Two companies have made launches today aimed at enabling service providers to achieve better performance control in their mobile backhaul networks.
Accedian Networks is unveiling its SDN and SON-friendly “SkyLIGHT” performance assurance architecture. The aim of SkyLIGHT is to enable the more dynamic and flexible assurance and optimisation of links across the mobile backhaul network.
SkyLIGHT sees Accedian introduce a new API capability it is calling Network State+, that will deliver data (collected by its Performance Assurance platform from software agents and aggregation-layer nodes) up into SDN controllers and SON engines.
Patrick Ostiguy, CEO, said that the aim of SkyLIGHT is to create an architecture that will allow for the real time optimisation of backhaul networks that serve multi-vendor heterogeneous networks across a variety of physical access types.
Currently, Ostiguy said, elements such as RNCs and MSCs that are responsible for controlling services down to the access layer are trying to “cascade” services through a variety of links, all with different quality characteristics. That can mean, in effect, poor visibility into the performance conditions across the network, leaving links either underutilised or congested.
“What we are proposing is to say: what if the actual router now has visibility of the performance of the backhaul that is in front of it, in real time, before trying to send traffic downstream, and in real time modulate and shape that traffic to fit these varying types as they modulate over time. That would create the ability to fill pipes better, run them hot, and make much more out of the current wireless backhaul infrastructure.”
“What we are doing is, in a standards-based fashion, extending ownership of performance assurance to multi vendor base stations within a solution architecture.”
Ostiguy said the SkyLIGHT architecture is built on three elements
1. A software agent in the access point, that could be a miniature or virtual agent, that “instruments” the network, providing real time performance data.
2. A performance assurance node, typically sitting at the “head end”, the RNC or MSC, that collects the aggregated data collected by the miniature modules.
3. A software platform that acts as a bridge to higher layer SON and SDN controllers, tooled with Accedian’s Network State+ API.
Ostiguy added, “We’re seeing operators deploy SON for their multi-vendor access networks and we’re sure they’d like to do this for backhaul, which often sits in the same box as the access point. In terms of managing RF budgets, backhaul link budgets, interference and so forth and it’s totally important for them to apply that same thing to backhaul – hence the concept of Mobile Backhaul SON.
“Backhaul SON for us is something that needs to have that Network State+ level of information: very accurate one way latency, jitter, packet loss, bandwidth availability, packet reordering. It’s a much richer set of transport layer metrics that are not currently available in vanilla SDN types of implementations.”
Ostiguy said SkyLIGHT would also enable and benefit from the move to virtualise network functions – particularly of network management applications. “Operators are tired of buying a different appliance for every EMS and OSS they deploy and the virtualisation of these network management applications means this conversion happening from two sides and meeting in the middle, and that’s where we are aiming to work with SkyLIGHT.”
Telco Systems is also advancing its presence in the mobile backhaul space with the announcement of a 10 Gigabit Ethernet product for aggregation layer of mobile network.
The company has today announced the T-Metro 8001 10GE MPLS/Ethernet switch.
Irit Gillath, VP Product Line Management, of Telco Systems said that the 20 port unit offers twice the capacity (200GE) in a device a third of the size of any competitive product, at 50% less power consumption. The device offers license free (meaning operators can choose and deploy as preferred) MPLS/Ethernet in a 1RU chassis.
People know about using MPLS in the core but they are concerned about the price and complexity of deploying that to the edge
Gillath said that planned small cell deployments within macro nodes would drive the need for 10GE capability and MPLS all the way to the edge of the network.
“People know about using MPLS in the core but they are concerned about the price and complexity of deploying that to the edge. These two things are solved with MPLS at aggregation layer and the demarcation layer by building smaller, denser products that include MPLS/Ethernet in the same product license free. A lot of [competing] solutions are either Layer 3 based or started life as L3 MPLS devices in the core and have been downgraded into access devices – but with a design that is still L3 based, which consumes more power, memory, CPU and so forth. Our combination design gives the ability to run L2 and L3 features [OSPF, iS-IS, LDP. RSVP-TE) in a 1RU chassis, which is something unique in the market,” Gillath said.
The product also offers hardware-based hardware-based OAM, including Y.1731, RFC2544 and Y.1564 capability, for SLA measurement and assurance. Gillath said that the product can also generate testing flows, towards the customer or into the network.