IOWN the Photons
Several years ago, while surfing the web, the author of these lines discovered something unexpected… an organization with focus on optical networks he had not heard of before. Clear the stage for the IOWN (Innovative Optical and Wireless Networks) Global Forum, “a community of industry-leading organizations from across the globe working together to accelerate the innovation and adoption of photonics-based technologies.” Not bad, in this IT-driven age of xG, AI and Data Centers, to have someone publicly highlight the sometimes hidden value, and potential, of optical networking.
Years have passed, and the organization has grown beyond a more Japan-centric focus to become a global player, while expanding its expertise beyond networking. At this year’s 25thanniversary meeting in Stockholm, where yours truly had the honor of introducing Ribbon as a member, IOWN made an impressive demonstration of both achievements and vision.
At first glance, IOWN’s mission is somewhat difficult to grasp, as many topics sound familiar to long term members of the industry. What’s special about IOWN is their “bigger picture” concept. The value is not in the partially well-known ingredients; it’s in the final meal. IOWN is pushing for nothing less than a network infrastructure with the ideal balance of power consumption and bandwidth efficiency, the APN (All Photonics Network), seamlessly interworking with attached systems like distributed computers. These networks would be ready for the new AI-powered age, and built to ensure that the earth doesn’t catch on fire due to the energy required to keep the machines “thinking.”
IOWN’s recommended architecture is based on three building blocks plus the APN (All Photonics Network)-C, which hosts the controller function. Starting at the customer end, the APN-T (Transceiver), is the endpoint and wavelength generator. The next block in the chain is the APN-G, a type of gateway that communicates with the connected APN-Ts, handles wavelength path services and aggregates optical paths from the APN-T to the trunk network. And finally, the APN-I, an interchange for wavelength switching at the midpoint of an optical path, owns functions such as wavelength cross-connection, amplification, and adaptation between APN-G and APN-I as well as APN-I and APN-I… an Optical Cross Connect (OXC), for building meshed networks.
Does it mean the OXC is back, and that we can pull out our 25 year-old slide decks to finally start selling? Hold your horses, as is often in the case in telecommunications, reality is standing in the way. Existing networks can’t be upgraded by “just” replacing a DXC (Digital Cross Connect) with an OXC. Most amplifiers have to be replaced as well, technologies for keeping a network stable while switching dozens of lightpaths have to be in place, and processing data with granularity finer than a wavelength needs to be done differently. Looks like this road to power savings is quite a long one. Is building brand new all photonic networks the way?
Indeed, that would be a viable option, once the required amps and OXCs actually exist. The former is realistic and at least partially achieved, the latter strongly depends on vendor business cases based on real market demand. Besides hardware, the underlying idea of true end-end-lightpaths requires true end-end management, also covering customer owned routers and computers, or at least their transceivers. Doable? Absolutely. Acceptable? Time will tell… so far, customers don’t like the idea of anybody else managing even the smallest part of their devices. But as the endpoint for a wavelength path, the APN-T can easily be a transponder in front of the customer device, it’s possible to avoid this debate and still enjoy both savings and benefits within the rest of the network.
Optical Circuit Switching (OCS) was also mentioned.in multiple presentations at the recent OCP25 EMEA Summit, this time for applications within Data Centers. Although no groundbreaking developments were highlighted, a well-known chip manufacturer indicated that some of the developments surrounding SiPho (Silicon Photonics) based solutions were interesting, and that Co Packaged Optics (CPO) combined with OCS eliminating an electrical switching stage cuts total network power by 40% or even more But firstly, they want to see the technology mature, and the cost level is an issue as well.
IOWN’s vision appears to be in sync with market dynamics, a good start for real life adoption. The positive effects of photonic networks are widely understood and accepted, and the industry is moving in this direction. Despite the challenges, we are much closer to making all photonic networks come true than in the dot-com days. Optical modules have shrunk dramatically and gained capabilities no one dared to dream of in the hype phase. IPoDWDM applications are saving both space and power and are adaptable enough to cope with the effects of a dynamic photonic network. IOWN rightly acknowledges that just proving that it works is insufficient, Proof-of-Value (PoV) and Proof-of-Business (PoB) cases are crucial for successful rollouts. Eventually, and this should come as no surprise, the market is going to decide how fast we will get close to what some people might call the ideal network. Now it’s about turning concepts, use cases and PoCs into real life action.
