TM Forum Community

 View Only
  • 1.  End to End user experience

    TM Forum Member
    Posted Feb 02, 2017 16:49
    With the communications industry being formed by varying Supplier/Partner levels such as ISPs, Service Providers, Core Network Providers and Access Network Providers, who each have a component of the network capabilities (Layer1, Layer 2,…), then how will SDN/NFV realistically be able to provide a role that will give a good End to End User Experience?

    ------------------------------
    Pete Bains
    Ultrafast Fibre Limited
    ------------------------------


  • 2.  RE: End to End user experience

    TM Forum Member
    Posted Feb 07, 2017 08:34
    Hi Pete,
    Very valid point on E2E user experience. Network Services are realized across different tiers of networks that you have sighted like edge, core, transmission, IP etc. SDN/NFV is just the virtualization of these network functions across several tiers. Industry has been grappling to get E2E user experience.
    Particularly, the way SDN/NFV could possibly contribute to better E2E user experience is by the virtue of having an ability to do real time correlation and analytics of network statistics and events and apply policy and rules that trigger automated workflows for preemptive and corrective measures. Having following functional blocks; this could be realized in the software : 
    1. Common Information Model (TMF SID; which already supports additional data entities for SDN/NFV)
    2. Big Data Analytics : Have as many data sources, both static (traditional OSS, NMS data) and dynamic (software probes based real time streaming for SDN/NFV) to gather sizeable and meaningful data sets.
    3. Policy Control with Rules and BPMN engine for close control loop. 

    Hope this helps.

    ------------------------------
    regards,
    Dhananjay Pavgi
    Tech Mahindra Limited
    ------------------------------



  • 3.  RE: End to End user experience

    Posted Feb 09, 2017 06:59
    In addition to Dhananjay's comment, with current set of network features, the service configuration is more static in nature hence the service profiles are limited.  With network virtualization and programmability, the customer will have more control on the service. With reduction in "Time to market" time and Fast Fail concept, SP can easily enhance or modify the service more align to customer's need.  

    ------------------------------
    Ashish Doodhwala
    Infosys Ltd.
    ------------------------------



  • 4.  RE: End to End user experience

    Posted Oct 22, 2020 11:08
    Very true Ashish. Additionally, With more projects now adopting agile delivery, its very critical that we always have a working model in place and unless fail fast approach is adopted the improvements cant be put in place or by the time they put in place it there is a risk of those becoming irrelevant. So now customer experience can't be realised if they are considered outside application boundary. Design has to be user centred and continuous feedback to improve the working model. Additionally, High speed connectivity i am sure is going to make a paradigm shift and super fast fibre and 5G will play big role in improving E2E experience.

    ------------------------------
    SANJAY KUMAR SINGH
    Infosys
    ------------------------------



  • 5.  RE: End to End user experience

    TM Forum Member
    Posted Feb 10, 2017 07:31
    Hi,

    Dhananjay brought in a very important happening need of big data analytics - if I can add....I would like to improve E2E experience by  using the NFV /SDN for Radio network optimization,end user quality optimization and also terminal power consumption optimization. this will encompass improving call initiation rate ,set up rate  and spectral efficiency. the crux will be load balancing in real time using big data .

    Avadhut



    ------------------------------
    Avadhut Deshpande
    Persistent Systems Ltd
    ------------------------------



  • 6.  RE: End to End user experience

    Posted Oct 22, 2020 11:08
    I am sure neural network machine learning with additional deep learning will help extract the network info in more meaningful way and discerned to plan better in terms or both fail-safe development and proactively monitoring.

    ------------------------------
    SANJAY KUMAR SINGH
    Infosys
    ------------------------------