADARs and ship-borne systems. For the coastal and ship-borne RADAR systems, implementing physical separation of IMT systems and IMT sector avoidance are both inherently more achievable since IMT systems are not deployed over water bodies.

Moreover, as noted in Recommendation ITU-R M.1460, maritime navigation RADARs have been concentrated almost completely in the band 3020-3080 MHz for the past several decades. So, other than coastal surveillance RADARs the band remains notably under-utilised. And, since coastal surveillance RADARs are primarily focused out to sea (where IMT user devices will be rarely, if ever, found), use of IMT sector avoidance and geographic separation mitigation techniques offer considerable protection and sharing opportunities.

Telstra submitted detailed sharing studies based on a deterministic model, and which concluded that sharing appears to be feasible where geographic exclusion zones are likely to be in the range 3-7 km – or less, if sector avoidance and physical obstructions are used to advantage.

______

Page 1 of 9

[1]For example: United Kingdom ; European Parliament ; United States ; Latin America ; World Bank(1) ; World Bank(2) ;

[2]