Background
The unprecedented demands for high-bandwidth and low-latency connectivity have spearheaded recent government investments in complex WISP (Wireless internet service provider) technologies—ultimately seeking a cost-effective infrastructure solution to the rural digital divide.
While the race to 5G accelerates evermore across the densely-populated areas of the world, costly deployments in more remote regions have largely been ignored. With many regions in the US still disconnected or at the mercy of the few satellite WISP with comparatively expensive data plans and inherent latencies that make remote work in the area a difficult/impossible proposition.
In this article, we will look at the ambitious telecom infrastructure project for rural connectivity, Starlink, that seeks to bridge this digital divide and even serve as the backhaul for large percentages of all internet traffic as it evolves. From its inherent strengths and weaknesses to the converging next-gen technologies (5G, AI, SAGIN) that can make all of what Elon Musk promised a reality and more.
LEO Satellites – Starlink’s Promise to Shrink the Digital Divide
The weekly software updates to Starlink’s CPE and the recent launch of 52 more satellites have delivered incredible metrics for the rural-connectivity WISP to its 10,000+ beta-testing subscribers.
Yet pervasive intermittent coverage gaps, equipment costs and life-expectancies of the V-LEO (Very Low-earth-orbit) satellites and its bottlenecks still have industry veterans doubting the scalability and many of the claims from the aspiring Mars colonizer.
Starlink Strengths:
- Wide Coverage
- High-Bandwidth
- Low-Latency (~40ms)
- Ultra Low-latency (inter-satellite laser links physically faster in air than glass)
Starlink Weaknesses:
- Reliability/Coverage Gaps (depending on the density of satellites available in the future these intermittent line-of-sight signal losses can become scarce or eliminated)
- Terminal Costs (1500 USD currently sold at a loss to consumers for 500USD)
- Terminal Size (59cm dish for phased-array technology, can theoretically be slightly smaller but this still limits many integrated/low-power IoT applications)
- Startup Costs (Launching satellites, even for industry-leader SpaceX, requires sizable investments in the billions that will not see returns for many years)
A great way to subsidize loss-leading solutions like rural connectivity might be the venture into more profitable applications of the evolving infrastructure, that’s where 5G backhaul and last-mile resellers come in.
5G – Cost-Effective Next-Gen Wireless Services
The most difficult part of deployments in rural areas has always been laying the actual physical fiber, microwave links, and cell towers needed to service these remote and low-density areas. An innovative take on the best coverage-per-node network solution around (satellite) might be the only way to cost-effectively scale out rural broadband and meet demands.
Far from directly competing with mobile or fixed wireless operators, the president of SpaceX Gwynne Shotwell has reiterated their plans for Starlink to be a profitable infrastructure venture across the telecom industry and a boon for the economy.
5G backhaul in rural areas has been a difficult prospect, with neither Fiber, Microwave, or traditional GEO satellites providing the capability to cost-effectively deliver the services.
LEO-based satellites alone are far from ready for this either, and with the economic costs of unmet demands an increasing factor, a more dynamic and complex approach is being adopted and the likely key to Starlink’s long-term profit ambitions – SAGIN (space-air-ground integrated network).
AI – Solving Space-Air-Ground Integrated Networks Complexities
The High dynamics of V/LEO, GEO, and MEO, ground base stations, and the different networking technologies in these layers are difficult to integrate effectively. Throw the large bag of technologies that are involved in the 5th generation of mobile carrier networks and you have an extremely complex task at hand. While Elon has initiated Starlink with mostly LEO technologies, his future roadmap leads to this convergence of technologies requiring increasingly sophisticated and autonomous systems that can learn, predict and federate the infrastructure to scale.
That’s where AI comes in, with machine learning, deep learning and reinforcement learning driving the latest advances in automation and analytics, it comes as no surprise that AI has become a key enabler for the future of Starlink, 5G and other SAGIN solutions.
To maximize its efficacy these solutions will require flexible software-defined infrastructure, and the best way to scale such limited resources starts from the edge.
Enterprise-Edge SD-WAN – Scalable QoS with Application and Network Intelligence
When it comes to cost-effectively scaling such a high-demand and critical commodity, intelligent usage, optimizations, and mitigation techniques need to be implemented well in advance—all coordinated with the fundamental limits of the service.
Innovative and longstanding QoS-focused technologies like MPLS and SD-WAN have shown how effective customer-premise application and network edge intelligence can be across unreliable or limited networks alike.
This move towards end-to-end application awareness can be seen throughout service providers well-before Starlink’s inception, such as the commoditized flexible hardware evolutions (universal Customer Premise Equipment CPE’s, Software Defined Networking SDN and Network Function Virtualization NFV ). These investments in disaggregated networks have paved the way for software-defined networking and network function virtualization that removes centralized compute and networking resources to distributed workloads closer to the edge to reduce latency for workloads what can be processed at the edge and to visualize entire network topologies and effectively manage its resources.
SD-WAN, being a state-of-the-art data-focused business connectivity solution that overlays a flexible network on top of current ISPs, harnesses these on-the-fly network intelligence capabilities across the entire network infrastructure.
