Space; the final telco frontier

Reach out, reach out and touch someone

Everyday there is a news story about SpaceX, Blue Origin or Virgin Galactic. We are fascinated by the hubris of billionaires Bezos, Musk and Branson to expand humanity off of Earth and out into space. The more I delve into the NewSpace industry , the more I am astounded by how much it reminds me of my industry, telecom. Let me explain.

The late 1980’s and the 1990’s were a time of disruption and change in telecommunications around the world . Deregulation and liberalization were in vogue as competition was introduced into previous monopoly markets. It started with upstart competitors like MCI and Sprint trying to break into AT&T’s long distance monopoly in the USA. Then the Telecommunications Act of 1996 in the US creating a whole new batch of mini-Bells.

The UK similarly saw the privatization of government owned British Telecom in 1984 and its monopoly was challenged by Mercury, a subsidiary of Cable and Wireless communications . Other countries , Canada, Australia, New Zealand, the Netherlands followed suit. It spread globally.

Governments worldwide embraced privatization of their former telco monopolies (PTT’s or Postal Telegraph & Telephone) to become publicly listed commercial entities and introduced competition into these markets.

Simultaneously we saw the rise of the worldwide Web and consumers were introduced to the Internet. Dial-up modems provided access at first, but were quickly replaced by broadband connections. To feed the growing demand for bandwidth for competitive telecom and the Internet there was a flood of new fiber optic cables laid. According to KMR Research about 80 million miles of optical fiber was installed between 1996 and 2001 in the USA alone. New submarine fiber cable routes between continents were planned and laid. The forecasts were that Internet traffic would double, every year, so we needed MORE.

Telecom was the new Wild West, the new Klondike gold rush. Capital markets paid attention and newly minted companies that had telecom, broadband or Internet in the name got funded to operate in this new environment of explosive growth. The stock markets in Canada, the USA and Europe were filled with telecom and broadband backbone providers, submarine cable companies , hardware vendors. It was as big if not bigger than the dot com boom.

My own career has been imbued by this zeitgeist . After engineering school at Waterloo I started at CNCP, which was owned by CP and CN the two giant Canadian railways. It operated the Telex network across Canada and provided the main competition to the incumbent provincial telephone monopolies. When long distance was finally deregulated in Canada I was part of the management team of ACC TeleEnterprises which we grew and took public on the Toronto and Montreal exchanges. At ACC we also had the worlds first trans-Atlantic private line that hooked up to the public phone network connecting London, UK with switches in Toronto, Montreal and Vancouver. After ACC, I started AurorA, which provided international telecom services, both voice termination and bandwidth to the competitive industry in Canada, the USA and telcos around the world.

So why do I write today about this ancient history ? Because the nascent NewSpace industry reminds me very much of those days in telecom. The parallels are tremendous.

Access to space used to be limited to governments, specifically those of the United States, Russia and China. The “Space Race” in the 1960’s between the Soviet Union and the Americans during the Mercury, Gemini and Apollo programs could only be funded by national governments . Government contractors built the rockets and other equipment and they were typically subsidiaries of the defence industry. Those contracts were for cost-plus and programs were large, expensive, bureaucratic and slow.

About the time of the go-go telecom years we saw the beginnings of a commercial space sector.. The race to the moon had given way to government cutbacks to NASA; the focus in the ’90s became the Space Shuttle program and the International Space Station (ISS). Telecom dreamers wanted to put up broadband in space, a way to “connect the unconnected”. This led to ideas like Iridium, Teledesic and GlobalStar that were able to access the public markets during the telecom boom to fund their dreams. There were also startups backed by legacy aerospace companies like Boeing, Lockheed and Alcatel. They all sadly failed and went bankrupt. (I wrote a review on a great book on the history if Iridium here).

Then came SpaceX. Through the vision of Elon Musk, his newly minted fortune from PayPal and the incredible fortitude of a young team of talented engineers, SpaceX was able to challenge and disrupt the Space industry in the early 2000’s. They were able to provide a new way to build rockets, using the agile, iterative approach of Silicon Valley. Their unlikely success revitalized a commercial space sector by dramatically lowering the cost of launch.

Thus was born the NewSpace industry. There have probably been about 1000 start-up space companies in the last 6 to 8 years. It was hard for them to get funding initially as many are very capital intensive and there really were no space focussed Venture capital firms until the last few years. Running a space company meant always looking for funding to keep the company alive.

Then came SPACs. Richard Branson’s Virgin Galactic needed more funding and he was able to get it in 2019 by going public via a SPAC deal with Social Capital Hedosophia run by Chamath Palihapitiya (who among other things is a UWaterloo engineer also !) to become the world’s first publicly traded commercial human spaceflight company.

Other space companies have since taken the opportunity to go public also via the SPAC route such as AST Science, Momentus, RocketLab, Astra, Spire Global, BlackSky, Redwire and Arqit Ltd. SPAC funding typically provides a space company with enough capital to fully fund their business plan so they can focus on execution. Today the space sector is worth about $400B but Morgan Stanley predicts it will be trillion dollar industry by 2030. That is probably conservative given the pace of change.

This is why I get excited about the NewSpace industry. What started off as a government monopoly became a competitive commercial industry fueled by the public stock market. Disrupted by new ideas, new processes and new technology. The new mantra is “You can’t spell space without SPAC” !

My final observation, is that even though launching satellites, hardware and humans into space does take a rocket scientist, in the end much of the applications really boil down to this being a telco in space. Whether it is providing broadband internet like Starlink or OneWeb (there is that same “connect the unconnected” dream), or Internet of Things (IoT) capabilities like Swarm or Kepler, Cellular or 5G from space like AST, Lynk or OmniSpace or any of the Earth Observation platforms (visual, SAR, RF or thermal) that connect their data stream to hyper provider clouds like AWS, Azure or GCP it all comes down to moving bits around, just like a telco. We’re just extending the tower up a little higher.

Ad Astra (Latin for “To the Stars”)

RAG Americas Online – Day 2

Art by Vickie Vainionpää, “The Third Insight” – https://vickievainionpaa.com/

This spring I was invited to be a speaker at the next Risk and Assurance Group (RAG) Conference. It was to be held in Denver, CO at Century Link (now Lumen). Alas , the COVID-19 pandemic scuttled those plans so until we can meet-up again live, it was decided to move the conference online and RAG Americas Online virtual conference was launched. September 16, 2020 was Day 2.

The highlight of Day 2 was the panel discussion on “Moving Telcos to the Public Cloud”. The panelists were Clodagh Durkan an expert in the fields of security governance, data protection, cryptography and perimeter threat detection ; Patrick Donegan the founder and principal analyst of Hardenstance, a research business focused on telecom and IT security ; and Danielle Royston former CEO of Optiva and Evangelist for the Cloud.

Most industries have already made the move to the cloud, after dipping their toes in it they have gone all in with their whole businesses. Even former skeptics and holdouts such as JP Morgan Chase have become cloud advocates .

The premise was that telcos need to move their software and datasets from their own data centres to the public cloud; specifically the hyperscalers ; Amazon’s AWS, Microsoft’s Azure and Google’s Google Cloud. The immediate benefits are dramatically reduced costs, increased flexibility and security and then abundant opportunities for growth and improvement by using the tools and software that the clouds provide. The clouds are not commodities nor necessarily competitors , they offer the opportunity for telcos to evolve and become better.

There were many questions from the audience and objections were raised on multiple fronts. The hyperscalers are all American companies (and we didnt include Alibaba from China and others on the list). It was noted that the public cloud companies have locations in almost every country and can contractually tailor the solutions to satisfy stringent government data localization requirements.

What about security ? The point was made that the hyperscalers are orders of magnitude more vigilant about security than even the largest telco would be. They have the staff, budget and experience to do it better. The telco can hand off that headache and hold the cloud companies feet to the fire via the contractual obligations.

Some large telcos have started to move towards the cloud. Examples such as Vodafone , Deutsche Telekom and Verizon were touted with specific projects. It may take time for telcos to fully relinquish their perceived investments in control, the desire to “hug their own servers”.

This is an area I will be watching with eager anticipation. I know if I was younger and just starting out again I would defiantly build AurorA 2.0 as a native cloud company and take advantage of the software tools such as Artificial Intelligence and Machine Learning that a company AurorA’s size wouldn’t be able to afford otherwise. The cloud really tips the scales to level the playing field for smaller entrepreneurial companies to be able to compete with even global giants.

If you didn’t get a chance to watch the live stream, the videos of the presentations will be available to view at the RAG website;
https://riskandassurancegroup.org/

I highly recommend searching out this panel discussion to view and digest for anyone in telecom, anywhere in the world.

RAG Americas Online – Day 1

Matt hard at work getting RAG Americas Online ready to rock !

This spring I was invited to be a speaker at the next Risk and Assurance Group (RAG) Conference. It was to be held in Denver, CO at Century Link. Alas , the COVID-19 pandemic scuttled those plans so until we can meet-up again live, it was decided to move the conference online and RAG Americas Online virtual conference was launched. Today, September 15, 2020 was Day 1.

There was a mix of about a dozen presentations and panel discussions interspersed with commentary from the Wise Heads comperes; Eric Priezkalns, Rachel Goodin, Nixon Wampamba and Tony Sani . Canadians were well represented with 3 speakers today and 3 tomorrow (including me) from companies like Telus, Rogers and Xplorenet.

From my vantage the most relevant panel was the discussion on Trends in Fraud Management, especially during this pandemic. Which frauds are up, which are down and what is new ? The consensus was a large increase in social engineering and identity fraud (all manner of phishing schemes), as well as increases in CLID spoofing. Reinforcing that was an expert panel just on stopping the spoofing of calls !

A recurring theme through many of todays presentations was Artificial Intelligence (AI) and Machine Learning (ML) and their application in Revenue Assurance and Business Assurance. It appeared as more than just a way to automate fraud management, more as a set of tools to be able to transform the entire discipline and provide better business results.

If the telcos manage to move their massive data sets (the industry with the biggest data sets apparently) into the public cloud and could use AI/ML tools along with the other benefits that cloud brings, we could really see a profound transformation of the telecom industry ! That will be the panel discussion on Day 2 that I am most looking forward to; Moving Telcos to the Public Cloud.

If you didn’t get a chance to watch the live stream today, the videos of the presentations will be available to view at the RAG website;
https://riskandassurancegroup.org/

I will post a summary of Day 2 later followed by a separate post of my own presentation on Least Corrupt Routing.

Elon and Jeff are brilliant ! Surely THEY can solve our broadband issues.

Two dishes got married; the ceremony was meh but the reception was incredible.

Much has happened since we last visited the wacky world of low earth orbit (LEO) satellite constellations and their use in providing improved broadband service to Canada’s rural and remote users. This past Tuesday, July 21, all of Iqaluit, the capital of the Territory of Nunavut was without communication services ; no Internet, no landline, no cell service, no cable TV – simply because it was raining ! In a first world country like Canada this is unacceptable. We need better broadband service in Canada’s North NOW.

There is a rash of breathless newspaper stories in the mainstream media touting LEO service as arriving soon to resolve our remote and rural broadband issues. I wrote about it before here, that Elon Musk is not coming to save us any time soon. I also wrote about the Chapter 11 bankruptcy of the early leader to provide LEO service to the Arctic, OneWeb, here. So where do we stand now on July 27, 2020 ?

Well on July 10, the US bankruptcy court of the Southern District of New York (SDNY) approved a joint $1 billion bid for OneWeb by Britain and Bharti Airtel. The UK government and Bharti Global, an arm of Bharti Enterprises, which part owns India’s Airtel, will each have roughly 45 per cent of OneWeb. The existing secured creditors, including SoftBank of Japan, OneWeb’s former biggest shareholder, will own the balance.

But the landscape has changed from before OneWeb’s descent into Chapter 11 in the spring. OneWeb’s original mission was to “connect the unconnected “ ; ie it wanted to provide broadband service to the millions of people around the world that do not have access to the Internet. The UK has invested $500M into OneWeb for other strategic reasons, mainly to mitigate the effects of Brexit on British industry. I sure hope they realize that it is going to cost them more, much more and that $500M was just the table stakes to play in the LEO game.

After Brexit, the UK would have been locked out of the EU’s Galileo mission. Its aerospace industry would have lost work and technology to the EU and the USA and it would have fallen behind. Investing in OneWeb, and taking the “golden share” that lets it decide who gets access to the network will let it be a cornerstone of new industrial policy for Britain. Jolly good show.

But a $1 billion investment is just the starting point, OneWeb will need to raise at least $1B if not more to fully fund the company’s plans. So to mitigate, the mission will be changed from targeting unserved consumers to one targeting governments and commercial sectors to generate revenue quickly. These use cases are more in sync with what Telesat is proposing to do in selling wholesale connectivity to telcos and ISPs, starting with the Arctic. The UK also wants to develop other use cases such as a new navigation offering different from GPS or Galileo that it could then sell to its Five Eyes partners as an adjunct to GPS. That requires adapting and redesigning the satellites to the new missions.

The UK also wants to use this transaction bring more aerospace manufacturing back to Britain. The current OneWeb satellites were made in Florida in a joint venture with Airbus. Some or all of this manufacturing is expected to move to the UK.

The transaction to close the deal to buy OneWeb from Chapter 11 is expected to close in Q4 2020. Thus the former timeline of providing service in the Arctic by late 2020 is out of reach, 2021 will be too aggressive and 2022 is more realistic. All of the changes wrought by the new ownership and new mandates have created issues and decisions that need to be navigated; technical, financial as well as political which all bring delays to actually launching services.

Meanwhile, the other proposed LEO constellations have not been idle.

Amazon’s Project Kuiper has won the backing of Ajit Pai, the Chairman of the FCC who is backing approval of the venture.

SpaceX’s StarLink venture has just begun to raise more funding to continue launching service. CNBC has reported that it is in talks to raise $500M to $1B funding at a valuation of $44B (July 23) . SpaceX has also been not so quietly signing up beta customers in Canada and the USA to test trial service of its direct-to-consumer internet service late this year.

That leaves Canada’s Telesat. Telesat was also a bidder for OneWeb during its Chapter 11 bankruptcy. That bid came out of the blue, and we still don’t know what motivated Telesat to launch the bid, nor who financed the deposit. Telesat has yet to choose a manufacturer for its satellites with apparently an announcement due any moment, perhaps at their Q2 results conference call on July 30. The biggest question mark remains around the financing of the project.

In July 2019, Telesat received $85M CDN from the Federal Government’s Strategic Innovation Fund. Once the constellation is in service, they have a commitment from the Feds to purchase $600M CDN in services over 10 years. That still leaves a huge question of how Telesat plans to finance the enormous capex to launch the LEO constellation before it sees any revenue. As we saw with OneWeb, there are not many investors around that are willing to sign the large cheques needed to finance such risky ventures. Even Softbank (funders of Uber and WeWork) balked at additional funding for OneWeb while they were the largest shareholder and creditor. Such investors need to be able to keep signing large cheques to keep funding capex. (are you paying attention Boris?)

Telesat’s current shareholders are Loral Space & Communications, a NY based holding company and Canada’s Public Sector Pension investment Boards (PSP). I am not sure either are up to the task of financing such a risky venture. Loral recently paid a special dividend of $5.50/share on May 28, 2020 distributing to shareholders funds it had received from Telesat. Not the kind of action you would expect to see from a party planning to underwrite billions of capex expenditures. The PSP is a pension fund that historically prefers ventures that provide stable cash flows at predictable rates, which is what Telesat’s legacy business provided. LEO constellations are highly speculative and very risky, the exact opposite of what the pension plan for Canada’s public sector workers want to see done with their retirement funds.

From my personal perspective (ie I am guessing) I can see a strategic buyer taking control of Telesat from its current owners and funding the LEO constellation. That buyer would be Amazon and its Project Kuiper.

Amazon’s strengths are its financial muscle and long term vision. Amazon can raise funds cheaper than most nation states. Amazon may covet some of Telesat’s strategic assets such as its protected spectrum rights from the ITU and its substantial base of satellite technology and experience. Both of those would give them a significant advantage over Starlink. Telesat and Amazon have a relationship already as Telesat signed a launch contract with Blue Origin, another of Jeff Bezos’s companies for multiple launch missions in January, 2019.

Amazon could also introduce another use case for the LEO sats that no other strategic buyer could. They could use the service internally to connect their global network of AWS data centres. In essence, it could be a part of the connectivity portion of their AWS cloud. Potentially at a lower latency than some sub-sea fiber cables. That use case may be appealing on its own.

It would give Loral and PSP gracious exits and Jeff Bezos a competitive advantage against OneWeb who will be be busy sorting out their new ownership and new mandates over the next two quarters and let Project Kuiper keep pace with Elon Musk’s Starlink as it beta trials consumer service this winter.

Never underestimate the power of bragging rights between billionaires.

We shall continue to watch this sector closely. My core message remains, do not expect LEO satellite constellations to be the saviour of Canada’s remote and rural broadband issues any time soon. We must continue to expand our network of fiber optic cables, including Arctic sub-sea cables.

Networking Pendulum

What was once old, is new again


One of the benefits of age is perspective. You notice that many of the “new” ideas are simply old ones that have come back into vogue, like the swing of the pendulum. The aptness of the metaphor is clear, as history demonstrates a tendency for human events to swing back and forth from one extreme to another.

We see this in politics (conservative vs liberal), we see this in fashion and in telecom networking. Early in my career, while still a systems engineer, I remember one of my first published articles being about the swing from companies using public networks based on X.25 packet protocol (like Datapac) to private networks using their own multiplexors and leased lines. (Note I had just moved from Bell Canada where I supported Datapac to General DataComm where we were selling muxes)

We’ve seen this pendulum swing between using public and private networks for a company WAN many times over the years. Leased private lines gave way to X.25 packet networks, which ceded ground to T1 or T3 networks (or fractional T1 like Megastream). Frame Relay, ATM and then the rise of MPLS, “Multiprotocol Label Switching”. Now we are seeing a challenger in the corporate networking world, SD-WAN, which uses the public Internet and extensive software to try to mimic and replace MPLS.

MPLS can be slow to implement, especially internationally, as it takes time to order and connect all of the connections, especially the final local Ethernet connections at each country. MPLS can also be expensive compared to DIA, Direct Internet Access. In MPLS’s favour, as with most private networks, is it’s inherent security, consistent latency, and guaranteed service levels and Quality of Service (QoS). For overseas voice circuits and critical enterprise data that is essential.

SD-WAN, based on the now ubiquitous public Internet, is now widely available, quickly deployable and seemingly less expensive. By using multiple business grade Internet connections (DIA which should be contention free) and some fancy software , it can approach the level of consistency of MPLS. Is it less expensive ? Well, vendors will make that case based on pure network costs, but soft costs of running and maintaining the equipment and connections have to be factored in.

The tension between secure and reliable private networks and less expensive shared public networks (like X.25, Internet and cloud) is one that has been going on for years, and watching this pendulum swing back and forth is something I find fascinating.

2019 Canadian ISP Summit – Day 1

I was looking forward to this event for months and Day 1 did not disappoint. Met up with many friends and customers within the first few minutes of arriving on Monday. And made new friends and contacts throughout the day . Thanks to TekSavvy for sponsoring my “office” at the show.

The keynotes on innovation and disruption in our Internet industry were very interesting. The talk about women in tech by Maryna Ivus of ICTC was very eye opening. She outlined StatCan numbers that show the number of women working in telecom fell from 35% in 1999 to 25% in 2019. Those numbers surprised and disappointed me as someone who has been in telecom all my adult life.

There was ample networking at this sold-out event; a cocktail reception sponsored by TorIX and QIX, a superb dinner at the CRAFT Beer Market by MBSI WAV and Cambium Networks and an after party by ADTRAN that I was smart enough to not attend . There are times I have to acknowledge my own limits ! Thank you to all the sponsors for making this event so special.

Day 2 has a full agenda with keynotes and panels on both the business of ISP’s and technical discussions. I will blog the highlights tomorrow after the Gala Dinner.