Study of Future Demand for Radio Spectrum in Canada 2011-2015

6.2 Fixed Wireless Access

6.2.1 Overview

Canada is one of the leading countries in Internet usage, access to broadbandFootnote 14 and broadband penetration (subscriptions)Footnote 15. An increasing number of Canadians are demanding greater access to content, including rich multimedia, audio, video and data.

In 2009, residential subscribers obtained their broadband services as follows: 54% cable modem, 39% DSL, 5% dial-up and the remaining 3% fixed wireless, satellites and other technologiesFootnote 16. Fixed wireless broadband represents a small percentage of the total broadband Internet access for the residential market.

Cost-effective technology is available for bands below 6 GHz to provide wireless broadband services. LTE is also likely to be available in instances where FWA spectrum is converted so it can be used as cellular spectrum, i.e. in the BRS band 2500-2690 MHz.

Enterprises represent 25% of the broadband Internet revenues and are served by cable modem, DSL, fibre optics, satellites and fixed wireless facilities. Broadband wireless access provides a number of conveniences not available with traditional wireline cable modem and DSL. This includes full portability to various locations within the serving areas.

Technology progress in the broadband spectrum above 20 GHz bands will enable more feasible deployments of broadband wireless facilities in those bands, and may be an alternative to cable modem and DSL facilities for enterprises and potentially for household Internet access.

6.2.2 Spectrum Inventory and Spectrum Utilization

The fixed broadband access spectrum consists of:

Below 6 GHzPoint-to-multipoint (PTM) technologies:

  • 30 MHz of WCS spectrum in bands 2305-2320 MHz and 2345-2360 MHz auctioned in 2004/2005;
  • 175 MHz of FWA spectrum in the 3475-3650 MHz range auctioned in 2004-2005 (three paired 25 + 25 MHz blocks and one block of 25 MHz);
  • As of 2005, more than a 100 MHz of MCS and MDS spectrum has been used to developed fixed broadband wireless access in the 2500-2690 MHz range; the band 2500-2690 MHz BRS is being rearranged for mobile use, with allowance for fixed broadband use in part of the spectrum. The BRS spectrum is expected to be auctioned by the end of 2012;

Above 20 GHz Point-to-point (PTP) technologies:

  • 400 MHz of BWA spectrum at 24 GHz, 1000 MHz of LMCS spectrum at 28 GHz and 800 MHz at 38 GHz was licenced using a comparative-competitive (beauty contest) process for wireless broadband serviceFootnote 17. Industry Canada has recently extended some of the licensing and has clawed back other spectrum for licensing on a first-come, first-served basis.
  • Another 1000 MHz of spectrum is available for point-to-point and point-to-multipoint use in the 26 GHz range. As well, the 31 GHz band has yet to be released for future broadband wireless.

These two groups of technologies and services are quite distinct: They serve different markets, their network architectures are different, and the methods of estimating their demand for spectrum are quite distinct.

FWA Services – the differences between Point-to-multipoint and Point-to-point

  1. Point to multipoint: Frequencies below 6 GHz
    • Rural areas of medium-low-population density;
    • Broadband for Residential and SoHo market;
    • Diverse technologies: Fixed WiMAX, FWA over Cellular technologies and Licence-exempt.
  2. Point to point: Frequencies above 20 GHz
    • Urban areas – e.g. fringes of city downtown;
    • Most customers are Tier 2 enterprises/SMEs;
    • Technology is similar to that used for Microwave Backhaul.

This Study provides separate projections for Service and for Spectrum Demand for these two sets of technologies/markets.

Accordingly, in presenting the results for FWA, this Report considers each of these two groups of technologies (PTM and PTP) and markets separately. PTM is covered in Section 6.2.5, and PTP is covered in Section 6.2.6.

6.2.3 Stakeholder Input and Research Analysis

Comments from stakeholders

  • A fixed wireless network carrier reported that it has more than 100 sites, covering portions of rural areas in three provinces using Motorola Canopy technology in the licence-exempt 900 MHz band.
  • Another operator reported plans the use of HSPA+ in the AWS band to offer broadband in rural Quebec, supported by government broadband subsidies.
  • Some stakeholders felt that the 3500 MHz band is well-suited for deployment in large urban areas, given the large amount of contiguous spectrum blocks available in Canada. Currently, WiMAX technologies are available for this band, and it is widely expected that LTE technologies will be available at this band as 3GPP defines the standards. The U.S. has recently identified 100 MHz of spectrum that could become available in this band.
  • It was reported that, in Canada, two main operators provide 95% of the residential broadband services using wireline, plus, their local Telcos provide most of the wireline business customers.
  • According to a stakeholder, the CRTC 2010-43 statistics stating 95% broadband coverage exists in Canada, tend to overlook that:
    1. at least 700,000 Canadian households are un-served by broadband access with download speeds of at least 1.5 Mbps (with 1.5 Mbps under-delivering what Canadians increasingly need);
    2. that 1.4 million households lack access to broadband access with download speeds of at least 4 Mbps; and
    3. that all of these households are located exclusively in rural and remote areas.
  • Some stakeholders reported that a significant amount of the fixed wireless spectrum that has already been licenced is held by the same companies that dominate the wireline broadband access market.
  • It was reported that wireless represents a small percentage of the total broadband Internet access for the residential market. According to this respondent, the demand for wireless access technology has likely peaked, and the demand growth will likely flatten.
  • A stakeholder felt that spectrum below 6 GHz has the characteristics to combine fixed and mobile service delivery. Unlicenced spectrum is used initially to meet the rural demand, but the long-term spectrum availability and reliability are not good.
  • Some respondents indicated that, while the 24/28 GHz, even the 38 GHz FWA, bands are assigned for point-to-multipoint applications, they have been treated more like PTP bands, and as broadband pipes, to businesses.
  • One provincial Telco reported having up to 5% of its broadband customer base using wireless access, mostly in rural areas. Customers located in extremely remote areas are using satellite-based Internet. In the future, the usage of fixed wireless access to provide voice and Internet data services may be a more economical alternative to replacement of longer spans of copper plant in rural areas.
  • In total, there are indications from respondents that 6% to 10% of Canada's population could be served efficiently using fixed broadband wireless for data and voice.
  • One FWA carrier, using mostly the 24/28/38 GHz bands, reported providing several-thousand business customers in more than 40 cities with business-grade broadband and voice services. Also, this includes the provision of backhaul facilities to a number of new AWS entrants.
  • It was reported that WiMAX is the technology of choice for the 3.5 GHz band. Although WiMAX is also available for the 2.5 GHz band, the LTE technology will eventually dominate.
  • A respondent felt that manufacturing cost-effective equipment for the 24 GHz BWA and 28 GHz band poses challenges. The same one felt that it is unlikely to be a business case for deploying FWA systems for residential customers in these bands in the foreseeable future.
  • In the opinion of a stakeholder, three main technologies will be deployed for fixed broadband wireless, namely WiMAX, DOCSIS-based wireless and proprietary OFDM equipment. WiMAX will be mostly used for the 3.5 GHz licenced band, as well as in the TDD 3.65 GHz band, which is lightly licenced. DOCSIS-based access technology has been deployed in Canada in the 3.5 GHz, 5.8 GHz, 2.5 GHz band, and in the newly available Rural Remote Broadband Service (RRBS) in spectrum available from unused TV channels 2-51. There is a potential market for some amount of proprietary OFDM equipment (or Wi-Fi-based gear) in the 900 MHz unlicenced spectrum.
  • A stakeholder stated that the 3.65 GHz shared licenced band (and 5 GHz unlicenced band) is providing broadband data and voice services in rural and some backhaul service to new AWS entrants.
  • One respondent reported energizing the 900 MHz licence-exempt band by using Canopy technology, the 2500 MHz band by using DOCSIS (covering more than 7.5 million households, or 63% of the Canadian residential market), as well as the 3500 MHz band also by using DOCSIS. No fixed broadband technology has been deployed in the 2300 MHz band as yet.
  • Respondents noted that, in the 2300 MHz WCS and 3500 MHz FWA bands, only a small portion of the licenced spectrum has been fully utilized in Canada. Many licence holders have not deployed their spectrum in any meaningful fashion.
  • A stakeholder reported that equipment in the 3.5 GHz band remains limited at present. The Canadian band plan does not align with the rest of the world. The short-range coverage of typical 3.5 GHz FWA stations also imposes high-cell-density costs, making blanket deployment very expensive. However, it will likely be utilized as a dense urban overlay of the urban mobility network, whenever handset technology evolves.
  • A provincial carrier plans to continue to use the 2.5 GHz band for FWA services and will replace the existing DOCSIS equipment with another wireless solution, such as LTE, for fixed wireless broadband services to rural customers who are outside the range of DSL service.
  • One FWA operator reported point-to-multipoint (PTM) deployment growth of 37% and point-to-point (PTP) deployment growth of 34% over the last five years; and projects an increase of 39% for PTM and 31% for PTP from 2010 to 2015.

6.2.4 Research

Information from Spectrum 20/20 proceedings, May 2011

  • According to a presentation made by Barrett, nearly 2.4 million households (HHs) could be defined as rural (less than 30 HHs per sq. km). Barrett suggested that FWA is an economical solution for rural areas with a density of between 6 and 30 HHs per sq. km. (i.e. 11% of 13.7 M HHs, which equates to 1.5 M HHs), and multimedia satellite service is an economical solution for rural areas with less than 6 HHs per sq. km. (i.e. 7% of 13.7 M HHs, or 0.9 M HHs). The average monthly data usage by rural subscribers was reported as approximately 16 GB.
  • Two new multimedia satellites are being launched this year and should provide broadband capacity to serve the Canadian market, especially the very low-density rural areas.
  • It was also commented on that Industry Canada's Tier 4 licensing of broadband mobile spectrum is not adequate to advance broadband services in rural Canada. For example, the Tier 4 area, which includes Calgary, has an area of 19,188 sq. km. and has 1 M people (est. 41,5000 HHs). 92% of the population resides in Calgary, and 8% of the population is in rural areasFootnote 18. In order to gain access to the 700 MHz and 2500 MHz for rural broadband, or the 8 % of the population, with Tier 4, one has to pay for 92% of the urban population. (Note: The Tier 4 should be modified to reflect rural areas to encourage broadband services. For the Calgary Tier 4, separating the rural areas would represent about 18,000 sq. km. and 80,000 people, or 33,200 HHs.)

CRTC Information published in 2010

In the 2009 CRTC Annual ReportFootnote 18, it was indicated that more than 93% of the broadband subscribers were being served by wireline facilities, such as cable modem and DSL.

The average broadband usage per subscribers was reported at 15.4 GB per month (12.0 GB downstream and 3.4 GB upstream).

100% of the urban households have access to broadband wireline-based facilities. In contrast, just 82% of rural households had access to broadband facilities. The remaining 18% in the latter group is the most natural market for FWA broadband services.

CRTC Information published March 2011Footnote 19

  • On average, Canadians spent 43.5 hours/month on the Web, almost twice of worldwide average of 23.1 hours.
  • Of 11 countries surveyed, Canada ranks first in the number of Web site visits per user per month at 95.2.
  • The CRTC estimates that 95% of Canadian household can access broadband services using a landline.
  • Satellites extend to virtually all households but have limited capacity.
  • The report to the Standing Committee of Industry, Science and Technology on Feb 3, 2011, indicated that:
    • The average usage per month per subscription (residences and businesses) was 15.4 GB/month in 2009.
    • 14% of all subscriptions (9.17 M) account for 83% of the Internet traffic. This means that:
      • 14% (1.3 M) of subscriptions average 91 GB per month,
      • The remaining 86% of the subscriptions (7.9 M) average 3 GB per month.
Table 6.2.1 — Information derived from tables on Internet and broadband availabilityFootnote 20
2008 2009 Growth Comments
Revenues [$ Billions]
> Residential
> Business
> Wholesale
> Other
$6.2B
$3.9B
$1.1B
$0.2B
$1.0B
$6.6B
$4.3B
$1.1B
$0.3B
$1.0B
6.3%
9.9%
1.2%
21.5%
-5.5%
>Good growth in revenues - indicative of GB traffic
>Business is approximately 25% of residential, in terms of GB traffic and subscriptions
Residential subscribers [M] 9.8 M 10.1M 2.6% >No data on number of business subscribers is available
Broadband availability to HH, excluding satellite
> National
> Urban
> Rural
95%
100%
82%
95%
100%
84%
Penetration (HHs)
> All speeds
> High speed
> BB ≥ 1.5 MB
> BB ≥ 5 MB
74%
69%
52%
41%
75%
72%
62%
44%
10.1%
7.3%
6.3% *
4.4%
>*Fig 5.3.4 of CRTC CMR estimates the number of BB connections greater than 1.5 MB to be 7.9 M subscriptions
(Cable 4.5 M, DSL 3.4 M)
Avg. Monthly Subscriber Traffic [GB /month/sub]
> Download
> Upload
9.1
3.2
12.0
3.4
> Total avg. monthly traffic per subscriber was 15.4 GB;
> 14% of all users (9.17 M) account for 83% of the total Internet traffic.

Source: CRTC CMR and Red Mobile Research)

6.2.5 Point-to-Multipoint FWA: Services and Spectrum Demand

This section presents the projections of broadband Internet subscribers and traffic to be accommodated by Fixed Wireless Access (FWA) facilities, the assumptions used to define the FWA market, the demand for spectrum and the results for alternative scenarios.

Point-to-Multipoint FWA – Clarification Regarding Services in the 2500 MHz Band

In reporting the demand for the High-Value Services, the primary principle in this report is to show the demands according to the types of service being delivered.

Given the changes to the BRS (2500 MHz) band, there are three distinct services operating in this band in the period of this Study.

The principle is applied to these services in the same way it is to all other services. This means that those in the 2500 MHz band are treated as shown in the box below.

Clarification: Analysis of Services in the 2500 MHz band

  1. FWA Services using Fixed WiMAX or any other non-cellular technology — mostly in mid- to low-density rural areas. This is accounted for here in the FWA Point-to-Multipoint Service, both for traffic and for demand for spectrum.
  2. FWA Services using LTE/HSPA technology — mostly in mid-to-low density rural areas. This, too, is accounted for here in the FWA Point-to-Multipoint Service, both for traffic and for demand for spectrum.
  3. Mobile Services using LTE/HSPA technology — in all areas, urban and rural. This is accounted for in the Cellular Services, both for traffic and for demand for spectrum.

Point-to-Multipoint FWA: Service Demand: Market Analysis

In developing the projections for service demand for FWA to deliver broadband Internet, the Study uses a combination of in-house expertise and primary research, which was reviewed against several sources of secondary research. This includes data from the CRTC Annual Report on the state of broadband subscriptions and types of facilities, number of households already passed by various access facilities, the average broadband consumption per subscriber and other information. Some operators provided information on their FWA facilities. This also included their views on broadband services and the economic opportunities for FWA and broadband satellite to address un-served or under-served rural areas.

In determining demand, the process began by reviewing the existing status of broadband Internet subscriptions in urban areas and the household broadband penetration rate. In studying the market, it is clear that there is a dominant use of wireline broadband facilities (cable/fibre with either DSL or cable modem) and limited use of FWA and broadband satellite. Furthermore, it can be noted that the situation in urban areas is, such that, close to 100% of the households (HH) and businesses are passed by wireline broadband facilities of some sort, with speeds of greater than 1.5 Mbps. With the high availability of competitive broadband wireline facilities, it was concluded that there are very limited opportunities to use FWA facilities (point to multipoint below 6 GHz) to serve urban households and small businesses. However, the need for large-capacity broadband wireless facilities (point to point at 24/38 GHz) was identified in urban and rural areas to complement broadband fibre distribution facilities.

For the FWA service, the large number of rural households and businesses were studied, especially in moderate and sparsely populated areas, which have limited or no access to broadband facilities. Broadband FWA facilities, including broadband cellular platforms, such as HSPA, will continue to play an important role in bringing broadband to un-served and underserved rural areas.

Furthermore, the role of broadband satellites was studied as suited to serve sparsely populated rural markets. It was concluded that, in moderate-to-sparsely populated rural regions, conventional wireline broadband facilities are either technically inadequate to provide high-speed services or are cost prohibitive. For many of these areas, some service providers depend on government subsidies to assist in bridging the high costs of broadband facilities.

A top-down analysis was conducted to project the growth of broadband subscriptions, which will be provided by a mix of technologies such as wireline facilities (DSL and cable modem), FWA (bands below 6 GHz and the bands 24/38 GHz), and broadband Ka-band satellites over the next five years.

The analysis included determining:

  • Market share across a mix of broadband facilities;
  • Subscription growth on FWA facilities (including fixed HSPA connections);
  • Average monthly usage in GB per subscriber;
  • Traffic mix – i.e. download/upload usage; and
  • Types of (rural) markets best suited for FWA and broadband satellite.

In modelling the service demand for broadband access and FWA, the following factors have been taken into consideration.

Approximately 20% of the rural households (700,000 HHs) have no access to broadband wireline and wireless services. Multimedia Ka-band satellites provide national broadband services, but satellite service is limited in capacity, and the Internet speed and price are not competitive with terrestrial facilities, especially for the more populated rural areas, where the business case for FWA holds.

Hence, in assessing the relative demand for FWA spectrum, it is important to consider that:

  • Existing FWA subscriptions are mainly in rural areas, and represent a small percentage of the total (urban-dominated) subscriptions for broadband services.
  • Broadband service in urban areas is very competitive with DSL and Cable facilities already available to 100% of the households, where these services are often bundled with landlines and cellphones.
  • According to stakeholder input and analysis, a percentage of rural households are still un-served, when it comes to broadband connectivity.

Point-to-Multipoint FWA: Projections of Subscribers and Traffic

The charts below summarize the projections for subscribers, data traffic, and total traffic for FWA (point-to-multipoint) below 6 GHz to households and Small-to-Mid-sized Enterprises (SMEs).

Figure 6.2.1 shows the number of subscriptions on FWA below 6 GHz, growing from 520,000 at the start of 2010 to more than 960,000 by 2015. This constitutes a near doubling of subscribers over the period from 2010-15. It also illustrates the evolution in the mix of technologies, with FWA services starting to be delivered using cellular technologies in the 2013-14 timeframe.

Figure 6.2.1 — Point to Multipoint, subscribers

Point to Multipoint, subscribers (the long description is located below the image)

Source: Red Mobile Research and Projections

Description of Figure 6.2.1

This figure provides the numbers of subscribers by technology, for point to multipoint systems. Technologies considered include cellular (HSPA+/LTE), WiMAX/OFDMA – MCS 2500 MHz band/ WCS 2300 MHz, License exempt – bands below 6 GHz, and WiMAX/OFDMA – 3500 MHz band. Numbers of subscribers are provided annually, from 2007 to 2015. Total annual subscribers start at 300,000 in 2007 and increase to over 950,000 by 2015.


FWA has a larger potential in rural areas, where large segments of the market are under-served. As a result, in this Study, it was forecast that FWA subscriptions, as can be noted in Figure 6.2.1, would grow from 520,000 subscribers in 2010 to 960,000 in 2015.

Traffic is projected to grow from 15 GB/month/subscription to 45 GB/month/subscription during the period from 2010 to 2015.

Total traffic grows from 8 M GB/month to 44 M GB/month, as shown in the chart below.

Figure 6.2.2 — Point to Multipoint, traffic per subscriber (GB/mo equivalent)

Point to Multipoint, traffic per subscriber (the long description is located below the image)

Source: Red Mobile Projections

Description of Figure 6.2.2

This figure provides traffic volume by technology for point to multipoint systems. Technologies considered include cellular (HSPA+, LTE), license exempt bands below 6 GHz, WiMAX/OFDMA - WCS 2300 MHz/MCS 2500 MHz band and WiMAX FWA 3500 MHz band. Total traffic volume in 2007 is around two million GB/mo and rise to around 44 million GB/mo by 2015.


Point-to-Multipoint FWA: Key Assumptions and Relationship between Service and Spectrum Demand

The key assumptions used for this calculation are as follows.

For each of the main established rural FWA technologies, including those in licence-exemptbands:

  • Assumptions are made regarding the approximate number of sites and sectors in which they have deployed and how this is likely to evolve out to 2015. The number of cells and sectors are not identified to protect the confidentiality of some of the information.
  • Data traffic per rural FWA subscriber is 15 GB/mo in 2010 rising to 45 GB/mo by 2015 (80% downlink, 20% uplink).
  • These FWA services are offered in areas of moderate-to-low-population density. The main market being areas with population density of between approximately 4 and 40 population per square kilometre, extending into more-densely populated rural areas. The services are not generally offered in urban areas (approximately more than 400 population per sq. km.) and are also not aimed systematically at high-density rural areas with a population of between 100 and 400. These assumptions are particularly relevant to FWA over LTE/HSPA — service offerings and take-up to 2015 are limited by the availability of LTE in areas of suitably low-population density. The assumption is that FWA over LTE/HSPA is not offered in significant volumes in urban areas, where the traffic would consume expensive capacity and spectrum aimed at cellular services.
  • Spectral efficiency is assumed to be 4.0 bits/sec/Hz for the WiMAX and licence-exempt bands, and 1.3 (rising to 1.4 by 2015) for the cellular technologies.
  • Frequency reuse factors are assumed to be 4 for the WiMAX technologies and the licence-exempt bands, 1 for the cellular technologies.
  • Busy-hour traffic is three times that of 24/7 average-hour traffic.
  • Required capacity, to allow for burst rates for individual subscribers, is 1.75 times that of busy-hour traffic.

Point-to-MultipointFWA: Demand for Spectrum

The considerable change in traffic growth, shown in the previous subsection, drives some demand for additional spectrum for these technologies, as shown in Figure 6.2.3, below.

Figure 6.2.3 — Point to Multipoint, demand for spectrum

Point to Multipoint, demand for spectrum (the long description is located below the image)

Source: based on Red Mobile and PA Analysis, and PA PRISM Modelling

Description of Figure 6.2.3

This figure provides demand for spectrum in MHz required to serve the offered traffic for Fixed Wireless Access systems. Technologies consider include (HSPA+, LTE), license exempt bands below 6 GHz, WiMAX/OFDMA - WCS 2300 MHz/MCS 2500 MHz band and WiMAX FWA 3500 MHz band. The demand for spectrum in 2007 is around 20 MHz and increases steadily to around 210 MHz by 2015.


Growth in the demand for spectrum is slower than the growth in traffic. The main reasons for the difference are gains in spectral efficiency and an increase in the number of nodes/sites/sectors to cope with some of the growth.

The projections are, of course, sensitive to the assumptions regarding numbers of sites and sectors and how these grow over time.

Point-to-Multipoint FWA: Sensitivity analysis – Assessing Spectrum Demand Using Alternative Metrics

For point-to-multipoint FWA, the effect of including further operational constraints, further increasing the measure of the demand for spectrum, was also examined the same way as was done for cellular. The results of this analysis are shown in the next chart.

The solid bar (1) shows the demand for spectrum, on the primary measure used above: i.e. what is needed to carry the traffic at an acceptable Quality of Service, in the Busy Hour, etc.

The two higher measures of demand are shown as lines on the chart. They are calculated in the same way as was done for cellular:

  • Solid line (2): including allowance for minimum channel widths per licenced operator/per technology;
  • Dashed line (3): also allows for channel widths and also for the fact that timing of demand growth cannot be predicted, and, therefore, allows two years either way on the demand figures for each technology;

Including the above operational constraints inevitably increases the figures, adding a further 100-150 MHz to the figures obtained on the primary measure of demand.

As with the Cellular service, the effect of using the different metrics for the demand for spectrum is additive (+100-150 MHz), rather than multiplicative (x2-x3).

Figure 6.2.4 — Point to Multipoint, includes operational measures of spectrum demand

Point to Multipoint, includes operational measures of spectrum demand (the long description is located below the image)

Source: based on Red Mobile and PA Analysis, and PA PRISM Modelling

Description of Figure 6.2.4

This figure provides the demand for Fixed Wireless Access spectrum in Canada from 2010 to 2015. The busy hour traffic is steadily increasing from 20 MHz in 2007 at around 210 MHz by 2015. The chart also provides lines to indicate the increased demand as a result of adding operational constraints. In 2015, after applying minimum channel widths, the demand rises from about 210 MHz to 270 MHz, and after applying busy hour traffic, the demand rises to over 350 MHz.


Point-to-Multipoint FWA: Assessment of Alternative Scenarios

Figure 6.2.5 below shows the projections for traffic growth in the alternative scenarios. The projections are higher in Scenario 2 (Wire-Free World), and slightly lower in Scenario 3 (Low Investment).

Figure 6.2.5 — Point to multipoint, traffic by scenario

Point to multipoint, traffic by scenario (the long description is located below the image)

Source: based on Red Mobile and PA Analysis, and PA PRISM Modelling

Description of Figure 6.2.5

This chart provides projections on offered traffic in GB/mo for FWA Point to Multipoint. The offered traffic is provided for the three scenarios, from 2010-2015 and is summarized in the following table. It is noted that the values in the chart use a small scale, and the following are estimates, rounded to the nearest million:

Point to multipoint, traffic by scenario
BAU WFW Low Inv
2010 ~8 Million ~8 Million ~8 Million
2011 ~11 Million ~13 Million ~11 Million
2012 ~16 Million ~21 Million ~14 Million
2013 ~22 Million ~35 Million ~18 Million
2014 ~31 Million ~56 Million ~23 Million
2015 ~45 Million ~93 Million ~30 Million

Projections for Spectrum Demand are shown in Figure 6.2.6 below. They follow a similar pattern, but slightly less extreme. The main reason for this is the differences in the projected rates of growth in sites and sectors. In particular, in Scenario 3, the slower growth in traffic is partly counter-balanced by lower growth rates for the numbers of sites and sectors, so the changes in spectrum demand are less dramatic than the changes in traffic volumes.

Figure 6.2.6 — Point to multipoint, spectrum demand by scenario

Point to multipoint, spectrum demand by scenario (the long description is located below the image)

Source: based on Red Mobile and PA Analysis, and PA PRISM Modelling

Description of Figure 6.2.6

This chart provides projections spectrum required to serve offered traffic in MHz for FWA Point to Multipoint. The demand is provided for the three scenarios, from 2010-2015 and is summarized in the following table. It is noted that the values in the chart use a small scale, and the following are estimates, rounded to the nearest MHz:

Point to multipoint, spectrum demand by scenario
BAU WFW Low Inv
2010 ~90 MHz ~90 MHz ~90 MHz
2011 ~100 MHz ~105 MHz ~100 MHz
2012 ~110 MHz ~130 MHz ~110 MHz
2013 ~130 MHz ~175 MHz ~130 MHz
2014 ~150 MHz ~240 MHz ~150 MHz
2015 ~200 MHz ~425 MHz ~180 MHz

As with the spectrum demand for Cellular services, there is likely to be some further balancing feedback loops between available spectrum and spectrum demand.

The most evident example of this comes from considering the projections for Scenario 2 – Wire-Free World (high growth), which suggest that the demand for spectrum may grow rapidly over the period 2013-15, and this may be felt as pressure on spectrum. There are ways for operators to alleviate some of this pressure, although, in general, this comes at a costFootnote 21 to operators and/or consumers, so, the greater the pressure on spectrum, the more these means of alleviating the pressure are likely to be deployed.

6.2.6 Point-to-Point FWA: Services and Spectrum Demand

Point-to-Point FWA: Subscribers and Traffic

The FWA point-to-point part of the model does not directly use projections of traffic. Instead, it starts from assumptions, regarding the capacity required on each link. Accordingly, this Report does not provide charts for traffic projections.

In terms of the development of the market, the projections are for slow continued growth in links (on the order of up to 5%-10% pa) and a rather more-rapid growth in traffic, on the order of a threefold growth in traffic per link, from 1 TB/link/month to 3 TB/link/month.

Point-to-Point FWA: Key Assumptions to Get From Traffic to Spectrum Demand

Key assumptions are as follows:

  1. The point-to-point links are highly directional, and the general characteristics resemble those of microwave links much more than they resemble the other FWA technologies.
  2. Current links are at up to, approximately, Ethernet speeds, with required capacity averaging 80 Mbps in 2010, rising to 200 Mbps in 2015. But, the majority of these links operate most of the time in the range of 2-20 Mbps.
  3. Spectral efficiency is 4 bits/sec/Hz, frequency reuse is 6 (i.e. six frequencies are needed in order to provide channels for all links), and sufficient channels — for two independently planned networks — are needed to operate in each neighbourhood. These assumptions do not vary over the period of 2010-15.

Because of the way that the point-to-point part of the FWA model works — i.e. it starts from assumptions about the required capacity per link and applies factors for frequency reuse and the other aspects of network design — the calculations do not directly use assumptions regarding total traffic, busy-hour factors or required headroom.

Point to Point: Spectrum Demand

The growth in demand for spectrum is shown below. Growth in traffic and the resulting capacity required on the links are the drivers of increase in demand for spectrum. Changes in the number of point-to-point FWA microwave links are not a major driver of growth in spectrum demand, and so long as they do not increase the allowance that needs to be made for frequency reuse, they do not have an impact.

Figure 6.2.7 — Point to point, spectrum demand

Point to point, spectrum demand (the long description is located below the image)

Source: based on Red Mobile and PA Analysis, and PA PRISM Modelling

Description of Figure 6.2.7

This chart provides projections spectrum required to serve offered traffic in MHz for FWA point to point technologies (24/28 GHz). The demand from 2007-2015 is summarized in the following table. It is noted that the values in the chart use a small scale, and the following are estimates, rounded to the nearest MHz:

Point to point, spectrum demand
2007 ~112 MHz
2008 ~115 MHz
2009 ~119 MHz
2010 ~230 MHz
2011 ~300 MHz
2012 ~390 MHz
2013 ~490 MHz
2014 ~600 MHz
2015 ~760 MHz

Point to Point: Assessment of Alternative Scenarios

For point-to-point FWA, the alternative scenarios were modelled using different assumptions regarding the rate of growth in the required capacity per link.

The implied traffic growth figures for 2015 are a factor of two higher in Scenario 2 (Wire-Free World) and 25% lower than in Scenario 3 (Low Investment).

The chart below (Figure 6.2.8) shows the spectrum demand projections for the alternative scenarios.

Spectrum Demand follows a similar pattern to the implied traffic projections.

Projections for Scenario 2 — Wire-Free World/high growth — suggest there may be a high level of pressure on spectrum by 2015.

As with other services, there may be ways for operators to alleviate some of this pressure, including migrating traffic to frequencies above 38 GHz or substituting from microwave to high-speed fixed connections. Like cellular, this service also exhibits the presence of a balancing feedback loop between spectrum supply and demand.

The projections are sensitive to the assumptions regarding the degree of coordination between operators in the same local area. If infrastructure completion were permitted or encouraged in a way which made less-efficient use of spectrum — for example, if there were twice as many operators in each local area, each with their own set of frequencies — this would have a corresponding pro-rata impact on the amount of spectrum required.

Figure 6.2.8 — Point to point, spectrum demand by scenario

Point to point, spectrum demand by scenario (the long description is located below the image)

Source: based on Red Mobile and PA Analysis, and PA PRISM Modelling

Description of Figure 6.2.8

This chart provides projections spectrum required to serve offered traffic in MHz for FWA Point to Point (24/28 GHz). The demand is provided for the three scenarios, from 2010-2015 and is summarized in the following table. It is noted that the values in the chart use a small scale, and the following are estimates, rounded to the nearest MHz:

Point to point, spectrum demand by scenario
BAU WFW Low Inv
2010 ~230 MHz ~230 MHz ~230 MHz
2011 ~300 MHz ~300 MHz ~300 MHz
2012 ~390 MHz ~420 MHz ~350 MHz
2013 ~490 MHz ~600 MHz ~400 MHz
2014 ~600 MHz ~850 MHz ~470 MHz
2015 ~760 MHz ~1200 MHz ~520 MHz

6.2.7 Conclusions

Conclusions are given separately for Point-to-Multipoint FWA and for Point-to-Point FWA.

Point-to-multipoint FWA: Canada has one of the highest penetrations of broadband Internet access of the OECD countries. One of the opportunities for FWA growth is that more than 14% of the rural households are un-served or underserved. These are the most natural markets for growth in point-to-multipoint FWA services.

Canada has licenced greater than 200 MHz of FWA spectrum, in addition to the provision made to use licence-exempt bands and locally unused TV broadcasting channels.

FWA point-to-multipoint (PTM) technologies will play a growing role in delivering broadband to households and small businesses in rural areas. The existing networks using WiMAX and other technologies will continue to operate and grow, in terms of their traffic. Finally, HSPA/LTE cellular access will provide a further means to serve part of the rural market by 2015.

The Study forecasts a significant service and spectrum demand to serve the broadband rural market using FWA (licenced and licence-exempt) PTM facilities and advanced HSPA-LTE cellular networks. Figure 6.2.4 projects the spectrum demand to double (from 150 MHz to 300 MHz) for the 2010-15 period to meet the broadband-access service demand.

Point-to-point (PTP) FWA: The high-capacity point-to-point links are providing an effective way to serve part of the large enterprise market and other high-speed connections requirement, in addition to multiple DS-3 (SONET) fibre-optic and wireline carriers.

PTP links require extensive spectrum, as shown in Figure 6.2.8 above. By 2015, the spectrum requirement varies according to the traffic scenarios and is projected as being between 500 and 1200 MHz.

Finally, the Study has identified a likelihood of the demand for FWA spectrum being influenced by some balancing feedback loops, in the same way as seems likely for Cellular services. This balancing feedback will tend to bring the demand for spectrum into closer alignment with the supply of spectrum, for both point-to-multipoint and for point-to-point FWA.

Date modified: