It feels like 2006 once again: vendors are creating fanciful and colourful presentations about SDN and operators are discussing about the need to move from silos to horizontal platforms and networks. In a way, almost the same story was told 6 years ago for IMS, but deployments were far smaller than expected. So is SDN following the footsteps of IMS?
Before answering this question, it’s worth looking at the telco SDN story from the beginning and why the market is a particularly interesting – and ripe – environment for SDN.
What is Telco SDN?
Telco SDN seems to be everything that is described by the following mantra: network elements implemented in software running on commoditized IT platforms. Each vendor has a different implementation, strategy and even definition of SDN, and there seems to be confusion in the market regarding both the actual deployment of SDN in the telco network and the value proposition of the new technology. For example, one vendor claims that in a true SDN, every element in the network needs to run in software (as far out as the antenna), while other vendors consider that an SDN should be implemented in the data centre for control/data plane abstraction.
In the IT domain, SDN largely refers to OpenFlow (although SDN is theoretically a much broader concept), where the control and forwarding planes are separated to enable more flexible management of traffic through software programmable routers. Virtualization is usually considered to involve running several instances of servers in the same data centre, a practice which is quite well established in the cloud market. However, both SDN and virtualization may refer to different network aspects or techniques, which depend on vendor strategy, network environment and legacy.
To summarize, telco SDN currently refers to the replacement of network components with software running on commoditized IT platforms. Several telecom network elements are particularly interesting for this scenario:
- Legacy and out of date equipment: Instead of having to maintain, support and upgrade equipment that is enabling legacy services, SDN can replace their functionality in software which runs on standard IT platforms. Network elements that are in this category include Broadband Remote Access Servers (BRAS) and PSTN network elements.
- Processing-hungry infrastructure: Software driven elements that can allocate processing capacity real-time are very relevant, so that overprovisioning can be minimized. Such elements include video or traffic optimization gateways or anything that requires a high processing load.
In order to address the cacophony of different definitions for telco SDN, operators have setup the Network Functions Virtualization (NFV) forum which is described by the following:
Network Functions Virtualisation aims to address these problems by leveraging standard IT virtualisation technology to consolidate many network equipment types onto industry standard high volume servers, switches and storage, which could be located in Datacentres, Network Nodes and in the end user premises. We [the NFV] believe Network Functions Virtualisation is applicable to any data plane packet processing and control plane function in fixed and mobile network infrastructures.
We would like to emphasise that we see Network Functions Virtualisation as highly complementary to Software Defined Networking (SDN). These topics are mutually beneficial but are not dependent on each other. Network Functions can be virtualised and deployed without an SDN being required and vice-versa.
Key operators, including AT&T, Telefonica, Deutsche Telecom, Vodafone, France Telecom and Verizon Wireless and several vendors including Alcatel Lucent, Cisco, Huawei, IBM, Juniper Networks and NSN are founders of the NFV forum. An interesting departure from the standard membership model of telecom bodies is that IT vendors have joined, indicating that there is considerable potential in the telecoms environment.
The member list of the NFV forum can be found here and the white paper outlining the mission of the NFV here.
SDN and NFV arrive at an appropriate and interesting time for carrier networks. The following reasons are key factors for the increasing interest in telco SDN:
- The telco and IT environments are merging and there is a requirement for clear definition –and perhaps standardization – of telco SDN.
- Following LTE, infrastructure vendors have considerably fewer opportunities for revenue growth. LTE-Advanced is expected to be an incremental revenue driver while small cells will not be able to match revenues for nationwide LTE deployments. In a way, telco SDN is the next potential revenue driver for infrastructure vendors.
- Operator cloud strategies are evolving and carriers are deploying data centres, where the IT version of SDN (primarily OpenFlow) is deployed. It is natural that this will propagate to network elements, which may be implemented in software in these data centres.
- The increasing drive for lower OPEX, less reliance on over-provisioning, ability to react to unplanned events and better efficiency in integrating new services (or partners) are all becoming a necessity. These are also all advocates for the flexibility and scalability that SDN offers.
On the other hand, telco SDN currently has some serious challenges to overcome:
Each vendor has a different definition of SDN, but hopefully NFV will force vendors to align to its vision.
- NFV compatible (or standardized) elements will arrive in the market at least 1-2 years from now.
- NFV elements may require forklift upgrades for existing infrastructure, where there may not be a clear revenue opportunity, but only cost savings.
Is telco SDN the new IMS?
Telco SDN essentially promises the same thing that IMS promised 6 years ago: horizontalisation of the network without clear revenue opportunities. However, telco SDN (and the NFV) have serious advantages over IMS: SDN is already being implemented in the IT domain and operators and vendors will have learned from their involvement in IMS. Also, SDN is attempting to enter the market by converting practices and technologies from the IT domain so that they can be applied in the telecoms environment. Contrary to this, IMS was a completely new – and very optimistic – concept.
The future is brighter for telco SDN. Although vendors may be currently marketing it similarly to IMS, operators will surely make sure that outcomes of the NFV address their key concerns – and tone down vendor hype.