Nokia and ARM are at the heart of a bid to revamp the ageing TCP/IP stack to make it better suited to networks that need to operate at very high speed and/or low latency. Among the plethora of industry alliances at the intersection of telecoms, IT and IP, this looks like one with a genuine contribution to make to the evolution of future standards, including 5G.
The network connection itself is only a part of the total telecoms platform these days, and in the evolution of 5G, it will be critical that the 3GPP’s core standards are closely aligned to work being done in the IP, cloud and data areas too. LTE may be IP-based, but it was a complicated process to get IP and mobile technologies fully married and as networks get faster and more virtualized, there is a real risk that the venerable TCP/IP stack is failing to keep up.
In 5G, there may be the opportunity to develop the next generations of these two key network standards in parallel – and indeed, that will be crucial to success and deployability, given that one important reason for developing 5G at all is to enable the next waves of the Internet, the Internet of Things and the future Tactile Internet.
Revamping the TCP/IP stack
Many believe that a critical success factor for 5G will be a fully revamped TCP/IP stack, optimized for the massively varied use cases of the next mobile generation, for cloud services, and for virtualization and software-defined networking (SDN). This is the goal of the new OpenFastPath (OFP) Foundation, founded by Nokia Networks, ARM and industrial IT services player Enea. This aims to create an open source TCP/IP stack which can accelerate the move towards SDN in carrier and enterprise networks. Other sign-ups include AMD, Cavium, Freescale, Hewlett Packard Enterprise and the ARM-associated open source initiative, Linaro.
Nokia and its allies hope to accelerate the platforms to create open but secure network applications, which harness IP packet processing and support a whole range of use cases, some requiring very high throughput, others ultra-low latency (or both). These are, of course, also central requirements for 5G. The new Foundation is a non-profit, open source software initiative which promises to build an innovative developer community around a standardized and accelerated new-generation TCP/IP stack, the FastPath Stack, which would be optimized for SDN-ready network functions in carrier or enterprise networks. It would also, as its name suggests, support faster packet forwarding, via low IP latency combined with high capacity, and so reduce deployment and management costs by making networks more efficient.
It would consist of a high performance user space TCP/IP stack. User space networking aims to get TCP/IP out of the kernel – kernels are code-intensive, and using them for packet processing involves a number of operations (moving packets into memory, then to the kernel, then back out to the interface) which could be streamlined to reduce latency. The BBC has built a user space stack to allow its video servers to push put HD video, for instance, and user space networking is also seen as an important way to accelerate mobile network processing.
In the initial release of OFP, some functions will still have to be handed off to the kernel or another software packet processor, but that will change as the features are completed. An “optimized callback-based zero-copy socket API” aims to keep packet processing in user space as far as possible.
An open, cloud-focused architecture
The new fast-path TCP/IP stack will be based on the open source FreeBSD operating system and will initially comprise the User Datagram Protocol (UDP), Transmission Control Protocol (TCP), and Internet Control Message Protocol (ICMP). Other functions will be added in the future. The Foundation will also build an open source community to develop a best-in-class IP stack, and draw on best practice and innovation from existing proprietary optimized IP implementations.
Nokia says it will meet the “stringent requirements for core and radio applications in the cloud”, unlike older implementations, and this may be its key differentiator. The stack is will be optimized for ODP (Open Data Plane) programming interfaces, which enables hardware acceleration of the fast-paths, and allows the protocol to be programmable via the ODP environment. This will be important to cloud applications and to communications processing on bare metal switches in an SDN world.
As SDN moves into the mainstream of operator and enterprise plans, IT solutions will increasingly rely on transport layer capabilities, and so both layers must evolve in tandem. That IP transport layer, whether supported on wired or wireless networks, will itself have to adapt to the new requirements of the IoT and beyond, a process which will need to draw on technology and skills from the telco and the open source IT domains.
This will require significant levels of cooperation between different ecosystems, though these links have already been building over the years as telecoms and IT have converged and the wireless network has been increasingly driven by data. So the OFP Foundation has a tougher challenge even than creating a replacement for the creaking TCP/IP – it also needs to take a significant role in bridging the IT and telecoms worlds.