counter for iweb
Silicon Photonics

Published book, click here

« Edgecore exploits telecom’s open-networking opportunity | Main | Using an open-source model to spur AI adoption »

TIP launches a disaggregated cell-site gateway design 

Part 1: TIP white-box designs

Four leading telecom operators, members of the Telecom Infra Project (TIP), have developed a disaggregated white-box design for cell sites. The four operators are Orange, Telefonica, TIM Brazil and Vodafone. BT is also believed to be backing the open-design cell-site venture.

 Source: ADVA

The first TIP cell-site gateway product, known as Odyssey-DCSG, is being brought to market by ADVA and Edgecore Networks.

TIP isn’t the only open design framework that is developing cell-site gateways. Edgecore Networks contributed in October a design to the Open Compute Project (OCP) that is based on an AT&T cell-site gateway specification. There are thus two overlapping open networking initiatives developing disaggregated cell-site gateways. 

ADVA and Edgecore will provide the standardised cell-site gateways as operators deploy 5G. The platforms will support either commercial cell-site gateway software or open-source code. 

“We are providing a white box at cell sites to interconnect them back into the network,” says Bill Burger, vice president, business development and marketing, North America at Edgecore Networks. 

“The cell site is a really nice space for a white-box because volumes are high,” says Niall Robinson, vice president, global business development at ADVA. Vodafone alone has stated that it has 300,000 cell-site gateways that will need to be updated for 5G.



A mobile cell site comprises remote radio units (RRUs) located on cell towers that interface to the mobile baseband unit (BBU). The baseband unit also connects to the disaggregated cell-site gateway with the two platforms communicating using IP-over-Ethernet. “The cell-site gateway is basically an IP box,” says Robinson. 

The Odyssey gateway design is based on a general-purpose Intel microprocessor and a 120-gigabit Broadcom Qumran-UX switch chip.

The white box’s link speeds to the baseband unit range from legacy 10 megabit-per-second (Mbps) to 1 gigabit-per-second (Gbps). The TIP gateway’s uplinks are two 25-gigabit SFP28 modules typically. In contrast, the OCP’s gateway design uses a higher capacity 300-gigabit Qumran-AX switch chip and has two 100-gigabit QSFP28 uplink interfaces. “There is a difference in capacity [for the two designs] and hence in their cost,” says Robinson.


The cell-site gateway is basically an IP box


The cell-site gateways can be connected in a ring with the traffic fed to an aggregation unit for transmission within the network.          

Robinson expects other players to join ADVA and Edgecore as project partners to bring the TIP gateway to market. To date, no software partners have been announced. First samples of the platform are expected in the first quarter of 2019 with general availability in the third quarter of 2019.

“Cell-site gateways is one of those markets that will benefit from driving a common design,” says Robinson. The goal is to get away from operators choosing proprietary platforms. “You have one design hitting the market and being chosen by the different end users,” he says. “Volumes go up and costs go down.”

ADVA is also acting as the systems integrator, offering installation, commissioning and monitoring services for the gateway. “People like disaggregation when costs are being added up but end users like things - especially in high volumes - to be reintegrated to make it easy for their operations folk,” says Robinson.

The disaggregated cell-site gateway project is part of TIP’s Open Optical and Packet Transport group, the same group that is developing the Voyager packet-optical white box.    


Source: Gazettabyte



ADVA announced recently that the Voyager platform is now available, two years after being unveiled. 

The 1-rack-unit Voyager platform uses up to 2 terabits of the 3.2-terabit Broadcom Tomahawk switch-chip: a dozen 100-gigabit client-side interfaces and 800 gigabits of coherent line-side capacity.

Robinson admits that the Voyager platform would have come to market earlier had SnapRoute - providing the platform’s operating system - not withdrawn from the project. Cumulus Networks then joined the project as SnapRoute’s replacement. 

“This shows both sides of the white-box model,” says Robinson. How a collective project design can have a key member drop out but also the strength of a design community when a replacement can step in.  

TIP has yet to announce Voyager customers although the expectation is that this will happen in the next six months.

Robinson identifies two use cases for the platform: regional metro networks of up to 600km and data centre interconnect.

“Voyager has four networking ports allowing an optical network to be built,” says Robinson. “Once you have that in place, it is very easy to set up Layer-2 and Layer-3 services on top.” 

The second use case is data centre interconnect, providing enterprises with Layer-2 trucking connectivity services between sites. “Voyager is not just about getting bits across but about Layer-2 structures,” says Robinson. 

The Voyager is not targeted at leading internet content providers that operate large-scale data centres. They will use specific, leading-edge platforms. “The hyperscalers have moved on,” says Robinson. “The Voyager will play in a different market, a smaller-sized data centre interconnect space.”   


We will be right at the front and I think we will reap the rewards for jumping in early


Early-mover advantage

Robinson contrasts how the Voyager and TIP’s cell-site gateway were developed. Facebook developed and contributed the Voyager design to TIP and only then did members become aware of the design. 

With the cell-site gateway, a preliminary specification was developed with one customer - Vodafone - before it was taken to other operators. These companies that make up a good portion of the cell site market worked on the specification before being offered to the TIP marketplace for development. 

“This is the right model for a next-generation Voyager design,” says Robinson. Moreover, rather than addressing the hyperscalers’ specialised requirements involving the latest coherent chips and optical pluggable modules, the next Voyager design should be more like the cell-site gateway, says Robinson: “A little bit more bread-and-butter: go after the 100-gigabit market and make that more of a commodity.”  

ADVA also believes in a first-mover advantage with open networking designs such as the TIP cell-site gateway. 

“We have been involved for quite some time, as has Edgecore with which we have teamed up,” says Robinson. “We will be right at the front and I think we will reap the rewards for jumping in early.”


Part 2: Open networking, click here

Reader Comments

There are no comments for this journal entry. To create a new comment, use the form below.

PostPost a New Comment

Enter your information below to add a new comment.
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>