Does it look impudent marketing to designate the 5500 kms Mumbai-Chennai-Singapore subsea route as superhot submarine superhighway! Its relevance as standalone subsea route from the international gateways of global connectivity in north-west and south-east coast of India to APAC business hub of Singapore is as important as segment of the larger SEA-ME-WE Eurasian subsea corridor.
Despite the requirement being felt for new submarine cables on this route for quite some time, not much seems to be visibly happening, or so is it felt. Metaphorically, it's so close yet so far. What are OTT, IIP and Telco Carrier community apprehensive about! What challenges they are likely to face! How could they collaborate to address them! This article aims to take a closer look at cable architecture, market dynamics and address the palpable indecisiveness to sign on the dotted line.
Let's characterize the subsea route and understand its anatomy. It is uniquely interesting to note that the otherwise linear route represents a triangular topology with Mumbai-Chennai, Chennai-Singapore and Singapore-Mumbai segments. Each segment has its own significance, that is different from the other two segments.
Mumbai-Chennai subsea route could be an alternate national-long-distance or NLD route, even though most of the route lies in international waters. New cable developers are expected to weigh the merits of subsea alternative to terrestrial long haul. They are likely in dialog with the OTTs and Telco Carriers for this never before domestic subsea connectivity proposition and deliberating the regulatory perspective. Considering $40,000/km for 16FP cable system spanning 2000 kms, it leads to fiber pair ownership economics of $5M/FP. How does this compare to SLA-equivalent terrestrial long haul fiber pair ownership that will require three diverse routes! Considering the three diverse routes add up to 5000 kms and $5000/FP/km for dark fiber pair IRU/ownership, lighting up and O&M, it’s awfully more expensive, nearly 5X. The $ numbers can be debated, 5X could be 4X, but it won’t change the conclusion, and therein lies the attractiveness of the domestic long haul subsea route.
Chennai-Singapore is the shortest route to Singapore. Unless, you consider the Chennai-Andaman & Nicobar Islands (CANI) cable system under deployment and plan for extension from Port Blair to Singapore. That’s unjustifiably wishful proposition currently given the challenges to overcome. The Chennai-Singapore span has some interesting possibilities. First is forked cable landing in India at a location other than Chennai. The two likely locations are Vishakhapatnam (Vizag)and Digha (180 kms from Kolkata). There have been lukewarm conversations on developing CLS at these locations, however nothing tangible has surfaced till date. Newfound interest seems to be sprouting again for Digha that gives access to Kolkata, and with possibility of terrestrial connectivity to Nepal, Bhutan and even Bangladesh. Unfortunately, Kolkata and Eastern India does not present optimism for growth of size and scale that will encourage investment in submarine cable branch. Having said that its possible that if the submarine cable branch is built, it will trigger the much-awaited scale of growth. Hence someone needs to take the risk and bite the bullet. Digha being 1400 kms north of Chennai will entail investment of the order of $60M. It’s anybody’s guess that OTTs need to take the lead to make it happen.
Chennai-Singapore segment is well placed for subsea to connect four (4) countries enroute – Bangladesh, Myanmar, Thailand and Malaysia. Bangladesh is tightly controlled by incumbent monopoly and only BSCCL is licensed to land submarine cable and operate CLS. The NTTN & ITC providers have campaigned to get license to land submarine cables but have not succeed so far, and presumably given up. A possibility that’s been talked about is extending 260 kms terrestrial connectivity from Digha to Benapole, the Indo-Bangla border crossing point for terrestrial fiber interconnect. This PowerPoint proposal is easier said than done when it comes to business case justification. Meanwhile, Myanmar has gained considerable attention recently and set to land SIGMAR (2020), MIST (2022) and SMW6 (2023), other than SMW3, SMW5 and AAE-1, a stark contrast to Bangladesh, with 3X the population and GDP of Myanmar. With respect to Singapore, Malaysia (Penang) and Thailand (Satun) are low hanging fruits as the subsea route grazes past these countries. Hence it is expected that new submarine cable design will include provisions for BU to land submarine branch to these countries. Looking back TIC and i2i that has ruled the Chennai-Singapore corridor for 15 years also had similar plans with omnibus fiber pairs set aside and BU positioned, but it did not materialize eventually.
Singapore-Mumbai is the express route to extend new and existing Marseilles-Mumbai cable systems to Singapore. It could work the other way around as well for Singapore, though not as prominently, to extend new transpacific and intra-Asia cables to Mumbai. Going forward cable system interconnects in Mumbai will gain traction and that highlights the need for open cable landing. It remains to be seen if one landing party can land Marseilles-Mumbai and Mumbai-Chennai-Singapore cable system that will significantly facilitate and optimize interconnect, or will it require painstaking CLS-to-CLS dark fiber interconnect.
Singapore-Mumbai express route merits some more clarity. For some it means direct route from Mumbai to Singapore without touching Chennai. That’s possible if the cable system has Branch Landing in Chennai. However, in some way it's perceived to demote Chennai and its potential compared to Mumbai and, would not be acceptable to other cable systems. Thus, Chennai is expected to have mix of Branch Landing and Express Route Landing, and both will offer express route connectivity in its own right. Its been a mixed bag for consortium cables as well with SMW4 having Express Route Landing while BBG has Branch Landing.
Cable System Design Options
Triangular topology leads to interesting variants in cable architectures and fiber pair allocations. New cable systems that are more focused on Singapore-Mumbai as low-latency high-reliability express route are expected to go for Branch Landing in Chennai. On the other hand, cable systems that are equally emphatic about the three segments of triangular topology with greater flexibility in subsea connectivity architectures are likely to go for Express Route Landing.
Express Route Landing would mean two cables landing – one Mumbai-Chennai and the other Chennai-Singapore and hence require 2X cable landing facility compared to Branch Landing (and Termination Landing in Mumbai) terms of BMH and space/power/cooling for PFE. In fact some cable system design could ask for two separate BMH for the two cables to land for the purpose of diversity, but the two cables need a common meeting point where Mumbai-Chennai and Chennai-Singapore fiber pairs interconnect for Mumbai-Singapore express route. This is considerably simplified if the two cables land in the same BMH with co-located facility to host PFE and optical switching equipment. This advantage is largely nullified if the cable system design entails extension of the cables to a carrier neutral DC where the optical switching equipment along with SLTE would be hosted.
A likely question while comparing Branch Landing and Express Route Landing in Chennai would be – what is the Mumbai-Singapore latency differential that the two cable landing alternatives lead to! Singapore-Mumbai route length will be less with Branch Landing from the trunk cable as Mumbai-Singapore fiber pairs will not land in Chennai. With Express Route Landing the Mumbai-Singapore fiber pairs will traverse nearly twice the length of the branch, land in Chennai and interconnect somewhere inland. Length of the branch from the trunk cable is expected to be 500 kms and traversing it twice translates to nearly 10 ms of added latency. It would not materially impact subsea connectivity over Marseilles-Singapore span. Will it really matter for Mumbai-Singapore!
With respect to fiber pair allocation, an equilateral architecture with 16FP system would lead to 8FP for each arm of triangular topology. It gets interesting if Mumbai-Chennai domestic subsea connectivity is not considered as high-demand segment as the other two and hence assigned lower count of fiber pairs. The fiber pair count doesn’t add up to 16 equitably on the three arms and that creates relevance for fiber pair switching technology. Fixed vs flexible architecture for route allocation of fiber pairs would be a topic of insightful discussion going forward.
Who are the contenders to build new submarine cables!
It’s a no-brainer - OTT, IIP and Telco Carrier. It leads to seven (7) possible combinations. It’s highly unlikely that any one of them would invest and develop a submarine cable system independently without involving others. This brings down the possible combinations to four (4). IIP+Telco Carrier is ruled out at it does not have OTT in the mix. OTT+IIP could work provided IIP takes NLD/ILD license. The best combination however seems to be consortia of OTT, IIP and Telco Carrier with each having a significant role to play. Interplay of cable system developer, anchor-tenant and landing-party is expected to drive OTT, IIP and Telco Carriers to join hands and collaborate. The framework of collaboration may differ, but it revolves around three aspects - Investment, Cable System Design & Deployment and License & CLS.
Instinctively we associate OTT with Investment and Telco Carrier with License and CLS. That leaves the IIP for Cable System Design & Deployment. They are certainly expected to play lead role with that responsibility and challenge telco dominance of submarine cable landscape. Some of them are expected to set up subsidiary in India, get permission for 100% FDI that the telecom sector allows, and pursuant to it, acquire NLD/ILD license to be the licensed landing party. While acquiring NLD/ILD license is not majorly challenging, none has been awarded license yet. Optimism prevails that licensed entities would unveil their identity in the next 6-8 months. That could determine when CIF is announced. Further, having acquired license, aspirations could soar, and IIPs decide to build CLS in order to have ownership and control of cable landing. Should we extrapolate any further, consider the IIP’s could also build metro fiber network and risk overstepping the realm of practicality!
We went little too far without pausing to ask, is it mandatory for an IIP to acquire NLD/ILD license and build CLS ground up in order to land a submarine cable! Certainly not. For regulatory and licensing compliance IIPs can forge alliance with Telco Carriers, such that the Telco Carrier with CLS is the licensed landing party.
OTTs are in the forefront of new submarine cables globally and India is no exception. They reportedly have FP ownership in i2i, TIC and FALCON. The FP capacity or T/FP remains an interesting question with closely guarded answer. Given these are 15-year old cables, SLTE upgrades could possibly overstretch the system to deliver up to 8T/FP. This over a period of time could turn out to be flogging the dead horse and unlikely to be preferred. Hence OTTs would prefer to go for new cable systems for attractive economics of fiber pair ownership.
An OTT or OTT+OTT alone cannot build a Mumbai-Chennai-Singapore cable system, for they are not licensed operator in India. What is the possibility that OTTs will take NLD/ILD license for greater ownership and control on submarine cable build, cable landing and backhaul connectivity! The answer in unexpectedly simple. Consider an OTT with NLD/ILD being served a notice to pay 8% of their Adjusted Gross Revenue (AGR) pursuant to the recent judgement of the honorable Supreme Court. Hence OTTs are expected to stay away from taking license. This leads to MAREA or DUNANT like build with involvement of a Telco Carrier as the Landing Party in India, and OTT's Singapore subsidiary as the Landing Party in Singapore.
The above arrangement with collaboration of OTT and Telco Carrier would mean, the new submarine cable will land in Telco Carrier CLS. However, the Telco Carriers are apprehensive in general to offer open cable landing with carrier neutral fiber pair breakout, for it could jeopardize the pricing dynamics for capacity inventory that they have in consortium cables and enjoy grant of capacity upgrades every year, with shared investment. However, financial stress and focus on asset monetization could drive them to think differently and offer open cable landing. At the same time there are Telco Carriers that does not carry burden of telco legacy and poised to offer open cable landing with flexible landing architecture.
Can the three Telco Carriers, Airtel, Tata Communications and Vodafone Idea join hands to build a 24FP cable system where each party has ownership of 8FP with 20T/FP. By doing so they could get away from exasperating and even extortionist dependency on anchor customer signoff to develop the cable system. It would suffice to say, it's unreasonably wishful under current circumstances and looks unlikely.
How many submarine cables connect Mumbai-Chennai-Singapore at present!
There’s only one – SMW4. However, it’s not meaningful practically, as it lands with Tata Communications in Mumbai and with Airtel in Chennai. Reportedly Airtel has learnt its lessons and will provide Express Route Landing in Chennai and Mumbai for SMW6.
BBG comes close with Mumbai, Chennai and Malaysia, and with 750 kms terrestrial extension from Penang to Singapore. However, BBG has the same limitation as SMW4. In Mumbai it lands with Vodafone and in Chennai it lands with Reliance Jio.
Competing cousins i2i and TIC from Airtel and Tata Communications have had great success with Chennai-Singapore connectivity. Airtel has had added advantage of protected connectivity with i2i+SMW4. Both the cable systems interconnect with respective Chennai-Mumbai terrestrial long-haul. However, the resulting Mumbai-Singapore connectivity has not proved to be reliable enough for hyperscale connectivity. It remains to be seen if Airtel and Tata Communications plan for successor to i2i and TIC. It will be necessitated by devaluation expected to be triggered by the new cable systems.
Where are the new cable systems!
Need for three (3) new submarine cables has been palpable. It's but anybody's guess that active discussions are underway between OTTs, IIPs and Telco Carriers on new cable system build. Everyone is vying for the best option, such that they contract for truly open cable landing are not outrun in terms of fiber pair ownership economics with competing cable systems. However, till date none of the new cable systems has announced CIF. A month back, in December 2019, MIST from Orient Link revealed its identity.
Reportedly four (4) new submarine cables are in various stages of planning and development. Some have reportedly made significance progress and expected to announce CIF in the next 3-8 months, though this expectation has persisted for last one-year. These are would be high fiber count submarine cables with SDM technology with 12FP or more likely 16FP configuration with fiber pair capacity ranging from 12T/FP to 22T/FP. Understandably Chennai-Singapore will have higher T/FP but lower $/FP compared to Singapore-Mumbai.
A lot is expected to happen this year with new submarine cables on Mumbai-Chennai-Singapore route. Some will be never before development. Fiber pair ownership economics will be intensely debated with it's acquisition, ownership and operationalization modeled for regulatory compliance. Rational to charge RIO/AFC will be challenged. IIPs will acquire license. OTTs will make up their mind and sign on the dotted line. Several Landing Party Agreements (LPA) will be signed with Open Cable Landing offered as managed infrastructure service. Series of public announcements before or on CIF for new cable systems will put an end to speculations. Let’s connect to India - will reverberate far and wide and usher a new era of carrier neutral network infrastructure cleansed of telco stranglehold.