National Broadband Plan: Discussion (Resumed)

I welcome, from the Regional Internet Service Providers Association, RISPA, Mr. Marcus Matthews, managing director; Mr. James O'Sullivan, director of Whizzy Internet; Mr. John Gartlan, director of Net1; Mr. Gurmukh Neote, director of BBnet; and Mr. Martin List-Petersen, director of Airwire.

Before we begin, I draw their attention to the fact that, by virtue of section 17(2)(l) of the Defamation Act 2009, witnesses are protected by absolute privilege in respect of their evidence to the committee. However, if they are directed by the Chair to cease giving evidence on a particular matter and continue to do so, they are entitled thereafter only to qualified privilege in respect of their evidence. They are directed that only evidence connected with the subject matter of these proceedings is to be given and are asked to respect the parliamentary practice to the effect that, where possible, they should not criticise or make charges against any person or entity by name or in such a way as to make him, her or it identifiable. Any submission or opening statement witnesses have made to the committee will be published on the committee's website after this meeting. Members are reminded of the long-standing parliamentary practice to the effect that they should not comment on, criticise or make charges against a person outside the Houses or an official either by name or in such a way as to make him or her identifiable.

I ask members and witnesses to turn off their mobile phones or put them in flight mode as they interfere with our sound system. I invite Mr. Matthews to make his opening statement.

Mr. Marcus Matthews

On behalf of the Regional Internet Service Providers Association, RISPA, I thank the Chairman and members of the committee for the invitation to appear before the committee and contribute to its discussion on the national broadband plan, NBP. As part of the delegation today, I am joined by several subject matter experts and representatives of RISPA members who will be answering technical questions.

RISPA is a trade association that was founded by a number of Irish-owned Internet service providers, ISPs, that wished to establish an industry representative body that could provide a coherent voice for the more than 40 ISPs that have been successfully delivering broadband services throughout the rural regions of Ireland for the better part of 20 years. RISPA's aim is to raise awareness of regional ISPs and address the issues they face so that the roll-out of reliable high-speed broadband services can be accelerated and delivered to rural Ireland within two years.

In respect of the purpose of RISPA's appearance before the committee today, RISPA recognises the important role that the committee serves in scrutinising all matters pertaining to communications policy and appreciates the privileged opportunity to assist the committee's examination of rural broadband issues. RISPA's appearance before the committee today is intended to help the committee's investigation of the NBP's process and provide advice about the options for accelerating the roll-out of rural broadband. In support of its appearance, RISPA has provided a written submission that outlines issues relating to the NBP's process and proposes a different approach to solving Ireland's rural connectivity gap. RISPA's suggested approach is based on a hybrid model of complementary technologies and a series of proven policies that collectively represent an alternative to the NBP. RISPA intends that the information it presents to the committee will provide its members with new insights on the NBP and the best options for rolling out cost-effective and timely high-speed broadband services.

RISPA has serious concerns about the NBP. Regarding the accuracy of the intervention area, a recurring issue throughout the NBP's procurement process has been the mapping of premises that do not have access to a minimum 30 Mbps broadband service. The implications of these mapping inaccuracies were most evident when 300,000 premises had to be removed from the intervention area because Eir concluded that they were commercially viable for the provision of its fibre optic broadband service. Such mapping inaccuracies were exacerbated by the fact that the intervention area failed to take into consideration the approximately 125,000 premises that are customers of regional ISPs, many of which had already begun rolling out next generation broadband services to rural Ireland when the procurement process began. Ultimately, the accurate mapping of the intervention area is necessary to minimise the financial burden on the State and prevent the overbuilding of infrastructure. Consequently, RISPA advocates that a real-time market monitoring system be implemented and maintained by ComReg and that such a system would be designed to allow ISPs to demarcate the areas that they serve and provide a facility for citizens to appeal such information.

Regarding the issue of encroachment, related to the problem of an inaccurate intervention area is the issue of the encroachment clause in the NBP's proposed contract. Such a clause would compensate the winning bidder should the number of premises in the intervention area fall below 542,000. Considering that approximately 125,000 of these premises are already customers of regional ISPs, RISPA is concerned that upon signing the NBP contract, the State may be immediately liable for the payment of large encroachment fees to the winning bidder.

Regarding the potential removal of retail restriction, a leaked aide-memoire from the Department of Communications reveals that the winning bidder is to be the retailer of last resorts. In the event that the uptake of fibre broadband services is lower than expected and the cashflow of the winning bidder suffers such that it cannot afford to build out and operate the network, RISPA believes that there will be pressure put on the Department to lift the retail restriction. If this were to happen, the ISPs that retail broadband services over the network could be disadvantaged and possibly exit the market, which would leave the winning bidder in a monopoly position that would run contrary to the very purpose of the NBP.

In respect of EU state aid rules, RISPA notes that the EU guidelines for the development of broadband networks to correct market gaps direct that member states should limit the scale of their interventions as much as possible. The guidelines instruct member states to intervene only after the private sector has failed to deliver services commercially. Where a member state elects to intervene, the guidelines clearly state that interventions should begin with the least intrusive policy instruments, such as those described in RISPA's written submission. Only after such policy measures have failed does the EU recommend that member states consider more intrusive interventions like the NBP. In short, the guiding principle of the EU rules is that interventions should target market failures and not crowd out existing market operators from areas in which they plan to invest or have already made network investments.

RISPA fully supports the laudable aim of rolling out high-speed broadband services to rural Ireland. RISPA believes that the NBP will fall considerably short of its target for 100% of premises to have access to broadband services and that the project does not represent good value for money. RISPA would argue that the collective policies it has outlined in its written submission represent an alternative approach that would achieve the NBP's aim in two years instead of seven and for a net cost to the State of €402 million compared with €2.97 billion. Consequently, RISPA advocates that the NBP's procurement process be halted so that a comprehensive evaluation can be undertaken of the recommendations that it has outlined in its written submission, which the committee is now invited to discuss and ask questions about. This concludes RISPA's opening statement. I now yield to the Chair.

I will start by asking why RISPA did not enter the procurement process back in 2015.

Mr. Marcus Matthews

In terms of the way it was set up, it was for the use of licensed spectrum. For the wireless option, the restriction was that one had to use licensed spectrum to do that. I think the other witnesses would be more familiar with the timing of the auction. I cannot remember exactly when it was but I believe the auction for the licensed spectrum had just passed-----

Mr. John Gartlan

It came in around that period for the 3.5 GHz licence, so the two things happened in parallel, and as a result, the operators of unlicensed spectrum were excluded from taking part in the submission process at that time.

Was that the reason RISPA did not enter the procurement process?

Mr. Marcus Matthews

That was the reason the industry did not enter the procurement process. Obviously, RISPA did not exist as an entity during that time. RISPA is a new entity designed to bring the industry together to operate as a coherent voice. However, there have been other organisations down through the years that have repeatedly attempted to influence public policy and the process itself. There were consultations and recommendations, paperwork was filled out and options were explained to the Department.

Mr. Martin List-Petersen

There is another item. Regarding the 3.5 GHz licensed spectrum, before the auctioning off of the spectrum so the change ran parallel to the start of the NBP procurement process, each provider could apply for a 20 km radius - a chunk of Ireland. They would buy them in the areas they were serving. Obviously, a regional provider will only serve a part of Ireland. With the change of licensing of that spectrum, all the providers that had been using 3.5 GHz spectrum were basically booted out of that spectrum and forced to move to unlicensed spectrum because it was not affordable for them to buy a nationwide spectrum block.

Have the witnesses consulted ComReg with regard to spectrum?

Mr. Martin List-Petersen

I personally expressed interest in Airwire participating but the figures it was aiming at were out of our budget.

What did ComReg say?

Mr. Martin List-Petersen

Basically, it was a nationwide spectrum. It would not have made any sense for us to participate in that.

Airwire consulted ComReg. What was its response?

Mr. Martin List-Petersen

Obviously, we could have participated in the bidding.

It could have.

Mr. Martin List-Petersen

We could have but where we were with regard to monetary expenses was very clear from early days.

Could Mr. List-Petersen explain what he means by monetary expenses?

Mr. Martin List-Petersen

The cost of buying a block of spectrum would not be something we could afford as a regional provider.

Mr. Marcus Matthews

Just to provide some more clarity, the policy in Ireland has been to auction spectrum nationally for the whole country. A multi-award model was examined for the 3.5 GHz element and that did divide up various areas. It was quite successful but it still covered huge chunks of the country. The United Kingdom, for example, has addressed this issue quite successfully and it has implemented a use it or lose it policy on spectrum. The mobile network operators can purchase vast blocks of spectrum to serve the urban and suburban areas they need. In the rural areas, however, where they do not tend to use the spectrum, what is happening is that regional ISPs purchase that auction for that region to use it in a service. It is quite a good model.

Mr. Martin List-Petersen

Another factor also made it unfeasible for us to partake. The technology requirement set for the spectrum auction mandated the use of long-term evolution, LTE, based equipment. We did not have the freedom of choice to use alternative transmission methods. If we were to purchase nationwide spectrum, it had to be LTE equipment. That equipment costs much more than alternative products we can get that can achieve the same speeds or similar.

I will move on to my colleagues and I call Senator Lombard.

I welcome our witnesses to this discussion on the national broadband plan. We have heard from witnesses, particularly from the Department, who have spoken about broadband being the game changer for the next 20 or 30 years. I would like some information on the speeds that will be provided in future. We have had presentations from the Department concerning elearning, efarming and ehealth. Developments in those areas will require major changes and high speeds. How would the witnesses fit their criteria within the context of the speeds required? Are they happy that their technology has the potential to provide those speeds?

Mr. Marcus Matthews

In our written submission we have given an overview of the evolution of broadband radio technology. We referred to the speed in 2012 as being 30 Mbps and remarked on the jumps experienced since then. The products themselves are now rated for 200 Mbps and some of the members here have connections installed that are achieving that speed in licence-exempt bands, which are noisy. Regarding the evolution of long-range technology, by 2025 the prediction is that 500 Mbps will be possible. I am referring to long-range and serving multiple premises from the same site.

At the moment, in respect of short-range provision, a new category of short-range broadband radio technology has come on stream which allows premises to be linked together. It is like laying wireless fibre. That is the easiest way to describe it. That technology operates at about 1 Gbps. In the second quarter of next year that technology is going to become multi-gigabit. It will then be in the range of 3.5 Gbps. If there are strings of houses within a reasonable distance of each other, it will be possible to install these radio units and it is possible to keep going-----

It is basically being bounced?

Mr. Marcus Matthews

Technically, it is a mesh networking technology. All of the elements interlink with each other. It is exactly like a fibre cable. When a fibre cable is laid, all of the homes that it passes will be connected to that same cable and all of those houses will share that same bandwidth. These multi-gigabit broadband radios do exactly the same thing as fibre cable except they do not need the wire. That is basically how it works. Regarding the evolution of speeds, we are already seeing multi-gigabit products coming onto the market now. It is admittedly a category of short-range radio technology. On the long-range technology, however, we are already up to 200 Mbps and the conservative estimate is 500 Mbps by 2025.

When I appeared before the Committee of Public Accounts I remember that there were discussions on this issue in the context of wireless technology. I heard figures mentioned of 20 Gbps. I tend to stick very close to hard research that I trust and that figure of 500 Mbps by 2025 is a very realistic. That is the evolution of where we are going. There are appendices in the back of our submission which contain information on the specifications under development. The fastest specification at the moment is just over 10 Gbps. In the context of short-range, we are seeing great speeds. To transfer that into the longer-range radios takes time. Regarding where this is going, however, and based on the technology shifts experienced to date, there is no reason to doubt that trend and trajectory will not continue.

RISPA technology is mainly based on line of sight.

Mr. Martin List-Petersen

That depends on which spectrum can be availed of. If we operate in the 5 GHz spectrum, then the technology will be line of sight. If we operate in the 3.5 GHz spectrum, however, technology exists that is near line of sight so it will cope with certain impairments. If we go down then in the radio spectrum to 700 MHz or similar, we will be talking about non-line of sight. In the United States, there is an unlicensed spectrum in the 928 MHz band. The US did not roll out the global system for mobile communications, GSM, in the 900 MHz band but in the 850 MHz band. It started with 1900 MHz. Fixed wireless operators in the United States are operating non-line of sight in the 928 MHz band to provide broadband to customers.

The current spectrum being used is basically line of sight, as far as I can determine.

Mr. Martin List-Petersen

It is

What would be the environmental impact of trying to ensure that line of sight capability is available all over Ireland?

Mr. Martin List-Petersen

We can never say that we will cover every premises. We will have to look at every individual scenario. The way we operate today with fixed wireless technology is that we have multiple base stations and repeaters, which means that if it is not possible to connect a premises from one location, then it will be possible to connect it from another. There have to be multiple options.

Mr. Marcus Matthews

In our model, we were trying to build the most stress-tested conservative model for coverage that we could. We reduced the distances in those sites down to below 6 km. When we get down to those distances, the likelihood that we will not have line of sight or the likelihood that we will not be able to cover somebody is very low. We also discussed in our submission the importance of allocating the correct spectrum bands. We discussed allocating low bands, namely, 700 MHz, which is non-line of sight as we have just heard, and allocating something in the mid-band, that is----

What infrastructure would be needed for that? How would an area be covered? Will it involve masts?

Mr. Marcus Matthews

The difference is that broadband radio technology tends to involve small and micro sites. Mr. Neote has set up an example of a broadband radio unit here today and it can be seen how small it is. Regarding what regional ISPs do, the technology involved does not tend to be massive tower masts on hilltops. The technology used tends to be on a high point in a local rural area. That could be a farmer's shed on top of a hill. It could be as simple as that. I reiterate that broadband radio technology is a different type of technology and not the same as cellular technology which requires those high sites which are trying to cover a larger distance.

The redundancy built into broadband radio technology and specifically the model we are talking about is multifold. We are talking about dealing with small distances from the sites to the premises and that is the first aspect of redundancy. There are also, as Mr. List-Petersen said, multiple different angles available because of the distances from the units themselves. It is different, therefore, in respect of the number of sites required. We did the calculations, and again, we were just looking at covering the entire landmass of Ireland, so it was a simple and very rudimentary calculation, and we came up with 795 sites based on that model.

Mr. Martin List-Petersen

A large number of those sites are not known to ComReg because they are not used by the average mobile operator. Many of the regional providers operate on a completely different subset of small repeaters. It is for budgetary reasons but also because we are happy enough providing to areas even if there are only 40, 160 or 300 customers there. In the circumstances in question, we cannot use a traditional mast – never mind the cost of gaining access to one.

Are the masts usually made by Mr. List-Petersen's company?

Mr. Martin List-Petersen

There are agreements with people in local rural areas who are as much interested in broadband as everybody else. There are basically individual agreements under which we have built our own masts. Usually, there are no other providers involved.

Mr. Marcus Matthews

To revert to the question on what infrastructure would be required, we are talking about the base station units themselves and a hybrid of backhaul broadband radios, which are slightly larger and bulk transport data from the site to the nearest access point. In some cases, if there is a particularly large number of premises in a given area, fibre would be run to the site. That is the beauty of the model. We are talking about fibre extension, not building huge swathes of network infrastructure. If one fibre extension is brought to the right geographical area, multiple broadband radio sites can be loaded on. Each one could have 160 to 200 customers. Very quickly, therefore, a large number of premises can be covered with the next generation access speeds we are talking about in a much more flexible and efficient way.

With regard to infrastructure, it is very scalable. That is the simple answer. It is not necessary to take up every road. Separate fibre extensions are installed and then technology of the kind I have referred to is used. It can be put up in an afternoon. One can connect a home in two hours. I refer to the antenna. One can see from the equipment I have to hand the antenna at the top of the stand. That is the sum total of it. One can run all the broadband services over the top of that. I refer to high-speed Internet access, a telephone service, television and whatever one wants. That is the sum total of the infrastructure.

My final question follows from what the Chairman asked. Mr. Matthews did not get involved in the process until 2015. The other three operators in the process were all opting for fibre to the home. He has suggested a totally different model. If the process were reopened, would his company come on board in the retendering? Does he believe the process has moved on such that the focus is just on the fibre itself?

Mr. Marcus Matthews

The process would have to be fundamentally changed and brought in line with what other countries have done. What we are proposing to do is abnormal. The only jurisdiction of which I am aware that has truly achieved 100% fibre is the island of Jersey. Everywhere else, there is a hybrid model where broadband radio covers the rural part. If we were to consider this again, we would need to consider what has been done in the United Kingdom, for example. We would need to consider a use-it-or-lose-it model of spectrum. We would need to consider what Scotland is doing. It tried to have 100% fibre under the digital Scotland superfast broadband scheme. It did very well but it does not have the same type of development pattern. It has a more clustered development pattern. It had very successful fibre roll-out but it concluded that it makes sense to introduce a hybrid approach. It is going at it in terms of the design of the consumer subvention programme. This involved connection grants, which are technology agnostic. Once the Internet service provider can meet the performance criteria, which is what we are talking about, it can use whatever technology it likes. It can run fibre to the home or it can use broadband radio. In terms of reopening the process, we would need to follow what other countries have done or elected to do and are doing successfully. We would need to consider a licence-protected model, such as that in France. A licence-protected model means the licence is not auctioned. It involves protecting a section of the band. The only people who may access the band are those who make an application to deliver high-speed rural broadband to premises. If they do not meet the performance targets, the application is rejected. It removes the cost burden and allows existing regional Internet service providers to start solving the problem immediately. By striking the decision in terms of the licence-protected model, it immediately opened up 2 million homes to coverage by existing regional providers.

Let me return to the question of reopening the process and what we talked about in our written submission. One would need a licence-protected model that allocates a suitable spectrum. One would need some low band and some midband. One would need several of the other policy instruments that have been outlined in our written submission. There should be a consumer subvention model to cover those premises that are not commercially viable. There would ideally be a State-backed affordable loan scheme, similar to what was done with the agricultural cash flow scheme.

What Mr. Matthews is saying is in some ways contrary to what we have heard. As I recall, representatives of Analysys Mason, a firm of international consultants, said when they were before us they had never seen any place like Ireland because of the ribbon development along roads. If I recall correctly, they concluded in their work on the national broadband plan that rural Ireland could not be served in the manner Mr. Matthews has suggested. Owing to development along country roads, it was found the only technical way of proceeding was to run fibre. It is difficult because we are not technical experts. He said the work can be done with a similar non-cable-based, house-to-house solution. Who do we believe technically?

Mr. Marcus Matthews

I recall the Deputy saying that at the committee meeting last week. I read extensively the various reports by Analysys Mason. None of what it said to the committee makes sense. It has more documents discussing the importance of wireless technologies for solving the rural broadband issue. It maintains a standing report on what other countries are doing with fixed wireless access technology. It monitors continuously how the countries on its list use wireless technologies to address the issue of broadband access in rural areas. It is quite detailed. In all the documentation one reads, the recommendation is that the wireless technologies comprise the way to cover rural areas and serve them with high-speed broadband services. I am not sure why the representatives said to the members that broadband radio is not the appropriate technology.

I believe they referred to topography where there is a strip of houses along a road. Even if there is a mast on a high point, such as at a local school, church or barn, it is very difficult to cover every house and the only way one can be certain of doing so is by looping from house to house. I suppose the representatives were not saying it could not be done with house-to-house fixed wireless, or radio technology. I understand them to say that one could not easily get a high point to cover all the houses along a road because they may be behind trees, a little hill or a turn. There is a meandering pattern of development everywhere that has no concentration of hamlets. There is just road-based housing development, which makes Ireland completely different from any other country. There is nowhere else in the world that has a road-frontage, sprawl-development model. Analysys Mason says it is just not physically possible to proceed with a fixed wireless solution.

Mr. Marcus Matthews

Mr. List-Petersen deals with this all the time.

Mr. Martin List-Petersen

I would make a point against that argument. More than 70 providers are registered with ComReg for a general licence for wireless broadband transmission. If more than 40 providers are actively selling wireless broadband in rural Ireland, what is the reason for that when it is not possible?

What they are saying is that it is very difficult to cover every house.

Mr. Martin List-Petersen

Yes, but as I outlined, most providers go in and find multiple spots so that they can get every angle. Obviously, there is the odd house, which we have come across, that is completely enclosed in trees and it cannot be done. I know of one scenario where the receiver was put up in a tree.

The technicalities of this are difficult. When we refer to radio, it is the radio wave spectrum. Fixed wireless is the same thing. Either term - wireless or radio wave - can be used. It is all on the invisible microwave spectrum. Mr. List-Petersen said that the next likely development is in the 700 MHz spectrum but he did not finish the point. I presume it was approximately 2 GHz.

Mr. Marcus Matthews

We were talking about suitable spectrum bands for implementing the model we have outlined. To briefly track back on Mr. List-Petersen's point, in terms of the position we are in now where we have the odd house which cannot be covered, it must be remembered that currently all regional providers are forced to operate in licence exempt bands, which are very noisy. The spectrum that has been allocated is not the ideal mix. It is very restrictive in terms of the options. What we have done in this written submission is outline the mixture of spectrum bands of which 700 MHz was one. That is a low band which the members have heard is for non-line-of-sight locations and to deal with trees, the drumlins and everything else one needs to bounce the signals through to make sure they are getting the 150 Mbps.

We have outlined also the position in terms of the midband. There are three options in the midband; they are in the written submission. I refer to the 2 GHz band. What is needed are continuous blocks of approximately 100 MHz per second assigned in a licence protected model. That means that the only people who can use it are, say, the regional Internet service providers, ISPs. If that licence protected model was implemented and suitable spectrum bands such as the ones outlined - a mixture of low and mid frequencies - were allocated, we would get into a situation where that lone house is suddenly covered.

We have taken that one step further and said that if we reduced the differences from the sites and took a very conservative limit - in reality one would probably do it over a longer distance - of, say, a maximum distance of approximately 6 km, it would be very interesting to see if there was a premises that was not covered.

From memory, was that 700 MHz part of the digital dividend? Was it formerly television and then switched to mobile cellular?

Mr. John Gartlan

The 700 MHz is currently occupied by 2RN. It is due to be cleared by March 2020. We do not know what will happen at that time. We imagine it will be auctioned off but it will be a highly significant and useful band width for use for broadband radio because it could provide coverage to, say, the final 5% of premises where trees, drumlins and small hills are in the way. Generally, the lower in the frequency we go, the better it penetrates through obstructions. I will use the analogy of driving from Dublin to Belfast and listening to RTÉ. One starts off in Dublin listening on digital audio broadcasting, DAB, which is on a higher frequency. The quality is very good but the range is quite short at 30 km to 40 km, which is good compared to broadband radio. However, as one drives up the coast and moves into areas where DAB cannot cover, it falls back to VHF, or what we call traditional FM. That provides coverage to approximately 60 km from one's area because the frequency is lower and it penetrates better through obstructions, hills, trees and forests. Driving along the motorway one enters Northern Ireland but once one goes 60 km in, one is out of range because the terrain has blocked it and one has to move to a lower frequency again such as AM or LW to continue listening.

A similar analogy can be given in respect of broadband radio in that the lower the frequency used, the better the penetration one gets from broadband radio. That is at a slight sacrifice in terms of speed but we envisage that will be used only for the 5% of premises we cannot service on the mid and high frequencies.

We hear about LTE or 5G. Where does Mr. Gartlan believe 5G will fit into that analogy he used? What does he mean when he refers to LTE? I am familiar with the term.

Mr. John Gartlan

LTE is classified as 4G. I believe what we are getting on 5G, and I am not the world's biggest expert in 5G technologies, is a next generation of 4G operating in the mobile allocated bands, probably using 3.5 GHz to 2,100 MHz, and perhaps some at 900 MHz. Would I be correct?

Mr. Marcus Matthews

I think it is, yes.

That is separate. That is not what Mr. Gartlan was talking about in that mix.

Mr. John Gartlan

Absolutely. It is separate.

Mr. Marcus Matthews

The 4G and 5G technology is a new category of mobile communications technology so it is coming out of the cellular family of specifications. When we hear about 5G, it is technically a 5G new radio line of code. It is a software upgrade to existing 4G technology. The straight up 5G in terms of hardware will come out later in the year. It is interesting that in the US, Horizon did some trials in terms of rural broadband with several mobile network operators. They found that in the rural regions the performance increase over 4G was not substantial.

If we go back to the original concept of 5G in terms of its development, it was to serve high density environments such as airport terminals and subway stations. It started out as a need to develop a new standard because everyone now has a mobile phone and we have scenarios where tens of thousands of people are in very close proximity to each other. We need technology that can deliver a large volume of throughput to all of these people in close proximity to each other. The scope grew as the technology was developed and it co-opted the rural broadband access issue. It has since become associated more with the idea of providing rural broadband access than, say, the original impetus for its development.

In terms of its development, 5G is not a technology designed for the type of rural broadband and rural connectivity gap we face. From its initial concept to the development stages through to the finished product, the technology used by regional ISPs has always been about connecting multiple premises from a single site at the highest speed possible.

In his presentation Mr. Matthews said that a leaked aide-memoire from the Department reveals that the winning bidder is to be the retailer of last resort. Can we get a copy of that?

Mr. Marcus Matthews

Yes. Clause 19.

The 40 ISPs have 125,000 customers. Some of those customers' houses will be passed. They will have the option of getting the fibre connection at a charge of €100. A total of 2% might be by alternative means but the vast majority will be fibre. Does Mr. Matthews believe his ISPs will keep the retail relationship and switch over to the fibre connection? Will he migrate that easily from his fixed wireless network to retail through the fibre? If there is not a sufficient number of customers, where is that risk that the Department would lift the retail restriction?

It has been fairly clear all along that this has to be open access. There could be open access where the provider was also acting as a retailer, I suppose. There is no sense whatever that Grattan McCourt has infrastructure or any interest in becoming a retail provider. Surely that is not the risk to the companies represented here. The risk is that some other service provider with another retail arm would use the fibre connection to compete with them and take out small, local, indigenous companies that have been doing a good job over the past 20 years. The bigger risk is not that the developer would take the business but that other retailers would do so.

Mr. James O'Sullivan

The regulated price for us to use Open Eir is €23.50 excluding bandwidth and VAT. Over a 12-month contract, we have to cover the cost of the connection fee and a decent wireless router, which is about €80. Looking at the pricing, we believe that to break even, we need to be charging about €50, with no margin. I am concerned at the price. Many of our customers can justify paying €45 for broadband access but not much more than that. That is why we are concerned about the take-up of the national broadband plan.

Why would there be a need for a wireless connection charge? Where is the wireless connection occurring?

Mr. James O'Sullivan

The wholesale connection for which we will be charged by national broadband Ireland, NBI, would be €23, give or take. On top of that, we have to provide bandwidth which, for current average usage, is about €6 per customer. Then we have to provide a wireless router. As far as I am aware, there will be a connection fee that-----

Would the wireless router be inside a house?

Mr. James O'Sullivan

Yes.

Mr. James O'Sullivan

A decent wireless router such as a FRITZ!Box or something that will deliver good WiFi throughout a house costs about €80 plus VAT. We then have to add a margin. We are not 100% sure if there is to be a connection fee but I understand there will be a connection fee of €100 for a house. All of that has to be built into a 12-month contract. Our concern is that all these customers will not pay €60 a month or more for their broadband service. The take-up will be smaller than expected and someone will be forced into selling themselves.

Where is Whizzy Internet serving at the moment and what monthly cost it is charging?

Mr. James O'Sullivan

We provide a 50 Mbps service with unlimited usage for €45 a month.

Where is that?

Mr. James O'Sullivan

It is the Gorey area of north Wexford. We have a faster service but the take-up is much lower.

Who does RISPA represent? Who are its members?

Mr. Marcus Matthews

RISPA was founded by 13 regional ISPs, as I said in the opening statement, to represent the broader industry. There have been previous successful attempts in which two previous trade associations were set up. When the national broadband plan process started, the industry as a whole was very active and spent an inordinate amount of time filling out paperwork and engaging in consultations, much of which was to no avail. This time around in terms of the industry having a third run of it, there was a feeling that we had done all this before when we submitted all those reports. In 2015, an independent report by the eminent economist, Colm McCarthy, went to the Department but nothing came of it. With RISPA it was a case of those who were willing to go again.

How many members did Mr. Matthews say there are?

Mr. Marcus Matthews

There are 13 providers.

Can we have the names of those or does Mr. Matthews want to submit them?

Mr. Marcus Matthews

We can submit them if the Senator wishes.

Perhaps we could have an example of one or two now.

Mr. Marcus Matthews

The companies joining me today are Net1, BBNet, Whizzy Internet-----

Could Mr. Matthews say the names slowly and clearly?

Mr. Marcus Matthews

They are BBNet, Net1, Whizzy Internet and Airwire. We also have HiSpeed Wireless, Total Wireless, Northwest Broadband, Wireless Connect and several others.

How many of them are indigenous providers?

Mr. Marcus Matthews

All of them.

Are there many providers in the market that are not members of RISPA?

Mr. Marcus Matthews

We have 25% or thereabouts-----

Of the providers.

Mr. Marcus Matthews

-----of the regional ISPs.

With respect, 25% is small. Mr. Matthews thinks it is enough to do what he is proposing, however.

Mr. Marcus Matthews

I understand where that misunderstanding has come from. Our written submission is not that RISPA does this exclusively. The written submission is about how the industry as a whole could do it.

RISPA is assuming in the written submission but it represents 25%.

Mr. Marcus Matthews

RISPA is just an industry representative body designed to gather together research and inputs.

I gather we are in public session so if anyone is watching they will want to know this. How many broadband customers do RISPA members have in rural Ireland?

Mr. Marcus Matthews

They have 20,000.

Mr. Marcus Matthews

The 13 founding members of RISPA have 20,000.

What level of equity would they collectively be prepared to provide for a new project under their proposal? What kind of money are they prepared to put in?

Mr. Marcus Matthews

In terms of the money, the regional ISPs have always self-financed the expansion of their networks, even when the State was subsidising other-----

Do they propose to put up an amount of money that would provide a security?

Mr. Marcus Matthews

The idea with the written proposal was that bonded insurance would be provided if loans and connection vouchers were being paid.

Otherwise RISPA providers would carry their own.

Mr. Marcus Matthews

They would carry the risk but a rigid compliance programme would have to be developed. It is written into the submission that it would have to be developed for any regional ISP to access or draw down a loan or connection voucher. It is the same as in other jurisdictions.

Has RISPA consulted ComReg on spectrum requirements?

Mr. Marcus Matthews

No. RISPA has not been in touch with ComReg yet on spectrum allocation. The policy-----

With respect, would that not be fundamental?

Mr. Marcus Matthews

Procedure dictates that we go through the committees first before moving on. We have been focusing on going through the Oireachtas and the committees. It is from the committees that the clerks of the Department will return the documentation, as I am sure the Senator understands. We have a meeting scheduled with the Department on 25 July. Our focus today is to follow due process in terms of transparency.

Is it proposed to speak to ComReg after that?

Mr. Marcus Matthews

When we meet the Department, we will go from there accordingly. In terms of spectrum allocation and changing that policy, there is a subtle relationship. ComReg has wide scope in the strategies and policies but is advised by the Minister on what it looks at and investigates. Section 13 of the Communications Act 2002 allows the Minister for Communications, Climate Action and Environment, with the permission of the Minister for Finance, to allocate any spectrum band for any justified purpose. The Minister can direct ComReg to do that. That is worth noting. It does not have to come from ComReg. The Minister can say he wants it considered and request a submission regarding how it would be implemented.

The question arises as to why RISPA did not get in previously. Its answer is that it relates to the terms of the procurement.

Mr. Marcus Matthews

The specific reasoning was that in order to submit a wireless proposal one had to use licensed spectrum and at the time the NBP process started - it still is the case but not so much with the 3.5 GHz - licensed spectrum was at national auction so one needed vast sums of money to win that auction. The reason, historically, that the mobile network operators have always won those auctions is that they have the scale and they are delivering cellular services nationally. What we are saying is that the trend in other countries is that many have realised that those mobile network operators are only using the spectrum in urban and suburban environments. What these countries are doing is looking at regional allocation models and deciding that if the spectrum is not being used in rural regions, they will hold regional auctions for that or they will have regional licences that can allocate that spectrum at a much more affordable rate. In the UK, for instance, if a mobile network operator is not using licensed spectrum in a rural area, a regional ISP can apply for the spectrum and pay £900 sterling for it per year. I believe that is correct.

Mr. John Gartlan

Yes, it is per year.

Mr. Marcus Matthews

The regional ISP can get the spectrum for three years to deliver it. If the mobile network operator then comes in after three years and there is a significant surge, it can then take the spectrum back. It is a much more efficient way of utilising the spectrum and one will get many more people involved in it who can then deliver more services.

Yes, but there are many risks. That is all a bit hazy from a Government perspective in terms of providing a service, getting it done on time and leaving very little room for doubt.

Reference was made to Mr. Colm McCarthy, which is fair enough, but is there any evidence that the companies represented here ever came together collectively in the past or made any effort to get involved? If we accept that the procurement process, as established, was too rigorous or constrained for them to participate, they did not lobby or come before this committee. These companies did not come to me as a public representative and put an alternative view. I take it that is also the case with my colleagues as I would have heard about it anecdotally. This is akin to a colleague buying a bad car which I knew it to be a bad car and then my saying to him that I knew it was a bad car when he had already bought it. Why were the witnesses not more proactive about getting involved at a critical stage in either forming our opinions or forming the Department's view. Why are they coming in at this point?

Mr. Marcus Matthews

In terms of the engagement with the process, there was extensive engagement with the Department by various regional ISPs and through industry representative bodies at the time. I am sure boxes of correspondence could be supplied if people require it. We would be happy-----

There is no sign of the witnesses making any concerted, collective effort at that time to say the process and its terms should be different and talking to the policymakers. I am a long-standing member of this committee. I remember when my good colleague, Deputy Eamon Ryan, was Minister that I was spokesperson in the Seanad on this area. We debated the broadcasting legislation and other Bills together. I am almost reluctant to say how long I have been on this committee, yet nobody ever raised this issue with me before now. I am far from having the solution but, collectively, the witnesses should have been talking to us and our political leaders. Does that make sense?

Mr. Gurmukh Neote

We were previously part of another association, ISPI, and we were represented by it. When was the Colm McCarthy report undertaken?

Mr. Marcus Matthews

It went to the Department in November 2015.

The association made very little impact. It did not get us to ask departmental officials questions on their propositions. I hope the position will change now that the association is more organised but there is no discernible evidence of any input from these companies previously or any trace of them trying to move the goalposts in the past.

Mr. Marcus Matthews

I accept that the footprint in terms of the committee was very light, but we had extensive engagement going back some time with the industry overall, in particular with the Department and ComReg. I object to the notion that the industry was not actively involved and trying to-----

Does Mr. Matthews accept that the committee was never involved or approached?

Mr. Marcus Matthews

I accept that the committee was not engaged with-----

I attend my parliamentary party meeting every week - Fine Gael is the largest party at this stage - but I do not remember any discussion with colleagues or any submissions at that level either. In any event, that is history. It is almost unreasonable at this stage for the association to be doing what it should have been doing a long time ago.

Mr. Marcus Matthews

That does not excuse the fact that the process is flawed and requires extremely close scrutiny, in particular given that the technology has now evolved to a point that we can do it more cost effectively in a shorter period of time.

In terms of the alleged flaws, does Mr. Matthews accept that Mr. Weckler from The Irish Times is an acknowledged professional expert and journalistic commentator?

Mr. Marcus Matthews

Is that Mr. Adrian Weckler?

Mr. Marcus Matthews

I would not accept that he is an expert in the subject matter. I do not think anybody here would. He is a technology journalist.

I would have thought there is a general view that he knows what he is talking about. He, among others, said that to achieve fairness, as policymakers we cannot make a decision. Coming from my constituency of Cavan-Monaghan, it would be negligent of me if I stood idly by – to use that awful cliché – while a less sound system of providing broadband to people in isolated areas was chosen. The view of Mr. Weckler and other technical experts – evidence was cited earlier – is that only through the fibre option can one be sure beyond doubt that fairness and evenness are created. I refer to the man at the back of the hill, a phrase I grew up with, in other words, the people who live in isolated rural areas. Fairness should be applied to them and their rights should be taken into account. Why should they be put into a two-tier system on a wing and a prayer and using less secure technology when the view is that the fibre optic solution is needed? It is the best solution. I am not saying that other solutions do not work, to a degree, and the witnesses are a crucial part of the market, but what we are talking about is whether there is a viable, workable alterative to the national broadband plan. That is basis of the terms of reference of the committee. Bearing that in mind, one must assume that we cannot have apples and oranges and that the solution must be all apples or all oranges, in other words, that the person in an isolated area such as Ballydehob must have the same access to broadband and the same level of technology as the person in south Dublin. As policymakers, we cannot make an alternative decision that would put people in rural areas on a different footing. We would be negligent if we were to do so.

Mr. Marcus Matthews

The Senator's argument is premised on the terms of reference and the quality of service criteria and that fibre is the-----

Yes, quality and certitude.

Mr. Marcus Matthews

On that point, what is very interesting is that the NBP process and the plan itself will already have a 2% component of broadband radio technology. They have already concluded-----

Mr. Marcus Matthews

No. Senator O'Reilly should please hold on for a second. The broadband radio technology they have already determined meets the terms of reference criteria. Already, they have concluded that it can achieve the quality of service criteria to deliver the target specifications of the national broadband plan. In terms of how other jurisdictions have approached the-----

If they thought that, it would be much greater than 2%. They are saying that there may be 2% in a doomsday scenario. They will not get 2% with fibre optic and in that case, that 2% facility is there. That 2% in very isolated locations would have to be met.

Mr. Marcus Matthews

We are not talking about the scale of the gap. The Senator said that the only way to guarantee the quality of service is through the fibre optic technology. What I am saying to the Senator is-----

That is the conventional wisdom.

Mr. Marcus Matthews

Right. What I am saying to the Senator is that already in the process they have determined that broadband radio technology meets that quality of service criteria. If the issue is, as it seems to have been in the past 18 months, that we have to guarantee a certain quality of service level, from a purely engineering standpoint - the committee can ask independent radio engineers and fibre optic engineers - it comes down to whether the chosen technology approach will achieve the specified quality of service criteria and the one that can do it more flexibility and more cost-efficiently from an engineering point of view is the technology that one goes for. The idea that we have to lay fibre because it is the only way to guarantee the target specification-----

Of the highest speed broadband to allow the person in the very isolated rural place to do work at home, have education at home, access education the same way as anyone else, access every service at home and be on a par and on the same standing as somebody in an urban centre. That has to be our policy objective.

Mr. Marcus Matthews

Absolutely.

We are told that the way to do that is fibre. That is what is inherent in the broadband plan, hence the extra cost. I will come back to cost in a minute. The certitude is there. Our fear in respect of Mr. Matthews' plan is that a two-tier world would be created.

Mr. Marcus Matthews

Even though, as I said, the technology already meets the target specification. The premise of the Senator's argument does not add up because it will be part of the NBP implementation.

It will be part, possibly for 2%.

Mr. Marcus Matthews

That 2% will become a lot bigger.

We are into semantics.

Mr. Marcus Matthews

It will become a lot bigger.

That is almost Jesuitical.

Mr. Marcus Matthews

The crux of it is that it already meets the quality of service criteria specification.

In that technical sense, in that instance.

Mr. Marcus Matthews

But that is what matters here.

For the generality up to that, for the 98% up to that 2%, one needs the fibre.

Mr. Marcus Matthews

One does not. The committee should not take it from me. Why not hear from some subject matter experts?

Why have Eir and SIRO chosen to use fibre in urban areas? Why did they not go for fixed wireless?

Mr. Martin List-Petersen

Open Eir has not used fibre in urban areas. They have used fibre to the cabinet. They are using copper technologies. The 1.7 million premises of their roll-out use copper. Of those, a large chuck are sub-30 Mbps. On top of that, because there is an issue where some of the copper lines still do not go to the nearest cabinet and they go to the exchange, to cabinets miles away, etc., which has not solved anything. That is why Open Eir has turned around, because it has this old legacy network that dates back 20 years or even more, and said that the only way it will fix this is by now also going fibre. This is what they do traditionally.

Open Eir, or, in its traditional sense, Eircom, also always had a nationwide 3.5 GHz licence. They have run WiMax trials. They have had customers on WiMax. They have never run the product because they are a traditional telecommunications company. SIRO, on the other hand, sprang out of ESB Networks. They are using ESB Networks power lines. Basically, the easiest way for them was fibre. It is down to the scalability with that. This is all urban. What we are talking about here is the national broadband plan, which is all rural. SIRO has nothing in such rural scenarios. It is cities, towns and villages to a certain extent. It makes a whole lot of sense with fibre there because if one runs with radio links and one does not have the spectrum to run non-line of sight, then one has houses in the way everywhere. In an urban scenario, one would need non-line of sight spectrum, and one would need a lot of it to do it, which is what the mobile network operators do. In an urban scenario, even though one has a 4G signal, often one does not have a lot of bandwidth because there are just too many people and there is not enough spectrum.

I am sorry, I interrupted Senator O'Reilly. I will let the Senator finish.

I am two issues away from finish. This next issue is one of human reality that, as policymakers, we cannot avoid. We have had a number of such issues. There is a considerable campaign in Cavan-Monaghan regarding the undergrounding versus overgrounding of the cables of the North-South interconnector. The not-in-my-back-yard, NIMBY, syndrome is evident all over Ireland. In the abstract, it is grand. Of course, people feel an unfairness and wonder why are all these structures, such as masts, lumped out in the sticks. Mr. Matthews and I can argue about the figure, but he will accept that he would need up to 6,000 masts all over Ireland to achieve his objectives.

Mr. Marcus Matthews

A total of 795. We included an accurate calculation in the document.

I was advised that it would be more.

Mr. Marcus Matthews

That comes from Mr. Adrian Weckler. He is focused on 5G, which, as we have explained, is a mobile communications technology. That would require a far greater number of sites. It is 795.

Here is our problem for Mr. Matthews to see. It is important we see his as well. It is being noted and will be part of our ultimate conclusions. However, he has to see our difficulty now. There is a political, policy and human rights imperative to get broadband to every home in the country quickly. Even under the Constitution, from a legalistic point of view, every citizen has to be cherished equally. Whether one looks at it constitutionally or one looks at it educationally, and when one looks at the PricewaterhouseCoopers, PWC, assessment of the good that broadband will do, putting the broadband plan in place will do such a range of things from home working right across-----

Mr. Marcus Matthews

Those benefits have been questioned by Mr. Robert Watt, who stated that the cost-benefit analysis, CBA, lacks continuity.

As a resident of rural Ireland, I believe in those benefits and I believe my people have a right to them.

Mr. Marcus Matthews

I did not say that they do not have a right to them-----

They have a right.

Mr. Marcus Matthews

-----but the benefits outlined lack credibility. According to the top civil servant whose job is to make that determination, the CBA lacks credibility.

Everybody in the country has the same right to this service. I dispute the 795 masts, but let us take that figure. By the time Mr. Matthews gets over the protests and all the civil disobedience and civil strife that will arise in the erection of them, there will almost be new technology. It will takes ages. Were he, for instance, to seek to put a mast in an area tomorrow, one could have protests and physical sit-ins.

Mr. Marcus Matthews

I have to ask whether the Senator was listening to the delegation because we made it clear how the technology works and that it does not require mast sites.

Is Mr. Matthews saying it will not even need the 795 then?

Mr. Marcus Matthews

The Senator is misunderstanding. A site does not necessarily have a mast on it.

How many masts will Mr. Matthews have then?

Mr. Marcus Matthews

Not 795. As Mr. O'Sullivan said, most of those sites are active.

Our problem politically is if one has any. It is not that we have a problem confronting the issue of the masts per se. Our difficulty is that the process will go on and that we will not get the broadband into the homes.

Mr. Marcus Matthews

I will let Mr. Gartlan come in here.

The concern is about the masts, planning, objections.

Of course. That is what I meant.

I will explain where the Senator is coming from. This has been raised a number of times during our hearings. ComReg, which is the trusted independent body in this field, has told us 6,000 masts would be needed to deliver 5G. I take the RISPA's point that it is proposing something completely different from 5G. The Plum report stated that fixed-wire operators would require 4,000 to 5,000 masts. If the witnesses are, therefore, proposing masts, we need clarity as to how many there would be, the planning issues that would arise and the timeframe for erection of the masts. We need that understanding.

How will this delay the process? If they want to make a written submission on this, that is fine; otherwise, they can comment now.

Mr. John Gartlan

I refer to the Senator's constituency of Cavan-Monaghan, with which I am very familiar, as our company serves it. We are currently rolling out what we call our next generation service. It is a 50-plus Mbps service. We have 19 sites in Cavan and Monaghan. They could be on a farmer's shed, the side of a pole in a haggard or a yard, or a hill, like some of them up in Slieve Glah, in Cavan, where there is a mast. From those 19 sites in Cavan-Monaghan we have close to 4,000 rural customers in the Senator's constituency. Regarding our penetration, we have recently obtained a test-and-trial licence from ComReg to operate on the 700 MHz low band. I think the Senator was at the launch last month in Teagasc's Ballyhaise College. That project is-----

I have another question but I just want to make it very clear that we are not questioning the efficacy of the RISPA's services or the merit of what its members do - far from it. They need support, and what they do is important. The launch in Ballyhaise College was exciting and wonderful, and the more such events the better. That is not the issue here. Our issue is whether this is a viable alternative to the national broadband plan. That is what our report will have to deal with specifically. Of course the witnesses' services are efficacious and do great good and there are beneficiaries - I know some of them - which I support. The question, however, is whether they can jump from this to an assumption that they deliver the entire national broadband service.

Mr. John Gartlan

It is a matter of scalability. RISPA is not advocating that it should deliver the entire national broadband service. We are putting forward an alternative argument as to how the speeds that are set out can be achieved in an alternative way that costs the State less money.

The final issue I wish to raise is relevant to people who are watching this or who will read about it in tomorrow's print media. RISPA made the point in its written submission that the projected cost of the national broadband plan has spiralled to €2.97 billion. We were led to understand by a number of other witnesses to the committee, by the Department, etc., that the State will get VAT on this back. That would amount to a reduction of up to €350 million. Do the witnesses accept that?

Within the €2.97 billion, approximately €527 million has been provided to avoid a scenario we have all been reading about recently relating to another large project, that is, to avoid the risk of escalating costs. The €527 million has been put in as a cushion such that the cost may go up or down but will not exceed the €2.97 billion. The likelihood is, with good management and the Department's rigorous protocols and controls, that the €527 million cost will be avoided. This takes the cost nearer to €2 billion. Furthermore, is it not the case that there is a clawback facility, assuming a high take-up? I think there will be a high take-up. When the ESB started to light up the homes of the hills of Donegal, among other places-----

Mr. Marcus Matthews

According to Robert Watt's memorandum, the Department of Public Expenditure and Reform strongly believes that the uptake would not be high.

I think it will.

I ask Mr. Matthews to let Senator O'Reilly finish. I will let him in afterwards.

Ireland is evolving and changing. Who would have thought 20 years ago that we would all have mobile phones? Who would have thought when the ESB started that everyone would want an electricity supply given people were almost afraid of electricity at the outset? This process will change and there will be an enormous take-up. This is the future. We have a political imperative; the witnesses have a commercial imperative. Everyone has an imperative to get this right. The €2.97 billion figure is not correct because of the VAT reduction, the cushion of more than €500 million to allow for escalating costs, which we do not anticipate will arise, and the clawback. In this way, the figure could realistically fall a little below €2 billion. Mr. Matthews said there will not be a high take-up. In the UK, one of the witnesses told us, the take-up was far greater than anticipated. I put it to the witnesses that the figure of €2.97 billion is not correct.

Mr. Marcus Matthews

Regarding the figure, the Senator said we must compare apples with apples, not apples with oranges. Our headline figures reflect the maximum and the worst-case scenario of having to build the model with an incredible volume of redundancy. One can talk about clawbacks and success rates, but that is like reading tea leaves as to what the future will be like. One must take the headline figure, the top figure, and presume that it could go that way, just as we picked figures and presumed that a loan scheme could cost up to €500 million and that connection vouchers could cost up to €412 million. In public policy planning, that is how one must look at this. There will be clawbacks and all these other mechanisms, but we does not know exactly how this will turn out. The bottom line is that the Senator is talking in billions of euro and we are talking in millions of euro. It is quite disturbing to me as a citizen that he would not be more prudent in examining the information and reading through in much greater detail the arguments that have been laid out. If we are talking about saving the country these large sums, I think that-----

We are questioning the proposition, as is our duty.

Mr. John Gartlan

That is perfectly fine, but subject matter experts are coming out and giving their evidence for the written submission about the plan, the credibility and the numbers. This warrants serious examination.

We subjected the other witnesses, on all sides, to a similar set of questions because the report must reflect that.

Did Mr. Gartlan say there are 19 sites in Cavan-----

Mr. John Gartlan

There are 19 sites in Cavan and Monaghan, most of them small - what we would call micro sites with something like what Mr. Neote has demonstrated.

How many premises would that cover?

Mr. John Gartlan

Currently, close to 4,000 customers in those two counties.

If RISPA is, therefore, trying to cover 5,400 customers, we are talking about more than 2,500 sites.

Mr. John Gartlan

Our current business model is that we go to a 15 km radius because we have set the bar at 40 Mbps to 50 Mbps, which was what was specified as the NGA speeds at the time when we designed the network. Of those 19 sites, three are main mast sites, in respect of which we pay rent and place our dishes on elevated hills and mountains. The rest are micro sites we have built ourselves.

Is planning permission required for those micro sites?

Mr. John Gartlan

They are planning exempt because they are just, for example, antennae on the sides of buildings.

Mr. Marcus Matthews

To clear the confusion, one can load multiple base stations onto a site. In our model we are talking about eight of these units per site. That is how we are able to cover the density in terms of the-----

So in the Cavan-Monaghan area we are talking about three masts. ComReg has said that for 5G, 6,000 masts are needed across the country. I want to ensure that people can understand this. When the witnesses talk about masts, are they talking about what ComReg is talking about? ComReg has given us advice on 5G or wireless technology.

Mr. Gartlan is stating that, as per what ComReg previously indicated, there would be three masts in Cavan-Monaghan. Is that correct?

Mr. John Gartlan

We refer to them as sites. A site would contain a broadband radio capable of giving service to 50 or 55 people. We can put up to eight of those on a site. Most of those sites are reasonably local and rural and we come to agreements with farmers to service the local area and they come to us daily asking for figures for people in their areas where there is no broadband.

ComReg referred to 6,000 masts in the context of the roll-out of 5G. How many of those kind of masts is RISPA talking about for its proposal?

Mr. John Gartlan

In Cavan-Monaghan, we utilise three telecommunication masts. They are on Sliabh gCleath, Croghan and Sliabh an Iarainn in the west towards the Leitrim border.

They relate to the 4,000 customers in the area. They are what I would call smaller masts.

Mr. John Gartlan

Yes.

That is for two counties. For 540,000, there would be many more masts required but Net1 does not have an exact figure in that regard.

Mr. John Gartlan

The masts would already be there.

Mr. Marcus Matthews

The three masts were not built by Net1. They house communications technologies from most of the different companies.

ComReg, the independent trusted adviser, has a credible reputation on this matter. We want to know exactly how many masts RISPA is proposing to install to serve the 540,000 people. The complexity of planning has been outlined by many individuals. We already know about the problems with mobile reception in some areas. Constituents contact us and we say that they need a mast in their area but there is an objection to putting up a mast for mobile reception. These are the political complexities surrounding this issue. That is why there are many questions about what exactly RISPA is proposing for the roll-out of masts and planning. I am not over-simplifying, that is our concern.

Mr. Marcus Matthews

The reason we put in such a small distance limiter in the model we put was that we would not have to build masts if the proposal was to be adopted and examined.

RISPA will not build any masts. Is that part of its proposal?

Mr. Marcus Matthews

As has been discussed, nobody wants masts. There are those of a not in my backyard, NIMBY, persuasion and objections. This technology can, as described by subject matter experts, be put on existing premises or, as Mr. List-Petersen described, even on trees. What matters is the length of the connection. If the distances are reduced, we do not have to deal with masts.

Is RISPA proposing not to have any masts at all?

Mr. Marcus Matthews

We would be proposing to have as close to zero masts as possible. We might have to build 20.

Is that with wireless?

Mr. Marcus Matthews

With this technology.

In other words, not with fibre.

Mr. Marcus Matthews

The key point is that we would have to do fibre extensions. We would have to bring fibre into certain rural areas. If we took somewhere like Wicklow, we might do four fibre extensions for the entire county and use broadband radio to connect the last mile network. Rather than digging up every road and town to run fibre everywhere, it would be extended where it makes sense and then we would use a flexible technical model which exceeds the target specification proposed in the NBP to deliver the services. To come back to the Chairman's original question about the masts, they are not desirable, the model was built in such a way that we would not get into building 6,000 masts. We were considering 20 or 30.

Mr. John Gartlan

They would be masts that probably exist at the moment and we would rent space through those providers.

Am I correct in saying that in its proposal RISPA could roll out wireless using the existing masts?

Mr. Marcus Matthews

Yes, provided the distance limitation that we specified is adhered to. That is why that mast infrastructure does not have to be used. It is used to cover huge distances and get a load of properties in the coverage arc, from the unit. If the distance is reduced, however, they can be located in a more targeted way. They cover fewer premises but the quality of service is much higher and more flexible and we avoid dealing with masts.

National broadband Ireland talks about a starting speed of 150 Mbps rising to 500 Mbps with a 1 Gbps option availability immediately for businesses. Is that not a good option for businesses?

Mr. Marcus Matthews

I would like two aspects of this point addressed, one, the technology cannot meet that 100 to 150 Mbps target-----

Which technology?

Mr. Marcus Matthews

The broadband radio.

That is RISPA's technology, which it is proposing.

Mr. Marcus Matthews

Yes, it does not meet the red line. The Chairman can comment on the red line of the technicalities in respect of that. The second point is about the gigabit connections. All of the companies here provide wirelessly extremely high-speed connections of 1 Gbps plus. It is possible to provide up to 10 Gbps with broadband radio technology. Businesses that want that type of connectivity already have it and have had it for a very long time. I could have 150 Mbps connection if I want to do teleconferencing or to work from home.

There are many businesses in rural areas that do not have it and that is our problem. That is what National Broadband Ireland is offering. It is not enough to say they have it if they want it. We want the roll-out of the NBP because there are many businesses in rural areas which do not have access because it is not commercially viable. That is why we are rolling this out. Mr. Matthews says RISPA cannot offer the same speed.

Mr. Marcus Matthews

I am saying that already, for many years, businesses that require that level of connectivity have had it and more than 1 Gbps.

Is Mr. Matthews stating that there is no need for it because businesses in rural Ireland can access it now?

Mr. Marcus Matthews

The statement was that it starts at 150 Mbps and will go to 500 Mbps and then to 1 Gbps services. It has been the case for many years, and some of the members of our delegation can attest to this if they like, that businesses which require multi-gigabit connectivity have had it for some time because they need it.

If the Chairman is talking about private households, the question is how many of them are taking multi-gigabit services. We discussed earlier how the uptake beyond entry level connections is very low. The technology is already there and subscriptions are available if people want 50 Mbps and 100 Mbps to their homes. I could have 150 Mbps now if I wanted it. I do not need it at all. I may need it in the future at which point the technology will have moved on further again. We have to look to see who needs that level of connectivity and when.

It is a question of empowering businesses and rolling it out to the regions and ensuring that people have it so that they do not have to be in the city. I suppose that is the whole point of the NBP.

When our guests talked about what they have to offer and the number of sites and masts they use now, I thought "That's great" but it is for a certain level of service. Net1 serves 4,000 in Cavan-Monaghan but the NBP will be for 540,000, which is very different.

ComReg is putting a price of €1.8 billion on that and 5,910 new sites, according to its evidence. How does Mr. Gartlan react to that? If I am misrepresenting ComReg, I will accept that. However, I do not believe I am doing so.

Mr. John Gartlan

Is the Senator referring to mobile coverage?

Mr. John Gartlan

I have not seen the ComReg report but it might be referring to microwave bands or high microwave 60 GHz bands with a range of approximately hundreds of metres to give those very high speeds. We saw the BBC demonstration that was carried out in London a couple of weeks ago with the launch of the EE 5G trials. The reporter had to sit with the telephone facing a certain direction to get it because the coverage was so small and was reflecting off buildings. Perhaps ComReg is referring to something akin to that. I am not sure because I have not seen it.

Mr. Marcus Matthews

What the Senator is saying in terms of mobile communications technology is the problem here. There is an issue that it is the same as the technology used by regional ISPs, but it is not. In terms of the cost and the number of sites, I have no doubt that is accurate if it was to be done via 5G, but we are not talking about 5G. We are talking about a technology that has been purposely designed and has been used for 20 years to deliver connectivity in remote and rural regions across the world.

Mr. Martin List-Petersen

With regard to the figures, the Senator is talking about 540,000 premises. That is the figure for premises passed that the NBP wants to achieve. When we talk about 4,000 to 5,000 premises that Net1 covers in the Monaghan and Leitrim area, that is 4,000 to 5,000 premises that are subscribed. There is also competition from ADSL, the copper technologies, mobile broadband and other platforms. Not every household in that coverage is subscribed to the service. It is apples and oranges again. One thing is uptake and the other is premises passed. One cannot just say that the 4,000 to 5,000 do not scale to the 540,000 nationwide because one must consider uptake. In addition, the Senator said he expects a high uptake under the national broadband plan. Even Open Eir, with a 300,000 roll-out, did not have a very high uptake. It did not anticipate that, for example, we are already covering many of these customers, who did not move because we could give them an adequate service or they were not interested in paying for the service needed to get the line into the property.

Mr. Marcus Matthews

On that point, it is interesting that regional ISPs have maintained their customer numbers in spite of the fibre roll-out of Eir's 300,000 intervention. It is something we were discussing a few weeks ago. It is quite intriguing. One can only come to one conclusion in that. As the number of premises took up that service, and it is reckoned that approximately 14% of premises passed during that Eir roll-out did, it would have been taking them predominantly from regional ISPs. The only way for regional ISPs to maintain their numbers would have been to extend their networks into areas that were not being served and acquire new customers. It shows quite a level of ingenuity on the part of regional ISPs that they are continuously adding customers to their businesses in terms of serving areas that are not served.

I missed the presentation because I was attending another meeting, but I read our guests' proposal and submission. I have a great deal of sympathy for the approach being proposed for a number of reasons. One is that I have worked in my constituency with a number of providers for a number of years. The NBP has been around for as long as I have been in active politics. It has not been delivered but certainly has been doing the rounds. I have worked with many different ISPs and have seen many different approaches and what has worked with communities where, perhaps, more technically proficient members of a community group used various ad hoc approaches and tried to improvise with varying degrees of success. I know from experience that there can be multiple ways to achieve broadband penetration into rural areas. It is not just fibre. There are many different technologies that work reasonably well.

One of the bugbears I had with the NBP was that these were not tried in further and greater measure. It is something I have said many times in this committee and in the Dáil. I have published legislation on it. The former Minister, Deputy Naughten, largely agreed with my legislation. In drafting a Bill and in various different testimonies such as this, I have said that it appears we are trying to do this the wrong way around. We are subsidising the winner of the tender to go into all the intervention areas. We are paying it a great deal of money to provide Internet to areas that cannot get it. Rather than stating that we will accept this as a given and we will give somebody a subsidy to do it, we should be asking why it is not happening already, why there is a market failure, understand the reasons for the market failure and tackle those reasons and the planning, regulatory and sharing of site anomalies. All sorts of things have been inventoried in the various task lists of different committees over the last number of years and I tried to tackle some of them in my legislation. Rather than saying we will fix all those issues, gripes and glitches and then see if the intervention area gets smaller, which it invariably would, we have taken an approach of saying there is a big black hole in an area so we will give loads of money to somebody to go and fill it.

As I said, I have a great deal of sympathy with the proposal. I agree with many of the points made. I am not convinced that we are achieving economies of scale with the current approach. By giving one tenderer a pay cheque to do it, we are not allowing for improvements in technology. There is no incentive to improve the process. There is no incentive to deal with a glitch between two different local area plans side by side in two different counties which have the same border. They are competing area plans. There is no incentive to choose the site sharing possibility. Instead of there being a common mast that is reused by many providers everyone puts up their own. There is no incentive to ducting along the side of a motorway. I understand it is cheaper to dig up the road and put in new ducting rather than reuse what is already there because of the rates the Department of Public Expenditure and Reform charges for access to ducting on the motorways. With all these crazy catches, there is no incentive to tackle them because we are going to give somebody loads of money to do it regardless. I have many issues with the proposed approach.

I have sympathy for what our guests are saying. They seem to be indicating, if I am reading their proposals correctly, more or less what I have just stated, namely, that there is a different way we can do it. One of the ways they are proposing contains a compromise of sorts. They are saying they will not give fibre to every home but that they will provide a variety of technologies, primarily broadband radio according to the submission. Perhaps there are other technologies as well. Perhaps there is a degree of fibre with the last mile being radio and perhaps a little 5G in the mix as well. The witnesses are indicating there is not, so it is primarily broadband radio, fixed wireless. They are saying they will do that and have a minimum standard of 150 Mbps. That will be delivered in two years at a cost €402 million. That is extraordinarily good value for money and an extraordinarily attractive proposition.

I am coming at it with an extremely sympathetic perspective. It bears out my direct experience as a councillor and then a Deputy working with companies in my area and working with local authorities and so forth. However, if something is too good to be true, it usually is. I would love to think this is the way forward, and there are many reasons that it might be, but if it was so obvious how come it is not happening?

Mr. Marcus Matthews

That is probably how I ended up becoming involved in this. I just could not believe it when the connection was installed. It was James O'Sullivan's company that did it. I was on ADSL for years and it was dreadful. He asked me how much speed I would like and I asked if I could get the Internet. I wanted it to work and not drop every two minutes and I asked if I could get 1 Mbps. He asked if I wanted 50. I was thinking that it would not work and that it would take weeks to be done, but he came up one afternoon and it was installed in two hours. I have never had to telephone the company about an issue of a technical nature. I carry out web conferences with the high definition video link with up to 40 participants and all the bells and whistles. There can be rain, hail, snow and fog in which one cannot even see the gate, but it does not make a difference.

When I heard about the NBP, I thought it did not make sense. Before long, I was involved in putting my public policy experience and various other skill sets to good use. Our submission provides case studies and figures, particularly in respect of Scotland, which tried the 100% fibre approach and got very good value for money at a cost of €442 million. It now has a top-up which will catch the final 5%. On things being too good to be true, other jurisdictions are bringing it in at this kind of efficiency and with these economies of scale.

On why the proposal was not raised previously, there were numerous consultation engagements on the part of the industry through the years. There are plenty of people who will attest to this and provide evidence and documentation in that regard. Two previous representative bodies also collectively engaged. One could say that it is appearing to the committee at a late stage. I would put that down to a lack of communication, possibly between the Oireachtas and the Department or on the part of the industry. The people in the industry are working very hard to serve their customers and they should not have to shout this loud to get heard. From my perspective in terms of public policy engagement, it is remarkable and worrying that this alternative was not examined as part of the cost-benefit analysis.

The criteria for a cost-benefit analysis and the public spending code require that one looks at one's project and an alternative and prepares a counter-factual that extrapolates what will happen if one does nothing in terms of how broadband access provision would progress. The NBP process did not really look at an alternative. The counter-factual had substantial issues in terms of 300,000 properties having to be removed. Two eminent economists stated that it does not live up to the stress tests. We are where we are. There is a tender process running. We do tenders for a reason, namely, to try to tease out problems and prevent bad decisions being made in the allocation of precious public funds. The process worked in one respect as a very credible proposal has come forward in time to be examined and for the current process to be halted for two or three months during which we can run a national mapping exercise. I am sure all of the regional ISPs would be very happy to provide their mapping data. We were looking at software last week that would probably allow us to do that mapping exercise and instantaneously figure out how many homes are covered for next generation access speeds and how many would need some type of assistance.

On whether it is too good to be true, I am of the opinion that we should put it to the test and stress test it before we make this decision and go forward with the biggest contract in our history. If we can achieve the same result in a shorter timescale for significantly less money, the Government has a fiduciary duty to examine that in great detail and robustly determine what is the best course of action.

We are trying to wrap up. Does the Deputy have another brief question?

I have a quick follow-up question. I thank Mr. Matthews for his response which makes a lot of sense. On putting it to the test, to what extent has RISPA or the providers it represents managed to penetrate into some of the areas in question? Has there been success in that regard? To an extent, an experiment has been running in recent years. If members of RISPA have managed to provide broadband in certain areas in a viable way which is worthwhile for the company, that is a success story in itself. If that has been done, are there things such as the glitches which I mentioned which could quickly be dealt with to enable the company to provide broadband in that way, which would turn the conversation around?

I invite Mr. Matthews to respond briefly.

Mr. Marcus Matthews

Yes, there are. We have outlined some of what would be required. We addressed the costs, including wholesale access. Backhaul is a major problem. The Deputy made the point that in many cases it is cheaper to dig the road than to buy it. That is a core issue that would have to be corrected. The subject matter experts who are present could tell the committee a significant amount about that, how it works and how it does not work compared with other countries. It is clear that we need to look at a regulator pricing model and, for backhaul providers that are directly or indirectly owned by the State, we need to consider implementing a ratcheting mechanism to drive those costs over time. We could tie that with an investment into the companies such that shareholders are not losing out and we avoid those types of problems while getting a good quality series of fibre extensions and a good scenario whereby regional ISPs can instantaneously start delivering more connectivity. That is a core issue to be addressed behind the scenes in this context.

There are other issues such as communications timelines in terms of different State entities. We would probably need to establish a communications management strategy, which is not dealt with in our written submission, and would address issues such as a fault or radio interference from another company. It is for ComReg to enforce the law and state that one cannot put something up in a particular place and start interfering with pre-existing equipment and other businesses. These types of things can take a very long time and that is a frustration for regional ISPs in the industry because their customers will question why their service has suddenly slowed down. It doe no good for the businesses to say that they can do nothing about it because they are responsible for enforcing the law. There are many issues behind the scenes that need to be fixed, but we could be here all day talking about them.

I thank our guests for attending.

The joint committee adjourned at 4 p.m. until 2 p.m. on Tuesday, 23 July 2019.