CSPs must go deep on network visibility

Staff writer
25 Sep 2014
00:00

Andy Huckridge, director of service provider solutions and SMEs at Gigamon, explains how a triple challenge of new technologies will change how service providers architect their networks

LTE Insights: What are some of the biggest concerns facing LTE operators in terms of new technologies?

Andy Huckridge: Service providers are contemplating how to de-risk the rollout of several new technologies, namely VoLTE, network virtualization and 100G transport links. Collectively, this triple challenge of new and largely untried technologies presents a major change in the way service providers architect their networks and how services will be deployed across them.

Carriers start by deploying one or two of the technologies and end up deploying all three, which causes a resource crunch. Carriers now confront a deep level of inter-relationship and interdependency between all three of these new technologies, such as the effects of provisioning, moving or de-provisioning a virtual server or virtual network element on a real time service such as VoLTE or video, for example.

The resource crunch is not obvious from the outset, when an operator over-extended in capex will find their opex affected, forcing the issue of visibility. Since two of the three new technologies (network virtualization and 100G transport) remove visibility from the network, the underlying problem is further exacerbated as many carriers haven’t realized the implications of deploying these three new technologies.

How does all this impact their network architectures and monitoring capabilities?

Although the three new technologies are complicated and largely untested, VoLTE is probably the most troublesome new technology to be implemented in recent years. It often needs a core and transport link upgrade to guarantee QoS needs. The reason is that voice (based on the RTP protocol) is a sensitive service to carry over packet-based networks. As for 100G, there are no 100G tools available today, and no way to tap 100G network segments effectively. With the introduction of NFV, layers of added complexity are placed on top of traditional virtualization, and none of them have inherent monitoring capability.

Related content

Comments
No Comments Yet! Be the first to share what you think!