Sunday, November 24, 2024

Building the new wholesale network (Planning)

 

If an entity stays in this world beyond the fulfilment of its purpose it will only cause imbalance and anarchy.

 

-------Old Proverb


1. Introduction

In the last blog that I wrote about ODU-XC and how it is fading and giving way to OTN-Transport and IPoDWDM I mentioned the shortcomings and the reasons why ODU-XC in now a non-viable solution. This is especially to those who are in the bandwidth leasing business or in the CoC business. The wholesale business is marred with margins and ODU-XC does not provide a suitable option over economical options like OTN-Transport and IPoDWDM. In this blog we are going to talk about especially the nuances of planning of the wholesale network for the new telecom era.

The fact that we are trying to establish in this section of the blog is related to the planning part of the wholesale network. A wholesale network is very different from that of a mobile back-haul or an ISP backbone. A wholesale network is supposed to serve many functions and many kinds of network. This is why the planning of such a network that is business oriented is extremely complex and has a lot of challenges. We will explore the requirements, the challenges and the new aspects of planning that goes into building this wholesale telecom network.


2. The requirements of a wholesale bandwidth provider

It needs to be clear that a wholesale bandwidth provider is not the owner of the content or is not generating any revenue from the content delivery. They are just pipes  or media who carry the traffic from one point to another with a standardized SLA. They are not owners of QoS or QoE, they are channelizers or media enablers and they are also paid in the same way. Thus the requirements of a wholesale provider is extremely flat and bland (if I am excused to say so)


2.1 Optimized bandwidth infrastructure

The wholesale provider has to play between the CAPEX and the OPEX of the solution that is implemented. The biggest asset it has is the fiber and that it has to capitalize to great extent. This capitalization comes from being able to have an infrastructure layer that will be able to carry large amounts of channels across the distance by consuming least pairs of fiber. Here we are focussing on delivery of bandwidth rather than the QoS of it, which is generally controlled by the end user. Thus the infrastructure is more important. Flex Grid systems with the ability to carry more bandwidth in flexible channels is definitely the good way to go forward. Flex Grid infrastructure provides that quintessential platform on which the premise of a wholesale business in the modern era can be based on.


 2.2 Cost effective transport modules

After the flex-grid ecosystem is created there needs to be a focus on cost effective transport modules that need to be provided in order to carry the traffic from one point to another. In this regard the focus should be on having modules that can aggregate more number of high bandwidth clients in a higher rate line output. This will result into the optimization of the revenue. Most of the times due to the ODU-XC and its inherent separation of the client and the line modules the cost of the solution increases, however if we look at the efficient muxponders that can provide the same solution at a lesser cost then it becomes very cost effective. 

Muxponders serving as simple transport modules


The muxponder that we see in the picture is a clear depiction of how the entire functioning of the ODU-XC can be compressed to a single module that can fit as a service card in the optical photonic shelf. This is an ideal situation for the wholesale bandwidth provider. The muxponder provider the end to end services for the wholesale provider to an extent that the CAPEX is kept minimal and the photonic shelves, that are extremely efficient in power and space, can handle the services. 


2.3 OPEX Efficiency

Another very important need for the wholesale bandwidth provider is the OPEX efficiency of the solution. The wholesale provider makes a full account of the cost that includes CAPEX and OPEX and the latter is a very important aspect as it is recurring in nature. Costs such as power, space, maintenance etc is a major part of the OPEX of the solution. OPEX efficiency, thus, forms an integral part of offering for the wholesale provider. Studies reflect the fact that the classical ODU-XC is extremely OPEX hungry and does not justify the RoI for the wholesale provider. In some cases the OPEX runs out to be more than three times of what it ideally needs to be. For wholesale bandwidth providers this situation is no less than a nightmare where the bandwidth has to be delivered in a wafer-thin margin cost structure. So definitely there have to be alternatives for the ODU-XC for such purposes so that economical ways of bandwidth delivery can be achieved. 


3. Why MNO requirements cannot be imposed on a wholesale network? 

Under normal circumstances it is always seen that the planning philosophy of MNOs often hijack the requirement scene of the wholesale bandwidth provider. This is the first step towards creating a loss-making network for the wholesale bandwidth provider. The wholesale bandwidth environment needs a network planning structure that is very different from the MNO planning structure.  There are certain factors which necessitate a separate design philosophy for the wholesale networks. 

3.1 Wholesale providers are not owners of the end-service

In the case of MNO the ownership of the end service and to an extent the content access lies with the MNO. The MNO is supposed to be the end to end provider of the service whether it is the wireless access or the back-haul. In the case of the wholesale bandwidth provider, this is not the case. The wholesale provider only owns the path and that too a part of the path than owning the entire service. The wholesale provider is like that of an active wire on which the service needs to pass. Since the ownership is limited the revenue expectancy is also of that category.  Therefore the technology that needs to be chosen under this case has to be extremely efficient in terms of cost whether it is CAPEX or OPEX. 

3.2 There are diverse routes in place of complex protection schemes

You agree or you do not agree the end customer will not solely depend on one wholesale network for the carriage of the bandwidth. The premium thus a customer will pay to you for a service that is protected by your network will be negligible in this respect. The focus should rather be on selling diverse routes that are unprotected while letting the protection mechanism work on the end customer. Diverse routes lead to a situation where the spectrum of the DWDM is more optimized and more static. Contrary to the cases of WSON or ASON where most of the resources are used up in protection. If we talk about WSON then there is a bit of economic sense involved as the resources are only blocked on the virtual channel level, however the ODU-XC ASON is of a really different level. Here we can see that the resources are pre-reserved in terms of hardware and interfaces. This increases the CAPEX dramatically.

The end customer on the other hand also ensures that there is enough diversity in the level of wholesale providers also. This is to ensure that if there is a major problem in the network of one wholesale provider the other wholesale provider can carry this bandwidth. This is a common phenomenon observed in the hyper-scaler sector where diversity is a key issue. 


Example of diverse route by the same provider


As seen in the figure above, we see that the diverse routes are provided here but the provider is the same for the wholesale bandwidth. The onus of the protection however is lying with the end service provider who does the switch between the main and the alternate path. Here the situation is a kind of a win-win.

For the wholesale bandwidth operator it is a sale of 2X bandwidth with low consumption of resources as there is no complexities of protection involved in the provisioning of the bandwidth. 

For the end customer there is a control of the main and the alternate path with full transparency. In addition to that the say 2X bandwidth can be used for passing best effort traffic when both the paths are available.

There is also a situation which is more ideal for the end service provider. Here the service provider does not rely on one wholesale entity but has two entities for the diverse carriage of the bandwidth. The diversity is more robust over here because here there is less reliance of the efficacy on one operator. 

Diverse route provided by diverse provider to the end customer

If we see the picture above then the situation is more pragmatic as now there are diverse routes provided by diverse providers. So for the end service provider instead of having an independent network of its own it can have a platter of bandwidths provided by different wholesale provider. The wholesale provider on the other side can be less robust and focus more on spectrum optimization rather than on the protection path.

Generally the end customer would seek such an arrangement with the wholesale provider(s) in order to maintain proper robustness in the network. This in a way lowers the load of onus on the wholesale network and thus can provide a more price optimized offer for the end customer from the wholesale side. 

3.3 Cost sensitivity is at its peak

As we can see from the above section that wholesale bandwidth is an extremely cost sensitive business, so the planning also needs to be cost sensitive. While the demands of technologies like WSON, ODU-Flex can be a sort of a wish-list, they do not have any kind of commercial leverage. So with the dwindling margins of bandwidth delivery there needs to be an innovative approach and cost effective approach in planning that is very much different from the planning that is involved to build a mobile network backhaul. A wholesale network will need to eliminate most of the features that are bulky in nature and are not required into the network. Focus has to be for delivering good quality bandwidth from point A to point B with all the SLA covered.

The planner has to realize that the margins in the wholesale network is less and it is more of a volume game. So, the mandatory aspect of a network needs to be covered for the wholesale network. More and more bandwidth is disposed off as a commercial entity in less and less cost incurred that will sustain the business to a greater level than creating space for more wishlists for the future. 


4. New aspects in the planning of the wholesale network

Now that we have understood the challenges that lure for the planning of the wholesale network let us delve into some points that consider the new aspects that are involved in the planning of the wholesale network. These new points will be extremely essential to build up wholesale networks rapidly and with extremely less cost. Before we venture into the building of the wholesale network we need to consider two fundamental principles on building it.

1. Faster network roll-out and faster delivery

2. Less CAPEX and OPEX involved in building the network. 

3. Prioritize service from the business point of view. 

Considering these fundamental principles we will explore some of the points that will make wholesale network building extremely efficient and cost effective. 


4.1 Build the infrastructure

The biggest requirement for a wholesale network is to secure the infrastructure. This forms almost 80-90 percent of the effort of building the wholesale network. More the delay in building the infrastructure more will be the delay in the delivery of the services. When we talk about the infrastructure we are talking about the following things. 


1. Fiber network readiness. 

2. Flex-Grid DWDM network with OTDR monitoring and fiber health monitoring. 

3. Proper channel/grid planning. 

4. Provision for the Alien wavelength/Spectrum on the network. 

This actually wins most of the battle. It enables the wholesale entity to start providing services on the day one with wavelength or spectrum basis. 


4.2 Identify high-bandwidth services

In a business of wholesale to sustain and to grow it is important to have an outlook and outreach towards cherry-picking the profitable kind of bandwidth deliveries that provide immediate revenue with minimal capex and opex. Our infrastructure is already done and ready. We can already provide spectrum and thus bandwidth on this logical infrastructure with minimal movement of CAPEX and OPEX. However, if there are operators who insist us to take ownership of the traffic segment also, the key guys would be the ones who are demanding bulk bandwidth. Let us take the example of an operator that is demanding bulk bandwidth across two points of 4X100G or more. Now for this the investment is quite small. 

We can assign muxponders that will aggregate this over a 400G or a 800G wavelength and provision across the two ports and be done with. This just is a module addition in the existing OLS not affecting the CAPEX and the OPEX as opposed to a full fledged system addition.

Thus, it becomes a very gradual addition to the infrastructure that we already have built without any additional hassle. Identifying the inventory is also easy as this mapped to the end user. Operations are not a hassle. 

4.3 Looking for those important low-rate subscribers

It is not rocket-science to understand that when you build an information superhighway then looking for a low-rate subscriber who wants the traffic also to be a part of your solution is a much expensive affair. For one such 10G there needs to be a line out towards 100G line or more and this built up can be a hassle for the network. Low rate customers can be the most expensive ones when it comes to the cost of delivery and the cost of maintenance. There cannot be a parallel line network for the 10G and for the 100G+. Therefore it is necessary to find the customers that are more suited to the network. A low-rate customer can be a very good strategic acquisition considering that there will be expansion in the future with growth of experience and the growth of the customers that the end customer has in this aspect. 

Generally low rate customers are coming from the enterprise sectors and more than the bandwidth what they require are. 

1. Reliability

2. Quality

3. Value addition

While the reliability and the quality part is already covered in the equipment and delivery of the wholesale provider the part of value addition is often missed out. Let us take some examples of the value addition that can be done for these enterprise customers. 

1. Customer network management (CNM)

2. Bundled services (Voice+Data Breakout)

3. SDWAN

4. Data monitoring and security. 

5. Encryption. 

Generally, when these value add services are provided with the wholesale bandwidth provider then there is a great deal of attraction for the enterprise services. 


4.4 Alien is very friendly these days

Today there are many service providers, MNOs, MVNOs, ISPs, MSOs etc who have the end device that can support the coherent but do not have either the DWDM infrastructure or the dark fiber. Under these circumstances they are looking for an infrastructure that can carry their traffic, which is already coherent in nature, on a specific wavelength or a spectrum. Hyperscalers, especially make a demand for such a service where they want just a chunk of the spectrum that is available in the DWDM infrastructure.

It is very important thus to plan for the alien spectrum and services like that because these are extremely quick and low maintenance businesses that a wholesale bandwidth provider can achieve. 


4.5 Automation is the key

Quick delivery, Express provisioning, Flexible BwoD and Self Healing networks. These are some of the important buzzwords that are in the market. The answer to all these buzzwords is “Automation”.  The automation has to be primarily in the network management system that the provider is choosing to manage the network. 

  • Configuration automation to ease the delivery of the services. 
  • Bandwidth on demand
  • Fault automation with AI/ML to have a more self-healing network. 
All these aspects are to be seen when planning for a proper management system for a wholesale bandwidth ecosystem

5. Summary

Thus, we understand that the planning of the wholesale network cannot be congruent to the planning of the MNO network. There has to be expertise involved considering all the new aspects of network building.

To cut a long story short. The planner of a Mobile Back-haul is Network-centric but the planner of a wholesale network is network-centric and business-centric. So, in order to get into the planning of a profitable and revenue yielding wholesale network such attitudes have to be developed.

 

I will be adding more of wholesale networks in my next blogs. Keep watching. 


Cheers, 

Kalyan

No comments:

Post a Comment