De-Risking New Service Provider Technology Deployments – What are the Deployment Inter-Dependencies?

By: Andy Huckridge, Director of Service Provider Solutions Andy Huckridge

Understanding the Technology Interdependencies

In the previous blog post we outlined the interdependencies between the three new technologies: IP Voice / VoLTE, high-speed transport pipes, and carrier network virtualization. These inter-relationships will make it difficult for the operator to gain confidence in rolling out each new technology, as well as challenging to pinpoint the source of problem areas. In this blog post we will detail a number of those interdependencies for further discussion.

What are the Interdependencies That Will Drive the Triple Challenge?

The below diagram shows the stages of a technology deployment and rollout, and denotes that whichever technology is used to start the process, resource constraints are experienced by the need to roll out the other two Triple Challenge technologies. Independent of the starting technology, the interdependencies and technology inter-relationships will cause the rollout of all three.

TC_3

Starting with the deployment of IP Voice / VoLTE
Deploying VoLTE leads to greater density 10GB, 40Gb or 100Gb transport pipe deployments
  • More data from packetized voice causes the need for greater bandwidth in order to guarantee dual bandwidth voice-RTP QoS
  • More data is driven into the core by the 2G to 4G/LTE RAN conversion which requires more bandwidth for enhanced LTE services
Deploying faster transmission pipes such as 40Gb or 100Gb leads to Carrier Network Virtualization
  • Removal of the final bottleneck: if the pipes are wide enough, the last bottleneck will now be the network elements (NEs) themselves
  • Do more while preserving existing CAPEX spend: Operators are able to go with “white box” and bare-aluminum solutions over single-use, non-upgradeable, dedicated, and often proprietary solutions
Deploying IP Voice / VoLTE leads to the deployment of Carrier Network Virtualization
  • New services will be brought out as virtual network functions (VNFs) on NFV-enabled networks. The ability to deploy a traditional upgrade on a legacy network may be short lived.
  • By virtualizing first, VoLTE could be deployed as part of a virtual EPC (vEPC) where a virtual IMS (vIMS) core could be deployed as a VNF.

 

Starting with the deployment of Carrier Network Virtualization
Deploying network virtualization leads to the deployment of IP Voice / VoLTE
  • Efficiencies can be experienced by collapsing multiple different service cores together
  • Carrier Network Virtualization allows cost reductions in the vEPC and new services to be deployed such as IP Voice / VoLTE
Deploying IP Voice / VoLTE leads to the greater density 10Gb, 40Gb or 100Gb transport pipe deployments
  • More data from packetized voice causes the need for greater bandwidth in order to guarantee dual bandwidth voice-RTP QoS
  • More data is driven in to the core by the 2G to 4G/LTE RAN conversion which requires more bandwidth for enhanced LTE services
Deploying Carrier Network Virtualization leads to greater density 10Gb, 40Gb or 100Gb transport pipe deployments
  • Virtualization of network elements with greater processing throughput pushes the bottleneck elsewhere
  • Now that compute and storage are elastic, the pipes have to be upgraded to deliver low latency, bandwidth contingent, transport and content delivery assurance, and be able to handle RTP service QoS

 

Starting with the deployment of greater density 10Gb, 40Gb or 100Gb transport pipes
Deploying greater density 10Gb, 40Gb or 100Gb transport pipes leads to the deployment of IP Voice / VoLTE
  • If the converged core is already running at increased speeds such as 40Gb or 100Gb, then the RAN coverage can be increased. Therefore, the 2G RAN is converted to VoLTE to enable the expansion of 4G/LTE data which produces more operator revenues
  • If 4G/LTE RAN is already running at increased speeds such as 40Gb or 100Gb and has available bandwidth, then IP Voice / VoLTE can be deployed. This allows for the re-farming of the 2G RAN to enable expansion of 4G/LTE data which produces more operator revenues
Deploying IP Voice / VoLTE leads to Carrier Network Virtualization
  • Reduced CAPEX spend by deploying a virtualized core, vIMS, instead of a traditional legacy IP Voice / VoLTE deployment
  • Virtualization will allow future ease of software upgrades, as well as ease of offering further enhanced or new software-defined services
Deploying greater density 10Gb, 40Gb or 100Gb transport pipe deployments leads to Carrier Network Virtualization
  • The cost of deploying wider transport pipes forces the move to virtualize to recover costs that would have been spent on single-use, often proprietary network switching or routing technologies. Use of SDN and NFV can save costs associated with routing and switching and allow the network to become more flexible
  • The ability to provide more efficient and elastic content farms and Big Data analysis is driven by high speed links
  • Elastic storage and compute are needed to power an operators future needs, even to offer cloud capabilities as a service to their residential and business subscribers/customers
  • In order to overcome subscriber bandwidth issues at the network edge, virtualized services are needed such as video transcoding, bandwidth treatments on-the-fly, and service treatments on-the-fly such as “throttling as a VNF”

How Will the Interdependencies Cause Network and Service-Related Issues? 

In the previous section we have demonstrated the interdependencies between the three new Triple Challenge technologies. Here we will explain the unique capabilities of a Unified Visibility Fabric™ architecture (UVFa) and how deploying one can bring a new insight to modern monitoring: understanding the inter-related deployment dependencies via cross-silo monitoring, allowing you to find the “needle in a haystack” faster, and oftentimes to vastly reduce the size of the haystack altogether.

In Order to Correctly Monitor the New Technologies, It Is Important to Understand What Is Needed and Why

  • IP Voice (VoLTE/VoIMS/VoWiFi) being based on RTP is a very sensitive service, complete visibility from edge to core is needed to debug complex transport/service layer inter-related issues
  • Bonded 10Gb, 40Gb, and 100Gb transport needs advanced processing across the fabric. Edge filtering and data optimization get the most out of the attached tools. Specifically today there are no tools capable of connecting to, nor monitoring 100Gb transport pipes
  • Carrier Network virtualization is a complex set of new technologies with no built-in monitoring capability. To deploy SDN or NFV is to remove the visibility from a large part of your existing network

Specific Issues Related to IP Voice/VoLTE

  • Effects of bursty traffic types and other RTP traffic types in the same transport pipe
  • Effects of server virtualization, network function or network element virtualization on RTP-based voice traffic
  • The effects of dynamic loading on RAN backhaul and RTP traffic QoS requirements

Specific issues related to 40Gb & 100Gb transport pipes

  • Effects of virtual servers being provisioned and de-provisioned, causing unpredictable traffic bursts
  • VNF provisioning overhead and monitoring needs
  • Multiple standards and changing technology associated with 100Gb transport pipes

Specific issues related to Carrier Network Virtualization

  • IP Voice/RTP QoS requirement overhead and associated transport pipe related issues
  • Effects of huge traffic draw on services and virtualization traffic-induced burstyness
  • Effects of vMotion and effects on other VNF’s / SDN controller decisions resulting in knock on traffic delay / jitter / latency or more generic throughput issues such as traffic fragmentation

Conclusion

There are clear interdependencies which will emerge when trying to deploy the Triple Challenge technologies. Monitoring can play a great part in de-risking the deployment of these three new technologies, and will allow service providers to fully understand these technology inter-relationships before deployment such that when trouble shooting, it is easier to find the real needle in the correct haystack.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s