15

Terrestrial Broadcast/Cellular Network Integration

15.1 Broadcasting in Historical Context

In 1923, Captain PP Eckersley,1 the BBCs first chief engineer launched a new era in broadcasting at Daventry in the Midlands. The first years of radio had individual stations transmitting programmes to individual cities across the country, from London to Glasgow.

The BBC transmissions in London for example were initially from Marconi House in the Strand, using a transmitter located in an attic room with aerials strung between towers on the roof adjacent to Bush House, home of the BBCs World Service. The transmission call sign was 2LO named after the transmitter illustrated in Figure 15.1.

Eckersley decided to have total coverage of the UK from one powerful transmitter located in the centre of the country. The site chosen had an ancient anglo saxon oak supposed to mark the exact centre of England, the Daven Tree, hence Daventry. The site was bought in 1924 and in July 1925 the world's first high-power low-frequency broadcasting transmitter was installed, known as Daventry 5XX and nicknamed the ‘old gentleman’ by its team of engineers. The 26-kW transmitter was the size of a tennis court and powered by two diesel generators each as big as a double-decker bus.

The transmitter achieved 85% population coverage, better than Eckersley expected. Within a few years smaller transmitters were built across the country but most radio receivers continued to have a mark on the dial saying ‘Daventry’. Foreign short-wave transmissions began at the site in the 1930s and the engineers received Christmas presents from listeners, including tea from India and coffee from east Africa. If locals had an earth fault on their telephone system, they would hear the BBC Far Eastern Service in Malay or Thai when they picked up the phone.3 The Daventry site also saw the birth of radar, after transmitters there were used to test if aircraft could be spotted by bouncing waves from them.

The transmitters were retired in March 1992 after 67 years of service. Most of the land where the antennas had stood was sold off, reverting to farmland or being used to build houses. The few remaining buildings were initially used for storage, before being sold to Crown Castle in 1997 when the BBC transmitter sites were privatised.

These long-wave and short-wave systems were an example of high-power wireless, beautifully engineered with frequency synthesisers built into seven-foot tall 19-inch racks. To achieve an 85% coverage was remarkable given that Northants had (and still has) poor ground conductivity, the combination of sandstone and a low water table), Orfordness, set in salt water was in later years to prove more effective. However, the site is also home to a DAB broadcast transmitter supporting local radio and localised infocasting and data casting, so may be playing a role in a newly developing radio network offer.

Similarly this bit of history might remind us that a rebirth of short-wave, medium-wave and long-wave long-distance transmission may be possible. Unlike the Voice of Russia, the BBC did not cut back on short-wave transmission and over the years, certainly up to a few years ago listeners steadily increased both in developing and developed countries. In 2010 the Foreign Office in the UK announced significant cut backs in World Service funding. A year later the decision was reversed and the Hindi, Arabic and Somali services were all reprieved. ‘Political radio’ and political broadcasting remains alive and well.

15.2 Digital Radio Mondiale

In the late 1990s, The Digital Radio Mondiale (DRM) Consortium4 successfully lobbied for an international standard for digital AM for frequencies below 30 MHz. Digital AM is comparable to FM mono in terms of sound quality and can be broadcast with a footprint of 1000 miles from one transmitter. DRM can be used for a range of audio content, including multilingual speech and music, and has the capacity to integrate data and text. The DRM signal is designed to fit in with the existing AM broadcast band plan, based on signals of 9 kHz or 10 kHz bandwidth. It has modes requiring as little as 4.5 kHz or 5 kHz bandwidth but also includes modes that can take advantage of wider bandwidths, such as 18 or 20 kHz with applications that include fixed and portable radios, car receivers, software receivers and PDAs. Remember that ‘wideband’ is a relative term and 20 kHz bandwidths below 30 MHz are to all intents and purposes ‘wideband’ in terms of their application potential.

Whether DRM can survive competition from the internet in the longer term may be subject to debate but this is a supremely energy-efficient mechanism for delivering information globally and should not therefore be discounted. It is not just radio but can distribute text, pictures and html files. As we shall see in later chapters it may also have a role to play in public safety and disaster relief. Short-wave radio irrespective of whether it is analogue or digital will almost certainly continue to be important politically and educationally. The National Association of Short Wave Broadcasters5 issues a monthly newsletter highlighting examples of the difference that short-wave radio continues to make particularly in countries where unrestricted access to the internet is not allowed.

15.3 COFDM in DRM

The DRM system uses COFDM with the number of subcarriers variable to suit channel allocations and required range (transmission resilience). There are three different types of audio coding. MPEG4 AAC audio coding is used as a general-purpose audio coder and provides the highest quality. MPEG4 CELP speech coding is used for high-quality speech coding where there is no musical content. HVXC speech coding can be used to provide a very low bit rate speech coder. DRM is therefore effectively a (very) wide-area version of DAB and DMB and was fully ratified as an ETSI specification (the ETSI ES 201 980 V1.2.2 (2003/2004) DRM system specification and 101.968 V1.1.1 (2003–4) data casting standard. In parallel, the ITU ratified DRM for use in the medium-wave AM and long-wave frequency bands in Regions 1 and 3 (Europe, Africa, the Middle East, Asia and Australia/New Zealand).

15.4 Social and Political Impact of the Transistor Radio

Integrating broadcast receivers into small form factor hand-held devices is not particularly new as a design concept. There is some dispute as to who produced the first ‘pocket-size’ transistor radio. A product called the ‘Regency’ introduced in October 1954 came to market in parallel with early Texas Instruments pocket radio receivers. This triggered a ‘form factor’ race with Sony producing the ‘world's smallest radio’6 (March 1957). Two examples are illustrated in Figure 15.2.

Form factor and functionality is still directly dependent today on device geometry. In 2005 ‘scalability roadmaps’ suggested that 95-nm devices would be replaced with 70-nm and 65-nm devices that would be replaced with 32-nm devices and this has duly happened. Device geometry determines DSP, microcontroller and memory functionality that in turn determines what we can do in a small form factor power limited handset. This in turn enables us to realise system value, which in turn realizes spectral value. Put another way, device bandwidth directly translates into application bandwidth, which directly translates into system value, which directly translates into spectral value.

In developing countries it could be argued that mobile phones are replacing, or will replace, the transistor radio as the lowest common denominator way of accessing information and delivering information. Governments have started to use SMS texting to distribute political messages, a parallel delivery route to TV and national radio so it could be argued that this process of transition is already well underway.

On the other hand, transistor radios are, however, significantly simpler than cellular phones, for instance they don't have a transmitter. This means that they have a lower component cost and less battery drain. For someone earning a dollar a day or less even the most basic mobile phone remains an unaffordable luxury. Almost anyone can afford to own and use a transistor radio. Short-wave transmissions like satellite transmissions also travel effortlessly over geographic boundaries. Partially successful attempts to jam short-wave broadcasts in some countries suggest that some governments remain acutely aware of the potency of this often overlooked delivery option.

Earlier, we mentioned that the UK sold off the BBC transmitter sites in 1997 and in common with many other countries is presently preparing to auction TV spectrum released by the transition from analogue to digital terrestrial TV. In a sense this seems to be trading short-term treasury gain against a potential loss of control over national TV and radio broadcasting, with a risk that political influence shifts to third parties, owners of satellite broadcasting networks being one example. In the US the National Association of Broadcasters7 is presently lobbying against the proposal by the FCC to take away an additional 120 MHz of TV broadcasting spectrum. 108 MHz, channels 52 to 69 had been taken away and auctioned in 2007. The additional spectrum would take out channels 31 to 51 that would mean (according to the NAB) that over 600 of the 1735 full-power radio stations would either need to close down or move to VHF, which would be rather impractical.

Looking at this globally, Table 15.1 provides an overview of the 500 MHz of broadcast (radio and TV) bandwidth available between long wave and 2 GHz. Table 15.2 lists the associated digital broadcast radio system options.

Table 15.1 Frequency allocations and digital broadcast system options

Radio Band	Frequency	System Options
Long Wave	3 kHz–300 kHz	DRM (Digital AM)
Medium Wave	300 kHz–3000 kHz (3 MHz)	DRM (Digital AM)
Short Wave	3 MHz–30 MHz	DRM (Digital AM)
VHF (FM radio)	88–108 MHz	DAB/DMB
VHF Band 3	174–233 MHz	DAB/DMB (218–230 MHz)
UHF Band 4	470–490 MHz	DVB-H, ISDBT, Media FLO
UHF Band 5	790–862 MHz	DVB-H, ISDBT, Media FLO
L Band	1452–1492 MHz	DAB/DMB
	1670–1675 MHz	DVB-H

Table 15.2 Broadcast receivers as at 2005 through to 2010

Broadcast receivers	Functionality
Analogue AM	Voice and low-bandwidth audio
Analogue FM	Voice, stereo radio, text and images (visual radio)
DRM	Voice, audio, text, data, images
DAB	Voice, audio, text, data, images, video
DVB-H	Voice, audio, text, data, images, video
ISDBT	Voice, audio, text, data, images, video
Media FLO (integrated with 1XEV)	Voice, audio, text, data, images, video
EDGE (Dual transfer mode).	Voice, audio, text, data, images, video
HSPA	Voice, audio, text, data, images, video

15.5 Political and Economic Value of Broadcasting

Broadcast spectrum has (always had) political value (Hitler, Mussolini, Churchill, Berlosconi), social value, evangelical value (Vatican radio) and economic value.

Economic value is a composite of license fee income (BBC in the UK), advertising revenue and (increasingly), participation bandwidth revenue. Participation revenues are a composite of text and voice value and (increasingly) image and video value, anything that works on a game show.

The realisation of fiscal value from spectrum measured in dollars per hertz or euros per hertz is dependent on delivering robust radio systems with adequate link budgets that ensure enough flux density to support consistent good quality reception of voice, audio and video content. This in turn is dependent on realising good receiver sensitivity, selectivity and dynamic range in TV and radio receivers. The earlier chapter on smart televisions (Chapter 8) also highlighted the value realisable from a return channel, in that example a broadband internet connection.

Note how participation revenues (which in some countries like Finland now exceed advertising revenues) are closely dependent on the robustness and consistency of this return channel. Participation revenues, in theory should increase as participation bandwidth increases. (This may not be true for participation margins where it will be hard to match the margin achievable on SMS voting and texting in terms of euros per hertz or euros per delivered megabyte.)

Note also the blurring of definitions between audio and video and text in these broadcast radio systems – visual radio with text is essentially competing directly with wider bandwidth digital TV. Any/all of the broadcast options are capable of triggering uplink bandwidth value.

So we are interested in the performance of the receiver in terms of sensitivity, application bandwidth and power efficiency and the performance of the receiver in the presence of locally generated transmit power (the return channel). We should also remember that we can also deliver radio and TV channels over existing cellular radio bandwidth.

This means that our choice of broadcast receivers could include existing analogue AM or FM radio, digital AM (DRM or equivalent), DAB/DMB, DVB-T or HSPA or EVDO mobile broadband products. Up until 2010 there was the additional option in the US of a system optimised for mobile reception known as Media FLO (forward link only) developed and promoted by Qualcom and an equivalent version of DVB T known as DVB H (handheld). Both these systems were vigorously promoted from 2005 and both failed for reasons we need to examine.

A number of phones included and still include integrated FM tuners. These work as well as most FM tuners when used in mobile applications, that is quite well sometimes but not consistently well in weak-signal or high-mobility conditions.

There is nothing wrong in principle in adding digital subcarriers to existing analogue bandwidth and using these carriers for data and image transmission – an analogue and digital multiplex. This provided a perfectly adequate basis for first-generation ‘visual radio’ systems, for example in Finland.8

In practice, over the longer term, it should be possible to get better bandwidth and power efficiency and a more consistent user experience from a digital broadcast receiver. The same applies for analogue AM systems.

All of the digital broadcast systems should in theory and generally in practice provide a performance benefit in terms of quality, consistency and bandwidth and power efficiency. DRM would be (relatively) easy to integrate into present cellular handset form factors and dongle products were introduced9 but failed to gain mass-market traction.

A DRM receiver works well under a wide range of operating conditions and is power efficient. It is, however, bandwidth limited and (at present levels of compression) incompatible with any existing TV content standards. There is a wide range of global programming available, Vatican Radio being one example. Up-to-date information on DRM receivers and DRM programming can be found on the Digital Radio Mondiale web site.10

15.6 DAB, DMB and DVB H

DAB/DMB is potentially useable across three radio bands – the VHF FM band, band 3 at 220 MHz and L band at 1.5 GHZ. It is MPEG2 TS and MPEG4 Part 10 compliant and supports useful features like multimedia object transfer, so it's definitely not now just an audio delivery system. At 2011 there are only 15 countries with DAB networks deployed, which is not enough to achieve global mass market adoption.11 It does work well in the UK and despite the limited global uptake there are nice receiver products available.12

DVB-H has proved more problematic. The intention was that cellular phones would have mobile TV delivered via a DVB H physical layer either at Band 3 VHF, Band 4 or Band 5 UHF. Crown Castle purchased a 5-MHz bandwidth allocation in the US between 1670 and 1675 MHz and in 2005 started a DVB H trial in Pittsburgh.

Technically, DVB H was quite attractive. DVB-H increased the power efficiency of the receiver by time slicing. If multiplexed with a DVB T signal for example, a 2-Mbit burst would be taken out of the DVB T 15-Mbps stream and sent in a 146-millisecond burst. The receiver then powered down for just over 6 seconds then powered up for the next burst. The 2-Mbit burst was read into and out of a buffer at a constant 350 kbps.

However, while this optimised the receiver power budget, it saved less power than might seem immediately apparent (the baseband processor still had to work pretty hard) and there had to be careful (and fast) synchronisation with the continuous and scattered pilots modulated on to the OFDM signal.

Extended receiver power down was, however, a well-understood technique and had been used in paging systems for at least 30 years, so it should have been possible to make this work satisfactorily. The time slicing also allowed for neighbour measurements and mobile initiated handover in multifrequency networks. This would have allowed a high-density DVB H network to be overlaid onto an existing cellular network.

There were, however, some practical issues when implementing a DVB H receiver in to a cellular handset. Producing a DVB H front end capable of accessing low band VHF, UHF (and L band) was challenging in terms of antenna design resulted typically in negative antenna gain of the order of between 5 and 10 dBi.

Equally problematic was the issue of GSM (or equivalent cellular) transmit power and wideband noise from the TX PA desensitising the DVB H receiver. This implied either a careful choice of DVB H channel allocation and/or some carefully designed (and potentially expensive and lossy) filtering.

There were various ways in which DVB H receiver sensitivity could have been improved, for example by using antenna diversity or additional time diversity (using the optional MPE FEC encoder) but this would not have overcome the problem of locally generated interference within the handset. Diversity gains within present handset form factors would also have been marginal. The other problem was that people didn't want to watch TV on their handsets, particularly if they had to pay for it and the battery went flat. Additionally there was scant incentive for operators to bear the cost of delivering content which they did not own and with no associated revenue stream.

ISDBT had similar problems to overcome. ISDBT, specified originally for the Japanese market, is arguably the most scalable of the present digital broadcast OFDM systems. It divides a 6-MHz channel into subchannel segments each of just over 400 kHz each segment has a variable OFDM multiplex, variable modulation (QPSK, 16 QAM, 64 QAM) and variable levels of convolutional coding. Table 15.3 summarises these capabilities.

Table 15.3 ISDBT

Although the system promises substantial deployment flexibility, performance (as with DVB H) will be dependent on achieving successful (RF) integration into existing and future cellular transceivers and global scale. Updated adoption status by country can be found on the ISDBT web site.13

Media FLO was Qualcomm's proprietary offering for broadcast receivers. It shared some common techniques with DVB H in that it used a time division multiplex, with one or more traffic packets being transmitted within a reserved slot to all users in the service area. Users received the same packets from multiple cells. An OFDM multiplex was used to slow down the symbol rate and the symbols were soft combined to improve downlink performance.

Media FLO was deployed into Channel 55 in the UHF band. It had the merit of being intrinsically compatible with existing and future 1X EV DO networks but no inherent compatibility to the US TV ATSC system. The deployment of Media FLO prompted ATSC to develop a mobile optimised ATSC implementation. The market failure of Media FLO and subsequent sale of Channel 55 to AT and T in 2011 has not provided much motivation to move forward at all quickly on the mobile TV front.

Similarly, it would have been feasible in 2005 (though not necessarily economically attractive) to use EDGE as a broadcast radio bearer. A function known as dual transfer mode would have supported some ‘ring-fenced’ broadcast bandwidth within the existing slot structure. There are already broadcast packet channels for signalling bandwidth, so it would not have been a great leap to deploy broadcast packet channels for broadcast content and there were a number of receiver optimisation techniques such as joint detection (single antenna interference cancellation) that could be used to improve receiver performance.

15.7 HSPA as a Broadcast Receiver

HSPA was a more persuasive candidate. Release 7 included specific work items on receiver optimisation using advanced diversity and equalisation techniques. The pilot symbols on the channel can be used in a similar way to present OFDM-based broadcasting to provide active channel characterisation). The HSPA MAC (2-millisecond-based admission control) is also more IP friendly than the Rel 99 MAC and arguably more power efficient for broadcast reception). The less than enthusiastic market response to mobile TV by this time, however, meant that little was done in practice to commercialise this as a product offer.

There were also a number of deployment issues that arise from the inescapable fact that legacy cellular networks were not and are not optimised to support broadcasting applications (other than for signalling functions).

Traditionally (in the 1980s and 1990s), cellular networks were radio engineered to provide a balanced link budget. Given that handsets had less power available than base stations, it was usual to provide some additional link budget gain on the uplink. A typical cell site might provide for example 17 dB gain on the uplink using diversity and/or dual polarisation antennas and 9 dB of gain on the (sectored) downlink.

Similarly in the access network, the A bis and Gb interface was designed to support symmetric bidirectional (largely duplex voice-based) traffic. In contrast, digital broadcast networks used (and still use) ATM to deliver a one way traffic multiplex that is (completely) asymmetric and highly asynchronous (variable rate). The Rel 99 air interface addressed this issue by introducing an ATM-based radio layer (10-millisecond frame-based admission control) and an ATM-based radio access and core network.

The debate, however, then moved to focus on the future role of IP in broadcast networks. DVB-H for example was explicitly being positioned as an IP-based broadcast system solution. This meant that the IP subsystem at the transport layer had to behave at least as predictably as an ATM-based transport layer, or in other words, the IP subsystem had to replicate ATM end-to-end functionality (the ability to control a complex multiplex in the time domain). This was not something that IP networks were ever designed to do.

The integration of flexible network bandwidth provisioning and QOS based broadcasting was particularly challenging and needed to encompass transport and network design, server architectures and performance and content management systems.

In 2007 at the annual GSM World Congress trade show, mobile TV was everywhere. By 2011 it had all but disappeared. Is mobile TV dead or dormant and how will the success or failure of portable TV over the next few years impact the mobile TV offer?

Apart from users not wanting it, mobile TV market adoption has been frustrated by multiple standards illustrated by the muddle described above and a failure to harmonise spectrum on a regional and national basis. These together mean that scale economy could not and probably cannot be achieved.

15.8 Impact of Global Spectral Policy and Related Implications for Receiver Design and Signal Flux Levels

Global spectral policy is set every five years at the World Radio Congress, the next one, when this book was written, being WRC 2012. Five years earlier WRC07 produced policy on the following areas:

• The extension of Digital Radio Mondiale from long wave, medium wave and short wave to include the FM broadcast bands.
• The extension of the present T-DAB/DMB multiplex in Band 111 to include160 kbit enhanced audio transmissions and related European digital TV and DMB deployments.
• Agreement on additional HDD broadcast DTT DVB UHF transmissions and additional DVB H multiplex options.
• Finalised plans for allocating and auctioning 112 MHz of UHF between 470 and 862 MHz for cellular FDD transmission.
• Agreement on L band allocations between 1452 and 1492 MHz including potential DAB extensions and cellular FDD or TDD opportunities.
• The repurposing of S Band for DVB -SH and related hybrid satellite and terrestrial broadcast delivery options.
• International agreement on the 2.5-GHz cellular extension band from 2500 to 2690 MHz.

The emphasis was on producing harmonised band allocations that could deliver sufficient economies of scale to support new cost- and performance-competitive cellular transceivers with integrated broadcast receive capabilities. Some of this spectrum was made available at the expense of present users, including broadcasting and programme making and special events (radio microphones and professional wireless cameras).

The repurposing of this spectrum therefore implied significant engineering challenges and related business transition management opportunities. The repurposing of the upper end of the TV bands for mobile broadband has proved particularly problematic both technically and commercially. The difficulty revolves around the risk of hole punching. Hole punching is when a TV receiver is unable to receive a TV programme because it has been desensitised by a locally proximate mobile broadband device. One answer is to improve the dynamic range and selectivity of TV receivers but there are a lot of them already out there. Also, the problem is rather subtler than just dynamic range.

A DVB or ATSC UHF tuner will generally be designed to tune across the band. It has to do this because digital broadcast channels could be anywhere in the band depending on the country in which the device is being used or the part of the country in which it was being used.

The alternative is to produce country- or region-specific variants that can just work on the locally available channels. The disadvantage of producing country-specific or regionally specific variants is that it is difficult/ impossible to achieve economies of scale because addressable markets are too small.

The disadvantage of tuning across the band is that it is hard/ impossible to deliver good sensitivity from the device and the device is vulnerable to interference from other inband transmissions, for example from other mobile cellular handsets.

The difference is significant. For example, a digital broadcast tuner working across the whole UHF band would typically have a sensitivity of between −92 and −98 dBm. A UMTS handset working across a much narrower band, for example 25 MHz rather than 400 MHz could have a sensitivity of better than −120 dBm.

Admittedly this is not a like-for-like comparison as the UMTS handset is working in a 5-MHz channel and the digital broadcast channel is 6, 7 or 8 MHz, but the difference is in any case substantial. And the difference is important. One of the lessons learnt from the BT Movio experience in the UK and Modeo experience in New York, two examples of mobile TV trials that failed is that users with an integrated cellular/mobile TV and radio receiver expect to listen to digital radio and for that matter to listen and watch digital TV wherever and whenever they have cellular coverage. The present sensitivities and selectivity available from digital TV and radio tuners are not sufficient to meet that experience expectation.

One option is to increase broadcast transmission power levels. To an extent this is happening as digital switchover proceeds. In the UK for example digital TV signals were increased by 7 dB as analogue signals were turned off in a particular area.

The second option is to retransmit digital TV and digital radio signals from cellular transmitters. The objective with options 1 and 2 is to ensure that received signal strengths as seen by portable and mobile devices, in other words the flux density, are similar irrespective of whether the received signals are cellular or digital broadcast, a solution that has come to be termed as a cooperative network. An advantage of this is that portable televisions would also work and/or lap tops with DVB T or ATSC demodulators could provide users with cost- and energy-efficient linear TV programming. The link budget and bandwidth needed for HD, and in the longer term super HD, are particularly challenging for terrestrial broadcasting.

The broadcasting industry and cellular industry have been mast sharing and site sharing for over 30 years. The broadcasting and two-way radio industry have been mast sharing and site sharing for over 50 years. More recently, there has been a trend for cellular network operators to share mast space, Vodafone and Orange in the UK being a recent example. There is therefore an extensive body of knowledge and experience that can be applied to ensure that cellular, two-way radio and broadcast transmissions can coexist.

From a business perspective this is also good news for manufacturers of RF conditioning products. Mast sharing and site sharing has historically been based on the two-way radio and cellular industry being given, leased or sold space on broadcast transmitter masts. Increasingly we may now see the broadcast industry being given, leased or sold space on cellular macro- or microsites for local digital TV and local digital radio transmission.

The broadcast and cellular industries have a strong engineering common interest. Generally, when a strong engineering common interest exists there is also a strong commercial common interest, though sometimes this is not immediately apparent to the parties and vested interests involved. The broadcasting industry also has extensive experience of implementing OFDM-based physical layer transmission networks.

The consensus to emerge from WRC2007 was that mobile broadband would take three to five years to implement at the top end of the UHF band. This has proved overoptimistic, though commercial networks are now launched in parts of Europe including Germany and in the USA (AT and T and Verizon).

Engineering issues include managing coexistence between very different radio network topologies working at very different powers. In the days of omnipresent analogue TV, a country the size of the UK could be covered from 50 large high-powered transmitter sites and 1100 lower-power relay sites. This provided the statutory required coverage of 98% of the population including all communities with more than 200 people. Getting 90% coverage for a 3G network at 1900 MHz takes over 7000 base stations.

Major urban conurbations such as London were and still are covered substantially by one TV transmitter. For example, Crystal Palace in London broadcasts to a ‘local’ population of 8 million people. The analogue transmitter delivers one megawatt, cosharing mast space with DVB, DAB, FM and medium-wave transmissions. The station has a range of 60 miles for analogue TV and 30 miles for digital TV. HDTV broadcasts started from the site in May 2006.

Digital TV transmitters typically have effective radiated power outputs (the power that is delivered as a result of the transmitter power plus antenna gain) ranging from 2 to 50 kW. Crystal Palace is 20 kW. These output powers though relatively modest compared to analogue TV are still substantial when compared to cellular base stations. There is also a height-gain effect that increases effective coverage.

This provides economic benefits in terms of rural coverage and urban in building penetration and a benefit to users in terms of reduced DC power drain when viewing digital TV content. However, this in turn depends on network density.

15.8.1 Network Density and Cooperative Network Opportunities

There is a trend in digital TV broadcasting towards implementing an increasing number of lower-power sites, particularly in urban areas. Many of these potential sites are owned by cellular operators or administered, owned and/or managed by common third parties such as Crown Castle.

A faster than expected transition to high-definition TV including high-definition TV and an earlier than expected transition to super-high-definition TV strengthens the rationale for providing TV broadcast bandwidth from smaller sites previously optimised for cellular. The same principle could apply for local digital TV and digital radio broadcasts that could potentially be supported on interleaved spectrum.

Successful cosharing is, however, dependent on closer integration of broadcast and cellular technologies (DVB in Europe, ATSC in the US) and now LTE mobile broadband and legacy cellular networks and effective mitigation of shared spectrum interference.

15.9 White-Space Devices

This brings us to white-space devices.

Terrestrial TV networks are either implemented as single-frequency networks or more usually multifrequency networks. The reason for implementing multifrequency networks is that these are or at least have been high-power networks and the signals from a dominant mast can travel tens of miles and in some propagation conditions hundreds of miles. Compared to cellular networks (relatively low powered with transmit antennas installed generally at a much lower height than TV) TV networks are conservatively planned, which means that interleaved channels are often unoccupied in some places at some times. White-space devices are cognitive devices that are supposed to detect or be told when they can and cannot use this unused bandwidth.

On 4 November 2009 the US chose a new president and the FCC voted to allow white-space devices to be deployed into the US UHF TV band. This simple decision has potentially profound consequences for the broadcasting industry, the cellular industry and other communities with an interest in how regulated and unregulated and licensed and unlicensed spectrum is used in the future, both in the US, the UK, Europe and Asia.

The white-space ruling is the direct result of robust petitioning from the Wireless Innovation Alliance14 supported by eight technology companies, Microsoft, Google, Dell, HP, Intel, Philips, Earth link and Samsung Electro Mechanics.

White-space devices based on spectrum-sensing principles already applied in the 5 GHz U-Niii band have been heralded as the birth of a new age in fixed, portable and mobile broadband access. In addition to the core companies in the WIA, other companies such as Motorola are actively promoting white space for municipal wide area WiFi. Alternative suggestions include the use of the spectrum for low-cost cellular wireless backhaul. The deployment of these devices is, however, controversial. Existing users in the band including TV and wireless microphones contend that interference issues remain unresolved.

Cellular operators who have either already invested in the band (AT and T and Verizon in the US) or are preparing to bid in future 700- and 800-MHz UHF auctions question the impact that white-space devices will have on present and future mobile broadband business plans.

Additionally, white-space devices could be deployed in other bands in other countries. This could be perceived either as a threat or an opportunity for cellular operators and the service provider community.

The assumption is that as this is unlicensed spectrum, wireless access can be provided on a more cost economic basis, for example for free. In comparison, broadcasters using this spectrum have to cover their costs through licence fee or advertising income. Cellular operators using this spectrum have to recover their costs by charging an access fee.

The assumption that unlicensed spectrum is inherently more cost economic than licensed spectrum is only partly true and dependent on certain preconditions. The 2.4-GHz and 5-GHz ISM spectrum is ‘free’ in as far as there are no license fees that need to be paid, but there are still access costs that need to be recovered.

In domestic and SME applications, costs are limited to buying a wireless router and paying for an ADSL connection. Hardware costs are low because the ISM band is, more or less, a global allocation and this has allowed vendors to realise substantial economies of scale.

However, in other applications, for example public WiFi in hotels, airports, trains and hospitals, significant hardware installation and real-estate costs can be incurred that can in turn prompt aggressive access pricing. Similar constraints would apply to white-space devices.

For white-space devices to be successful they must have similar scale economies to existing WiFi and WiMax and cellular handsets.

The devices must support equivalent or faster data rates than existing cellular products and/or cellular handsets and when used as portable devices should use similar amounts of DC power. However, the receive and transmit functions have to work across operational bandwidths that are far wider (when expressed as a percentage of the centre frequency of the band) than all other existing mainstream radio systems. Additionally, the receiver dynamic range needed from these devices is substantially more than cellular or WiFi handsets.

This introduces a number of practical performance and cost issues.

15.9.1 Spectrum Sensing

White-space devices need to be able to detect when signals are present and not present. The signals to be detected are either high-power TV or low-power wireless microphones

One of the points when discussing the detection of TV signals is how reliably the pilot tone in an ATSC signal can be detected at a range of flux densities.

Companies such as Adaptrum,15 founded by Robert Broderson the cofounder of Atheros, claim that time-domain matched filter techniques working across the whole of the 6 MHz ATSC channel provide robust signal sensing techniques. Motorola have added in geolocation to provide additional protection.

Broadcasters, however, remain sceptical. Even if sensing an ATSC signal can be achieved consistently down to low flux densities it is still feasible for devices to be in areas shadowed by buildings or hills – the hidden node effect. If the devices need to work in countries with DVB TV (with 7 or 8 MHz channel spacing) they need to detect a signal with a much more complex (and by implication hard to detect) pilot symbol structure.

Wireless microphones are different – these devices use FM modulation and are inherently low power.

The requirement to detect two different types of signals has to be realised in the presence of relatively high incident received energy from other adjacent channels. These signals potentially desensitise the receiver front end, thus potentially compromising the sensing function of the device.

15.10 Transmission Efficiency

To date, most technical work has focused on receiver performance parameters but white-space devices are supposed to deliver two-way wireless connectivity. The difficulty here is the bandwidth over which the devices will need to operate. In the US this would be somewhere or anywhere between channels 21 and 51 (512 to 698 MHz) or potentially from channel 2 upwards (54 MHz). The maximum power of these devices is relatively modest, of the order of 100 milliwatts, but difficult to deliver efficiently across these operational bandwidths.

Producing a device that would work in other markets, for example Europe, would mean devices would need to work somewhere or anywhere in UHF band 4 extending up to 790 MHz. Designers of cellular phones know how hard it is to deliver efficient RF power amplifiers at lower frequencies over extended operational bandwidths. Getting an RF power amplifier to work efficiently across 40 MHz at 800 MHz is challenging, let alone 100 MHz or more at 600 MHz.

As a comparison, the widest cellular bandwidth as a percentage of centre frequency is presently GSM 1800. 75 MHz at a centre frequency of 1747 MHz is equivalent to 4.3%. The 80-MHz ISM band at 2.4 GHz is 3.3%. White-space devices could be of the order of thirty to 40% – an order of magnitude greater.

15.11 Scale Economy Efficiency

One solution would be for white-space devices to be market specific that would mean white-space devices for the US market (Region 2) and rebanded devices for Region 1 (Europe) and Region 3 (Asia).

However, this would make it impossible to achieve sufficient scale economy. This would be particularly true if only deployed in the US. By 2014 the US will constitute less than 10% of the global market for cellular devices. RF development today is typically amortised over hundreds of millions of devices. Such volumes would be inconceivable from the US white-space market alone.

15.12 Signalling Efficiency

White-space devices will have to measure and continuously remeasure bandwidth occupancy and share these measurements with other devices or access points in the network. The IEEE presently has a working group specifying these protocols but essentially this is a complex measurement process.

This means that the devices will not be spectrally efficient and more important from a users perspective will not be power efficient irrespective of the network topology adopted. The use of mesh-network topologies for example would result in particularly poor spectral and power efficiency. White-space devices would therefore be likely to have much poorer duty cycles than equivalent cellular devices.

15.13 Power Efficiency Loss as a Result of a Need for Wide Dynamic Range

The need for a wide dynamic range in the receiver front end would further reduce session duty cycles by increasing DC power drain. This would effectively invalidate any business plans predicated on the use of mobile or portable equipment.

15.14 Uneconomic Network Density as a Function of Transceiver TX and RX Inefficiency

Unless white-space devices somehow escape the fundamental laws of physics, it will be hard to realise transmission or receive efficiency across the required operational bandwidth.

The lack of consistent channel pairing would make an FDD band plan problematic, so the assumption has to be that this would be a TDD interface. This will result in additional loss of receive sensitivity particularly when devices need to transmit and receive at the same time.

This will compromise the measurement capabilities of the devices (see spectrum sensing above) but additionally implies a network density that would be unlikely to be cost economic. Mesh networks are not a solution. (See the section on signalling efficiency above.)

TDD also implies a loss of capacity over extended distances, a function of the time-domain guard band overhead needed to accommodate round-trip delay. This implies a relatively dense network topology.

Note that TV receivers are inherently insensitive due to their need to tune over an extended frequency range. This does not matter if signals are being received through a high-gain roof-mounted antenna but it does matter for devices with internal antennas and/or for devices used at ground level. White-space devices will be used typically at ground level and may or may not have internal antennas. They will need to tune across similar bandwidths but additionally have to generate transmit power for the uplink, not something a television has to do. This will result in further receive desensitisation that will translate into lower downlink data rates and/or denser more costly networks.

15.15 Cognitive Radios Already Exist – Why Not Extend Them into White-Space Spectrum?

Cognitive radios already exist in mass market applications. DECT cordless phones are one example and it could be argued that cellular handsets are cognitive in that they measure channel quality. The admission control algorithms in LTE networks for example are based on the channel quality indicator measurements received from LTE handsets.

The difference here is that the search and measurement parameters are relatively modest. DECT handsets work over just ten channels within a 20-MHz channel allocation. WiFi measurement and access algorithms at 2.4 and 5 GHz are similarly comparatively straightforward. Most 2.4-GHz systems only use three 22-MHz channels across the 80-MHz bandwidth allocation.

In cellular networks, the measurement options are theoretically broad but practically narrow. Channel access will usually be in the same band and often on the same channel (a simple shift to another time slot).

In DECT and WiFi and cellular networks, the devices are measuring identical signal waveforms to the waveforms they need to demodulate. White-space devices have to detect different wave forms that may be similar or different to the waveforms they need to demodulate. This is a significantly more complex task.

15.16 An Implied Need to Rethink the White-Space Space

This suggests that white-space devices as presently conceived will have poor transmission efficiency that will translate into low uplink data rates and/or limited uplink range and a limited RF uplink power budget. The limited power budget will be compounded by a relatively high measurement and signalling overhead that will reduce session duty cycles. The devices will not work as well as cellular devices and/or WiFi and WiMax devices and will cost more due to smaller market volumes. This would tend to suggest that present white-space business models are fragile at best.

However, as white-space proponents point out, hundreds of MHz of spectrum are unused or under used at certain times at certain places. This white-space spectrum has a social and economic value. The problem is that white-space devices as presently conceived are not an efficient mechanism for realising that value. The answer may be to encourage the broadcasters (the NAB in the US the EBU in Europe) to develop a white-space device specification that could be integrated with existing (DVB) and planned (ATSC) portable TV specifications.

This would have several benefits.

The present dispute over potential interference problems would be resolved. A new business model could be developed based on a closer coupling between terrestrial TV and two-way wireless internet access. It would provide a broader industrial base over which RF development costs could be more efficiently amortised.

It would be even better if the cellular community could be encouraged to work with the broadcast community on a common white-space standard to provide compatibility with ATSC and DVB and LTE 700- and 800-MHz devices.

Similarly, the future of terrestrial TV will be dependent on building closer relationships with cellular service providers. TV transmissions from cellular infill sites are for example the only credible way forward if ATSC-based portable TV is ever going to work.

15.17 White-Space White House

So in theory, white-space devices are a great idea – an opportunity to realise value from presently unused or underused spectrum.

As such, the 4th November decision to make a swathe of new unlicensed spectrum available for innovative services must still look superficially attractive to the new US administration – a populist policy with tangible social political and economic benefits. In practice, these practical benefits can only be realised if devices can be developed to work in most, if not all, global markets. This is politically, commercially and technically challenging and implies a need to work with rather than against the broadcast and cellular community and to work across national and international boundaries.

In common with licensed spectrum, unlicensed spectrum incurs access costs that have to be recovered from access charges or from other sources such as tax revenue or market subsidy. To quote President Roosevelt ‘We have never realised before our interdependence on each other’.

Seventy five years on and faced with similar recessionary pressures we should recognise that interdependency implies a need for to explore and exploit collaborative rather than competitive market opportunities and avoid the unnecessary and presently unsustainable costs introduced by conflicting market interests.

15.18 LTE TV

This brings us to LTE TV.

As documented earlier in this chapter, considerable time, money and effort has been spent on mobile TV standardisation and network and handset development with to date no commensurate fiscal return.

It has been difficult to realise a return on investment partly due to competing standards, (DVB H, Media FLO, ISDBT, DMB DAB and MBMS, mobile broadcast and multicast service) and partly due to network and handset development limitations and performance constraints compounded by regional and national differences in mobile-TV frequency-band allocation and incompatible standards.

Although mobile TV has been successful in some markets, for example Japan, it has failed to achieve global scale and in most markets there has been a marked reluctance by consumers to pay for mobile TV content.

The longer-term economics and effectiveness of hybrid satellite and cellular networks such as DVB SH in Europe or ATC (ancillary terrestrial component) hybrid satellite and terrestrial networks in the US also remain unproven, a topic addressed in the following chapter.

This might lead one to question the merit of introducing yet another mobile TV standard to service a global market that is presently nonexistent. However, the inclusion of multicast broadcast single-frequency networks, MBSFN, as a work item in present and future LTE releases suggests that there is still a measure of vendor and operator confidence that mobile TV remains a worthwhile investment opportunity.

Remarkably, it may be and MBSFN could be used to realise additional revenue and improve returns on existing spectral and network investment both for the cellular and broadcast community.

15.18.1 The Difference Between Standard LTE and MBSFN – The Technical Detail

MBSFN is based on 7.5-kHz subcarrier spacing rather than the 15-kHz used in standard LTE. This doubles the symbol length from 66.7 microseconds to 133.4 microseconds and allows the cyclic prefix to be increased from 4.69 microseconds to 33.33 microseconds.

The cyclic prefix provides a time-domain guard band between symbols to compensate for delay spread in the radio channel. A 4.69-microsecond cyclic prefix allows a delay spread of 1.5 km; a 33.33 microsecond cyclic prefix allows a delay spread of 10 kilometres. The delay spread is the difference in path length caused by multipath in cellular networks. In SFN broadcast networks, the delay spread is also a function of receiving the same signal from more than one transmitter.

The 33-microsecond cyclic prefix used in MBSFN means that TV transmissions can be broadcast simultaneously from multiple node B cellular base stations without causing intercarrier symbol interference in the receiver.

Table 15.4 compares DRM, DAB, DVB broadcast networks with LTE and LTE MBSFN.

Table 15.4 Broadcast technology comparisons

The number of subcarriers defined for DVB T and LTE excludes edge-of-band subcarriers used to provide a frequency domain guard band. For example the 8 K DVB- T multiplex has 6817 one-kHz subcarriers nestled within 8 MHz (8 k times one kHz) channel spacing.

In practice, a number of DVB-T networks were implemented as 2-k rather than 8-k networks due to receiver chip set limitations, for example in the UK where initial implementation was realised in 1997. This prevented the implementation of a (more spectrally efficient) single-frequency DVB-T network. Although the LTE MBSFN is described as a single frequency network, it could be deployed as a cluster of SFN enhanced node B base stations with frequency reuse from cluster to cluster.

Present standards work is focused on either FDD or TDD implementation, known as downlink optimised broadcasting within existing cellular allocations, for example the TDD bands within Band I at 2 GHz.

15.18.2 MBSFN Investment Incentive – Cost Saving and Revenue Opportunities

This, however, fails to capitalise on some potential major cost saving and enhanced revenue opportunities that together could provide the required incentives for MBSFN investment. These can be identified by analysing the problems that the broadcasting and cellular industry have to solve, some of which we have identified earlier.

As can be seen from the table, the 8-k subcarrier OFDM used in DVB-T allows large cells to be deployed. This means that DVB-T broadcast networks can be very cost economic. However, to provide adequate coverage, low-power infill sites have to be deployed that add capital cost and running cost and reduce broadcast spectral efficiency.

Although this provides adequate coverage at an adequate link budget for TVs (including high-definition TVs) with a roof-top aerial, there is generally not enough signal strength to support portable receivers. Portable TV receivers have been the Cinderella of the TV industry for many years but the availability of lap tops with DVB or ATSC modems will potentially result in a much broader market reach. The top of the range high-definition Sony Vaio16 is one example. However, the signal levels in most countries are not sufficient to deliver a sufficiently consistent user experience to support mass-market adoption. Deployment of high-density SFNs based on existing cellular infrastructure would overcome this problem. The difficulty is that portable DVB devices do not work well unless connected to an outdoor aerial or a satellite dish or cable. These limit the portability of the device.

A roof-mounted antenna has about 12 dBi of directional gain. A built-in broadband TV antenna in a lap top will usually have negative gain, typically −7 dBi and will be looking for a signal severely attenuated by height loss (10 dB), additional building penetration loss (8 dB) and location variation due to standing waves (10 dB), resulting in a received signal up to 47 dB lower than the value planned for fixed outdoor reception. It would be prohibitively expensive to increase flux levels from the existing terrestrial broadcast networks to support acceptably ubiquitous portable TV reception, particularly portable HDTV reception.

Cellular base stations, particularly cellular base stations designed to support the 700- and 800-MHz cellular bands could, however, function as relay repeaters and could provide a cost-effective alternative to dedicated TV repeaters. Cellular handsets or at least some cellular handsets and mobile broadband devices including lap tops with embedded LTE modems or dongles will have 700- and 800-MHz LTE transceiver functionality. It is therefore not inconceivable to increase LTE 700- and 800-MHz receiver bandwidth to accommodate DVB T transmissions.

Operators might question the point of including DVB reception in mobile broadband devices particularly given that many of the services are available free to air. This, however, ignores the indirect revenue gain that is potentially achievable from coupling free to air TV reception with two-way mobile broadband and is pragmatically probably the only way in which cellular operators will ever achieve a return from 700- and 800-MHz cellular infrastructure and spectral investment.

The ability to receive broadcast TV off air would reduce traffic load on LTE data networks delivering a capacity gain. Hill-top TV transmitters could also be used to provide extended cellular coverage. An upper limit cell radius of 100 km is included in the LTE specification.

15.19 Summary

The proximity of cellular receive and transmit bands at 700 and 800 MHz to UHF broadcast TV channels opens up the possibility of supporting digital TV transmission via cellular base stations and digital TV reception in handsets and mobile broadband devices without significant additional hardware cost. Rebroadcasting of national and local TV from cellular transmitters would allow HDTV to be delivered to portable receivers without the need to connect portable devices to a fixed antenna.

Cellular site and hardware costs including backhaul overheads could be amortised across cellular and broadcast services. There would be no need to ring fence or repurpose TDD bandwidth at 2 GHz for broadcasting. These channels could be used as originally intended for mobile broadband access.

The coupling of local and national linear TV broadcasting with mobile broadband would unlock new direct and indirect revenue streams both for the broadcasters and cellular operators and would provide the basis for developing innovative mass-market consumer electronic products that could be clearly differentiated from present product and service offerings. The combination of the additional revenues realisable from a more broadly based user experience combined with more broadly amortised costs could significantly improve profitability for all involved parties.

15.20 TV or not TV – That is the Question – What is the Answer?

The launch of Apple's iPad has generated countless column inches of industry comment but no one seems to have quite nailed the answer as to what users will do with the device. The consensus seems to be that iPads will change the way that media is consumed and information is accessed on the move.

However, the form factor of the device, smaller than a lap top but larger than a smart phone, suggests a sit-down or sit-back experience – very different from how mobile phones are used today. If this is true, then some form of TV connectivity would seem sensible. This could be IPTV over WiFi or cellular but equally could include ATSC and/or DVB T.

It has become fashionable to dismiss terrestrial television as an increasingly irrelevant delivery medium, but in practice terrestrial broadcasters still hold a number of trump cards in terms of linear and nonlinear content, delivery bandwidth and political, social and business collateral.

The generic argument is that terrestrial TV will continue to be one of the most economic and power-efficient delivery platforms for content and information and that the use of MPEG and JPEG encoding and compression schemes at the application layer and IP protocols at the transport layer, if combined with closer physical bearer integration, would result in a closer coupling of these traditionally separate delivery platforms.

The BBC are presently promoting a new standard to be known as DVBT2M for mobile phones and have a mobile iPlayer offer which at Summer 2011 was getting 8 million requests per month. Attention was also being given to developing dual-screen delivery formats with a tablet displaying supporting content to a program being shown on TV.17 Commercially, a TV license fee could be a way of collecting money for digital content rights.

We suggested that LTE could potentially provide the basis for this closer integration and that this process of technology and engineering convergence would translate into new market and business opportunities. However, in order to maintain its ubiquity as a delivery medium, terrestrial digital TV would need to be received by portable terminals used indoors without an aerial. This implies a very substantial increase in signal strength and/or an improved link budget of the order of several tens of dBs. High-definition broadcasting would need another 10 to 12 dB, though whether this would ever be worthwhile for portable form factor devices is open to debate.

Either way, the only practical way in which acceptably consistent TV coverage could be realised would be to rebroadcast the DVB T or ATSC multiplex via local cellular transmitters or via inbuilding repeaters. Alternatively, TV signals could be piped down from a roof-top aerial and rebroadcast from an indoor femtocell. However, even with signal rebroadcasting, portable receivers would benefit from having improved sensitivity and selectivity. Several factors now make this a more plausible scenario.

Cellular base-station transmitters in the 700-MHz (US) and 800-MHz (European) UHF band need to be sufficiently linear to transmit the cellular OFDM signal with minimal distortion. It is therefore not unrealistic to consider adding ATSC or DVB T broadcast to the TX chain.

Cellular handsets and portable lap tops will have to add 700- and 800-MHz UHF LTE TX/RX functionality. It is therefore not unrealistic to consider adding DVB T and ATSC receive functionality to these devices. Indeed, it could be argued that extended RF connectivity might be a precondition of mass-market adoption for the emerging crossover form factor devices that combine traditional lap top, net book and smart-phone functionality, of which the iPad is one example.

However, there are some practical RF design and performance issues that have to be addressed before these products can be realised. Substantially these are determined by the characteristics of the UHF band and the need for these devices to provide cellular connectivity in the 850/ 900, 1800, 1900, 2100 and 2600 MHz bands.

So, for example, it might seem sensible to have an ATSC and DVB T receiver that worked across the whole UHF band from 470 to 862 MHz and a transceiver that could cover the cellular bands from the bottom of the US 700 MHz band at 698 MHz to the top of the proposed European UHF cellular band at 862 MHz.

In practice, such a device, even with adaptive matching and tracking filters, would have unacceptable sensitivity and selectivity on the RX path, poor transmission efficiency on the TX path and the dynamic range needed would incur an unacceptable power drain. Adding other bands introduces additional switch paths that introduce insertion loss and poor isolation.

The link budget gain of working at 700 MHz or 800 MHz is about 8 or 9 dB when compared to 1800 or 1900 MHz, and rather more when compared to 2600 MHz. This is a function of reduced propagation loss. Inbuilding penetration in particular should be notably better and should translate into improved coverage and/or higher average data throughput rates.

However, these gains are academic if offset by RX and TX efficiency losses in the user's device. The form factor of an iPad or net book or lap top provides some extra space but this advantage can be offset by higher noise floors generated by display drivers or other proximate processor activity.

Additionally, the decision has to be made as to whether the TV receiver should be expected to work simultaneously with broadband cellular. This would mean additional parallel processing but more importantly would also require careful implementation to avoid desensitisation of the TV receiver by the LTE transmit path.

Commercially, it would seem that this is a problem worth solving. Online connectivity viewed in parallel with broadcast content could provide the basis for a whole new generation of interactive two-way content coupled with real-time peer-to-peer or peer-to-multipeer communication.

Technically, this is a nontrivial challenge. The assumption is often made that this kind of dual functionality will become easier to deliver as software-defined radios and/or cognitive radios become more pervasive.

Unfortunately, this is only partly true. Cellular transceivers today still route band-specific signals through band-specific filters and band specific power and low-noise amplifier components. All of these devices have individual matching components.

Getting acceptable performance from a TV receiver integrated with a cellular multiband transceiver would probably also require discrete low-, mid- and high-band signal paths if the whole UHF band had to be covered. Even if digital TV is eradicated globally from the upper UHF bands then a single receive chain would still seem unlikely for a device capable of receiving DVB T and ATSC broadcast.

The alternative is to manufacture devices that are regionally or nationally or band specific, but this frustrates scale economy and prevents products being used universally. The holy grail is to produce high-Q tracking filters that can be channel rather than band specific combined with active components that can maintain their efficiency over extended operational bandwidths, but these devices remain an ambition rather than a present reality, at least in present transceiver designs.

But incremental progress is being made. The industry news flow tends to concentrate on baseband devices but in practice and as stated in earlier chapters the advances being made in new passive and active materials including RF MEMS, silicon on sapphire and BST-based devices are likely to be at least as important.

The future of mobile TV will therefore be dependent on materials and manufacturing innovation.

15.21 And Finally the Issue of Potential Spectral Litigation

Parts of this chapter have highlighted the coexistence issues implicit in the WRC 2007 decisions. In this final section we review the spectral litigation risks that are arising as terrestrial broadcast, cellular networks and white-space devices compete for UHF spectrum and why these risks may increase over time.

An important objective of any regulatory process is to avoid or minimise the impact of spectral ownership disputes. An increase in the number of disputes and related costs implies a failure of the regulatory process.

The US 700-MHz auction and its aftermath from 2007 to 2011 provides an example. The context is a block of spectrum called A block that a number of mobile broadband operators bid for when it became available postdigital switchover. A block is immediately adjacent to the top of the TV broadcast band. The bidding entities including Cellular South and US Cellular supported by Metro PCS and Cox Communications alleged that AT and T and Verizon Wireless misused their market dominance to prevent vendors from making user equipment and base stations available capable of accessing the spectrum. The litigation and advocacy work was done through a 700 MHz Block A Good Faith Purchasers Alliance.18

Although Verizon had invested $2.57 billion in A block there were considerable performance costs implicit in adding the band to either Verizon or AT and T base station and user equipment. AT and T had no technical or commercial incentive to cover A Block and were more concerned as to how they repurposed and integrated the TDD bandwidth at channel 55 and 56 acquired from Qualcomm.

For A block owners you could say this was either a case of caveat emptor or an example of the regulator bringing spectrum to market that was not fit for purpose. A regulatory environment that fails to take into account engineering reality inevitably results in the destruction of industry and end user value, trading short-term treasury gain against long-term loss to the national economy. In this case a $50 billion industry investment was compromised by the physical limitations of a few 50-cent components in the front end of a mobile broadband transceiver.

15.21.1 Other Prior and Present Examples

One would have thought the FCC would have learnt from prior examples. In February 200219 a decision was taken to make bandwidth available between 3.1 and 10.6 GHz for ultrawideband technologies. The operational bandwidth was not acceptable in other countries. Other regulators responded with their own variations, thereby invalidating any potential business case for UWB. A decision to let the market decide on technology produced two competing standards that created additional market fragmentation. As suggested earlier, similar mistakes are being made on white-space legislation presently being introduced with minimal consideration for global spectral implications or user equipment RF development economics.

15.21.2 The Cost of Technology Neutrality

If the US 700-MHz band was in some way unique in terms of regulatory approach then this would probably not matter that much. However, the disconnect between regulatory policy and physics is more generally pervasive. A doctrine of technology neutrality or at least an absence of technology direction compounds the problem and generally results in a loss of spectral efficiency. This in turn increases the risk of spectral litigation. Mixing CDMA, HSPA, WiMax and GSM in the 1900 or 850 bands is one example.

One of the reasons we have regulators is to arbitrate between competing spectral property interests. This extends both to interindustry and intraindustry disputes. The amount of litigation could thus be viewed as a proxy measure of the individual and combined effectiveness of the regulatory process. In the US the 700 MHz spectral disputes are presently confined to interoperator disputes, but in years to come could easily involve the broadcast industry challenging interference from A block mobile broadband TX and public protection agencies challenging interference from Verizon upper block mobile TX. Both are the direct result of a poorly implemented band plan. Similar disputes are likely in L band and S band as mobile operators realise the potentially adverse fiscal and operational impact implicit in the present repurposing of mobile satellite spectrum for terrestrial use. In Europe, similar disputes are simmering over the allocation and auction processes being proposed or implemented for the 2.6-GHz extension band linked directly or indirectly to 800-MHz deployments. Operators generally are concerned that their spectral holdings should be balanced across all available bands or rebalanced to address legacy allocations that in retrospect can be regarded as being competitively unfair.

This increase in tension can be ascribed to a number of factors.

Additional bands are being allocated with terms and obligations that differ from previous allocations. Each additional band introduces additional insertion loss and reduced isolation through the switch paths and filters in user equipment. This will result in unforeseen and unexpected interference issues but also reduces spectral efficiency.

Multiple technologies are being introduced into legacy bandwidth with minimal consideration as to the likely impact on proximate bandwidth and other user communities. This will result in unforeseen and unexpected interference issues, but also reduces spectral efficiency.

Existing bands are being extended to accommodate these multiple technologies. The 10-MHz extension to the 850 band in the US is an example. This will result in unforeseen and unexpected interference issues but also reduces spectral efficiency.

Channel bonding is being introduced to meet an assumed market demand for peak data rates of up to 1 Gbps. This will result in unforeseen and unexpected interference issues but also reduces spectral efficiency.

MIMO is being introduced in parallel with channel bonding on similar assumptions. MIMO may result in unforeseen and unexpected interference issues and will certainly reduce average throughput rates, effectively a reduction in system efficiency.

Operational requirements are being imposed on operators without due consideration of the spectral implications. Mandatory E911 support for example fails to take into account that the second harmonic of 787.5 MHz, between C band and D band mobile transmit in the US 700 MHz band falls directly on the GPS receive frequency in L band at 1575 MHz. Terrestrial use of L band may also compromise GPS front-end receiver performance, a problem that Light Squared had to contend with through 2011.

All of the above increase performance uncertainty both in terms of band-to-band performance, within-band channel-to-channel performance, sensitivity to hand and head effects and changing operational conditions including temperature and battery charge state. This makes quality of user experience service level agreements harder to model and manage. All of the above are therefore likely to increase rather than decrease ‘within industry’ interoperator disputes over spectral ownership rights, compounded by the combination of a reduction in spectral efficiency coupled with an increase in operator-to-operator, user-to-user interference. All of the above are also likely to increase ‘between industry’ interindustry disputes over interference and spectral ownership including disputes between the mobile broadband industry, the broadcasting industry, the mobile satellite industry, public safety radio industry and cable industry (set-top box interference).

An adversarial approach to future spectral allocation, for example the repurposing of C band between 3400 and 3800 MHz for FDD/TDD bands 22/41 and 23/42 makes it less likely that these separate industries will be able to work together to resolve technical issues. Interindustry cooperation at the technical level is already complicated and frustrated by competition policy. TDD/FDD coexistence will also introduce additional system level complexity which will be technically and commercially hard to resolve.

The problem is compounded by a false belief by the regulatory community that all that is needed is to define spectral transmission masks. This ignores the growing need to define and enforce receiver selectivity and dynamic range.

It is also increasingly inappropriate for the FCC to assume that the world should follow US policy. In terms of mobile subscriber market size, China and India are both more than twice as large as the USA. Given that 48% of all connections are now in Asia it should be obvious that the US now lacks sufficient economy of scale to support a nationally specific band plan and/or a nationally specific technology mix. Operator-specific band plans and operator-specific technology solutions within the US market make even less economic sense.

China is arguably the only market with sufficient volume and local design and manufacturing capability to be able to support national- or operator-specific band plans and technologies, but even this does not mean that such policies make any kind of long-term economic sense. The problem is that in all markets, China, the US and Rest of the World, spectral allocation and auction policy is based on economic modelling that fails to capture relevant cost and value dynamics. So as a reminder of what is in effect the ongoing narrative of this book there are, we would argue, five separate but coupled domains that need to be analysed:

15.22 Technology Economics

This is the area most directly coupled to the standards process. Our contention here is that false market ambitions, specifically high peak data rates per user are introducing unnecessary complexity and compromising system performance to the point at which performance loss/economic loss litigation becomes likely. For example, it might be expected that at some stage operators will need to protect network performance by introducing pass/fail margins into conformance testing. This is valid and understandable but will highlight vendors who are presently shipping products that only meet conformance performance requirements under an unrealistically narrow range of operational conditions.

Note that some devices can now take well over 1000 hours to go through conformance test – performance testing potentially takes longer. This is a cost that no one wants to absorb. This will result in an increase in litigation cost.

15.23 Engineering Economics

This is the area most closely coupled to the spectrum allocation and auction process. Our contention here is that false policy objectives, specifically the maximisation of short-term gains for national treasuries over longer-term economic gains have resulted in spectrum sales that ignore the laws of physics. The problem is compounded by present standards policy that compromises spectral efficiency. Caveat emptor may provide a measure of protection but an increase in litigation seems to be an inevitable outcome of the present policy approach.

15.24 Market Economics

The problem here is that standards and regulatory policy has failed to adapt to the change in relative market importance between Asia and the rest of the world. If subscale markets, for example the US, continue to pursue a nationally specific standards and regulatory agenda then it is understandable that Asian markets, particularly China, will want to do the same. The result will be an increase in litigation and more aggressive protectionist legislation.

15.25 Business Economics

This is the problem of the 50 cent components compromising the $50 billion investment. The fact that most RF components don't scale has made it hard for the RF component industry to deliver products that can meet associated performance requirements. Newly dominant markets are understandably using their market leverage to extract significant performance promises from this underfunded and underresourced sector of the industry. The result will be an increase in Chapter 11 filing that will further inhibit RF innovation and investment.

15.26 Political Economics

Back to where we started. Regulators will become increasingly exposed to litigation from entities who have bid for spectrum that is not fit for purpose or compromised either by poorly executed standards policy or, equally damaging, by the absence of a standards policy. The problem is compounded by regulatory policies that fail to take account of basic engineering reality and present RF component limitations and R and D constraints.

15.27 Remedies

The standards process would benefit from being less focused on an assumed need for high peak data rates and more directly focused on delivery cost economics. Standards and spectral policy need to be more closely coupled and engineering cost needs to be directly factored into band allocation and auction policy. Some mechanism must also be devised to encourage more effective collaboration between different sectors of the industry, mobile broadband, public safety, broadcasting and the mobile satellite sector and closer economic integration of wireless, cable, copper and fibre delivery systems.

Regulators are always under pressure to promote country-specific solutions to benefit their national stakeholders. The aim is laudable but history shows that this approach can be disastrous. In the early 1990s, Japan, at that time the world's second largest national economy – went its own way with mobile cellular. As a result, Japanese manufacturers were left floundering, unable to service both a quirky domestic market and a brutally competitive global market. In the USA, Block A Purchasers rail against manufacturers who ignore their special needs. But manufacturers who wish to compete in the global market cannot afford to divert expensive engineering effort to rescue spectrum owners left stranded by poor regulation.

In this context it is important to differentiate between constructive tension, the mechanism by which economic progress is achieved through efficiently ordered market competition and destructive tension, a general outcome of a poorly ordered market and misapplied regulatory intent.

A similar differentiation can be made between constructive litigation, the valid arbitration of competing market interest and destructive litigation, the outcome of poorly implemented spectral and standards policy. Destructive litigation is unnecessary and wasteful but can be avoided provided potential causes are identified at an early stage – a challenge and opportunity for the regulatory community.

1 Before being appointed as the BBC's first engineer, the multitalented Peter Pendleton Eckersley had been the producer, writer, and presenter of the first experimental half-hour programme that the Marconi company was licensed to broadcast each week from a studio in a former army hut in Writtle near Chelmsford. The programme was launched ‘on air’ on St Valentine's Day, 1922. The Captain had a gift for improvising on microphone. He became Britain's first radio star with an instantly recognisable call sign ‘Two-Emma-Toc, Writtle testing’ http://www.marconicalling.com/museum/html/people/people-i=17.html.

2 The author is unsure of the provenance of the 2LO description. Possibly the transmitter had two local oscillators or perhaps the call sign was meant to be Hello Hello- answers on a postcard please.

3 Thanks to Richard Lambley of Land Mobile Radio Magazine for sharing some of his early childhood memories of living close to the Daventry transmitter.

4 www.drm.org.

5 http://www.shortwave.org/.

6 http://www.sony.net/SonyInfo/CorporateInfo/History/sonyhistory-b.html.
http://www.sony.net/SonyInfo/CorporateInfo/History/.

7 http://www.nab.org/.

8 http://www.radioandtelly.co.uk/visualradio.html.

9 For example, a product from Coding Technologies subsequently acquired by the Dolby Corporation www.dolby.com/professional/technology/broadcast/he-aac-dolby-metadata.html.

10 http://www.drm.org/products.

11 http://www.worlddab.org/country_information.

12 http://www.pure.com/.

13 http://www.dibeg.org/.

14 http://www.wirelessinnovationalliance.org/.

15 http://www.adaptrum.com/.

16 http://www.sony.co.uk/product/vn-z-series.

17 Stephen Baily, General Manager BBC R and D, presenting at Cambridge Wireless conference in Cambridge 27 June 2011.

18 http://ecfsdocs.fcc.gov/filings/2010/04/30/6015588771.html.

19 http://www.fcc.gov/Bureaus/Engineering_Technology/News_Releases/2002/nret0203.html.