Written on 10:33 AM by ooe
Allocations made for narrowband wireless local loop applications vary from country to country but in most cases they are in the 2 to 4 GHz band. Individual operators have been granted spectrum allocations of approximately 10 to 50 MHz within these bands, which is relatively small. There is increasing interest in using these frequencies for mid-band systems offering data services as well as voice. The advantage over higher frequencies is that these systems can work over greater distances and use less sophisticated and hence less expensive components. However, there is inherently less spectrum available and it is in demand for future wide area mobile systems.
Broadband systems require more spectrum and the allocations that have been made, or are being considered, lie within the 10 to 40 GHz band. indicates the major allocations in the US and Europe. Shown are both issued and proposed licences, with the UK shown separately from Europe to illustrate that even within one region there are still national differences. It can be seen that a number of past licences issued for fixed wireless access in the UK are under consideration for reallocation, due to lack of use.
|
Written on 10:33 AM by ooe
A common licence condition for block allocations specifies whether or not spectrum trading is permitted. If it is, the winning operator may subsequently transfer their licence to another party for instance if they wish to sell it or merge their licence holdings with those of another operator. If trading is prohibited, an operator is not at liberty to sell their licence and if they no longer require it, it must be returned to the government for resale.
|
Written on 10:33 AM by ooe
An alternative method of awarding block licences is commonly known as a ‘beauty contest’. In most cases, an annual fee is charged for the licences, but this is an administrative charge, much lower than the likely auction price. The award is based not on which operator is willing to pay the most, but rather the operator who undertakes to make best use of the spectrum.
In a similar way to the minimum permitted performance licence conditions often imposed on auction winning operators, the beauty contest attempts to ensure that the government's original objectives in issuing the spectrum are met. While an auction process limits itself to specifying a minimum set of desirable parameters that the resultant network must exhibit, the beauty contest award decision is given to the operator whose proposed network offers the most desirable features. Exactly how these desirability criteria are judged will again depend on the original objectives.
One obvious flaw with this process is that an operator could over-promise in an attempt to secure a licence. For this reason, the promises usually form the basis of the winner's licence conditions. Additionally it is common for a detailed business plan to be required which shows achievable forecasts for revenue and expenditure.
|
Written on 10:32 AM by ooe
Block spectrum allocations are often awarded using a competitive auction. Holders of such licences can use the spectrum wherever and as often as they like, within a given country or region. Such awards usually last for a set number of years, with the issuing authority then claiming back the spectrum for reissue or an alternative use.
There have been many successful auctions for FWA spectrum around the world with considerable income being generated, for instance the 1998 US LMDS auctions (spectrum in the 28–31 GHz range) raised $600 million. However, the UK 28 GHz auction held in 2000 illustrates the risk to governments of using this process – 26 of the 42 licences offered failed to attract any bids, leaving large parts of the UK without any broadband FWA operators to compete with existing access providers.
Other risks occur depending on the government's original motivation in offering the spectrum for FWA usage. In many cases the objective is to increase competition in the access market by using the potentially low cost of entry offered by wireless networks compared to wired networks. One possibility is that the licence winner will ‘cherry pick’, only offering service in the highest profit areas. This can mean that only the most densely populated business districts will obtain the benefit of increased access competition. Alternatively, the spectrum might be used for infrastructure connections, such as point-to-point backhaul connections within the operator's network, rather than for its intended application in access networks. Outcomes such as these can be guarded against by introducing licence conditions such as minimum coverage areas that must be achieved within a given time and strict definitions on the permitted use of the spectrum.
|
Written on 10:32 AM by ooe
This type of spectrum licence charging tends to be applied to point-to-point systems, where dedicated spectrum is required for each customer connection. The operator must apply for the spectrum required for each radio link and the licence obtained is only valid for that frequency in that particular location. Generally, if the link is removed, the spectrum licence must be handed back to the issuing government. Charging may have temporal, capacity, usage and geographic elements.
|
Written on 10:32 AM by ooe
Some spectrum can be used without payment to the government that owns it. Some such spectrum still requires the user to obtain a licence prior to use and some is set aside for use without any permissions at all, provided certain technical constraints are met. The main example of the latter are the internationally agreed ISM (industrial, scientific and medical) bands, this spectrum often being known as ‘licence-exempt’. Whilst the technical parameters vary by country, the aim is the same – to make spectrum available that may be used on an ad hoc basis without need for prior permission. Example applications include microwave ovens and industrial heaters, wireless LANs and remote control units. Another emerging use of these bands in many countries is for commercial fixed wireless access networks, often focused on delivering data services and based on existing wireless LAN technology.
Within the UK there is an unusual constraint on these bands which limits their licence-exempt use to ‘private self-provided communications’ in effect precluding their use for commercial access networks without a licence. Thus far the government has only issued one commercial use licence per region within the UK.
|
Written on 10:32 AM by ooe
All fixed wireless access systems require a portion of the radio spectrum in which to operate. In almost all cases, a licence to use this spectrum must be purchased from the appropriate government agency before the system can be operated or even planned. The size of the radio spectrum is in theory infinite, extending from zero hertz through progressively high frequencies to light, x-rays and beyond. In practice only a very limited part of this spectrum is usable for commercial access systems. Demand for these frequencies is huge and growing all the time with competition to fixed terrestrial access systems coming from uses such as satellite communications, radio astronomy, industrial applications, emergency services, television broadcasting, telemetry, mobile communications, military and medical applications.
The large and ever changing appetite for such a scarce resource leads to a requirement for careful management on national, continental and global fronts, which in turn limits the amount of spectrum available for fixed wireless access use at any point in time. Suitable parts of the radio spectrum, or bands, are usually made available by a national government when agreement is reached with all other previous or potential users of the spectrum and there is judged to be significant interest from operators.
Would-be fixed wireless access operators therefore face the difficult situation where their ability to deploy a network is dependent on being able to obtain a licence for suitable spectrum from the local government, who may or may not make such licences available at some point in future. This makes spectrum an absolutely essential ingredient in any planning or operation of wireless networks.
|
Written on 10:31 AM by ooe
Mankind has been communicating via radio signals for a very long time, between neighbouring buildings, across oceans and even from one planet to another. Theoretically such technology can be used for the last mile of a telecommunications network to connect its customers. In practice, however, it is not well suited to such use and it is only over the last decade or two that tailored solutions have emerged, allowing large scale wireless access to become economically viable.
Wireless access splits into two classes – fixed and mobile. Fixed systems are known by many names including fixed wireless access (FWA), wireless local loop (WLL), multimedia wireless systems (MWS), broadband wireless access (BWA) and local multipoint distribution systems (LMDS). While mobile systems allow users to roam around the network, in fixed systems the radio units are permanently mounted in the same way as copper and optical fibre.
Using radio in the access network presents a unique set of engineering challenges, especially if service quality is to rival wired delivery. Securing suitable spectrum is a prerequisite to establishing a network. The transmission medium presents many challenges which are outside the direct control of the network operator, and those that are of particular importance for fixed applications have to be considered when selecting radio equipment. Careful planning of the network is required to ensure it meets initial and longer term market requirements and the initial technology and architecture choices are key to achieving success. When all these aspects have been considered, a network operator can assess the likely costs and revenue of a wireless system and can answer a more fundamental question – is wireless the best choice of access technology for this application? This chapter will explore each of these issues in more depth.
|
Written on 2:55 PM by ooe
The biggest barrier associated with deploying FTTH may not be due to technical or cost constraints but due to customer demand. However much the infrastructure and transmission equipment is developed and cost optimised, unless sufficient customers want the services that FTTH can deliver it will not be commercially viable. Most customers are interested only in the services that they have access to and not the network which provides them. As discussed earlier, services such as software delivery and High Definition TV require bandwidths that would greatly stretch the capabilities of most, if not all, technology options except FTTH. To drive FTTH deployment, one or more high bandwidth killer applications need to be developed which will increase the revenue per customer. Until such applications are found FTTH deployment is likely to continue at a more sedate, though no less stoppable, pace.
-
The deployment of FTTH is largely constrained by commercial and not technology issues. What is needed is both a reduction in the cost of deployment and an increase in the perceived revenue that may be achieved. Once an acceptable profit margin can be seen the speed of deployment will accelerate. This may fuel further cost reductions associated with volume production which may sustain the acceleration.
-
The development of a killer application, for which many customers would be prepared to pay a premium price, could generate a compelling commercial argument for deploying FTTH more quickly and as widespread as operationally possible.
-
Even with a concerted commitment, universal deployment of FTTH across the whole of the UK is unlikely to be achieved in less than five to ten years even if it were started tomorrow.
-
Current effort of the Broadband Network Engineering Network Modelling team is concentrated on understanding the costs involved and seeking ways to reduce these. Optimisation of the distribution network is seen as crucial for cost per customer reduction. Blown fibre is seen as a key enabler for FTTH.
-
Infrastructure cost per customer is highly dependent upon customer take up. This tails off at 50%, so ideally active marketing should be undertaken to achieve take up of at least 50% in most geographical areas. 11
|
Written on 2:55 PM by ooe
Labour costs, whether for direct employees or external contracts, form a significant part of the total deployment costs for any network. For FTTH a large amount of external network installation will be involved. The bulk of costs are attributable to this installation work, including duct build, cable installation and jointing/configuring cables. It is, however, important to consider the wider labour requirements associated with the whole lifetime of the FTTH network as shown in Figure 10.12. Not only does this add to the labour costs, it demonstrates the number of different functions required; these functions are usually carried out by different personnel.
Reduction of labour costs could be achieved by one or more of the following:
-
making the best use of the existing infrastructure (particularly duct network);
-
optimising network design;
-
more efficient practices;
-
more efficient processes;
-
better utilisation of personnel;
-
deskilling practices to enable lower paid personnel to be used.
|
Written on 2:53 PM by ooe
It would be technically possible to use current commercially available fibre infrastructure for FTTH. However, much of this infrastructure has been designed for dedicated fibre feeds to business customers. It has been realised that this plant is likely to be over specified for wide scale FTTH provision. Similarly operational practices have been developed for providing fibre feeds to individual or small groups of customers and are not likely to be optimum for FTTH. It can be concluded that cost optimisation is required for both infrastructure stores and labour before FTTH could be seriously considered. It should be noted that a significant reduction in transmission system costs is also required. This is discussed elsewhere [2].
The largest single factor that affects the cost per customer of installed FTTH infrastructure is customer take up. High customer take up in a geographical area enables greater sharing of network segments and consequently greater sharing of the costs associated with those parts of the network. The effect for the town of Ipswich is demonstrated in Figure 10.11 for both a 32 way split PON and a dedicated fibre network. It can be seen that the cost reduction tails off above 50%. For a PON network the infrastructure cost per customer is halved by increasing the customer take up from 10 to 50%.
|
Written on 2:53 PM by ooe
Detailed cost modelling of FTTH infrastructure has been performed for the town of Ipswich. This area includes five exchanges which serve a total of approximately 70 000 customers. The average cost per customer connected has been estimated by dimensioning a fibre network for an expected level of customer take up. The results of the modelling activity highlight the following cost trends.
Figures 10.9 and 10.10 show simple summary charts indicating the network infrastructure cost breakdowns for PON and point-to-point networks to serve an expected customer penetration of 20% (with no demographics applied). In both cases it can be seen that the distribution side of the network contributes the greatest cost per customer. This is because this network segment is shared between the least number of customers. Optimisation of the distribution network is therefore crucial in order to reduce overall costs. For both PON and dedicated fibre networks, labour costs contribute approximately 50% of the total infrastructure bill. A key part of this process is using NetMod to carry out cost sensitivity analysis and hence start to develop planning rules that ensure cost effective network deployment.
Detailed cost breakdowns have also been produced to identify those items of the network build that contribute most to the cost per customer connected. These individual items are now being targeted for potential cost reduction. For a PON network the highest cost items are distribution blown fibre and optical splitters. For dedicated fibre the highest cost items are primary cables and distribution blown fibre.
|
Written on 2:52 PM by ooe
Results from NetMod can be either numerical or graphical. Figure 10.6 shows an example of a graphical summary output from the demographics module. It highlights the number of potential customers in East Anglia for a particular system service set. With no demographics applied (100% of population) customer take-up numbers are shown within predefined bands, whereas when demographics are applied only targeted customers are identified.
These maps are useful for visualising the effect of varying estimated customer take-up of a particular service set/technology. They can also be used to develop strategies for cost effective network rollout based on targeting specific customer types that promise high potential revenue.
Simple networks designed by the E-side and D-side geography builders respectively are shown in Figures 10.7 and 10.8. Figure 10.7 shows a spine and spur network linking a number of primary nodes (positioned at PCPs) to their serving exchange. Figure 10.8 shows a network connecting secondary nodes (positioned at DPs) to their serving primary node (positioned at a PCP). By repeatedly running the geography builders for all the E-side and D-side networks within an exchange area the entire network within the exchange area can be designed. This can then be dimensioned and costed. This process can be repeated for all exchanges within the UK or for a representative sample of exchanges.
|
