Enterprise network undergoing a radical transformation

As enterprise networks become more complicated, distributed, and congested; traditional network architectures are undergoing a fundamental change.

In 2002, when Jeff Bezos founded Amazon Web Services (AWS), it was because he was very uncomfortable on how much time his developers spent on the undifferentiated, heavy lifting needed to create the infrastructure that was necessary to run the software they were building. He felt the need to allow enterprises to consume computers (and later storage) as-a-service. And following that, AWS experienced meteoric growth.

It may seem crazy to consider not building your own networks. But it was crazy to consider using software-as-a-service 25 years ago, and that turned out well. Same thing for compute, storage, and virtually everything else.

As organizations continue to look for ways to improve productivity, as-a-service models like networking-as-a-service (NaaS) represent a way to support new workloads and business requirements with confidence. It enables business resiliency and agility efforts all within budget requirements.

It is time to consume NaaS now. That means provisioning the network through a simple cloud portal. It means relying on a hyperscale prebuilt network instead of designing, purchasing, and building your own bespoke network. It means no longer worrying about hardware throughput or software licensing levels. It means an enterprise can once again receive SLAs that guarantee availability and reliable, predictable performance. And it means world-class security and the ability to maintain compliance with complex data sovereignty regulations.

NaaS is the logical outcome of many business processes moving to the cloud. It provides the networking hardware, software, and operational/maintenance services as an operational expense instead of the traditional upfront expense. In today’s fast-paced world, NaaS is the perfect fit as businesses can deploy and scale network connectivity quickly.

The growth in internet dependence is really what has been driving the cloud, because high-quality, interactive, user interfaces are critical, and the cloud’s technology is far better for those things, not to mention easier to employ than changing a data center application would be. A lot of cloud interactivity, though, adds to latency and further validates the need for improvement in internet latency.

Interactivity and latency sensitivity tend to drive two cloud impacts that then become network impacts. The first is that as you move interactive components to the cloud via the internet, you are creating a new network in and to the cloud that is paralleling traditional MPLS VPNs. This does not displace MPLS VPNs immediately, but it does necessitate elevating connectivity management above them; think virtual-network-over-MPLS. Since network connectivity is what users see of a network service, this pulls differentiation away from the physical transport elements of the network. That takes a critical step toward converting virtual network technologies into the framework for network-as-a-service, or NaaS.

The second as hosting is being encouraged to move closer to the edge to reduce application latency. That puts pressure on the network to reduce the latency it contributes, which is one justification for 5G, but also a reason to wring latency out of wireline broadband services.

NaaS can have different definitions, depending on who defines it. At the highest level, NaaS is defined as network infrastructure hardware, software, services, management, and licensing components consumed in a subscription-based or flexible consumption model. NaaS is different from other traditional as-a-service models that take advantage of cloud and virtualization capabilities because a significant amount of on-premises cabling and distributed networking equipment are required for network connectivity operations. Additionally, organizations may still require WAN interconnections to the cloud, which may or may not be included in a NaaS offering.

The variations in definitions can often be attributed to differences in the capabilities and expertise of the assorted providers. Public and hybrid cloud service providers, such as AWS, Microsoft Azure, HPE GreenLake, etc., usually include the packaging of compute, storage, and networking components as part of their configuration and implementation to provide a comprehensive solution. And NaaS deployment can be seen as a completely managed service with varying levels of service provided by the cloud computing service vendor. OEM enterprise network equipment vendors, like HPE Aruba along with their partners, provide NaaS solutions leveraging their R&D and deployment expertise in enterprise LAN, remote branch, and SD-WAN connectivity use cases as a foundation for their NaaS offering. The categorization falls primarily on providing the equipment through subscription-based or consumption-based acquisition, eliminating the upfront capital spending tied to purchasing the equipment. In these cases, the implementation of internet connectivity and cellular services are still required from a telecommunications carrier and internet provider. Network vendors would manage the hardware and software in conjunction with the internet provider and telecommunications carrier, leaving the overall management of the network infrastructure to the network vendor, partner, or internal staff. Like telecom providers, organizations are given a choice of deployment options that best fit their business needs. Organizations have the flexibility to control and manage parts or all of their network infrastructure. They have the choice to pass the baton of control completely to the OEM network equipment vendor or vendor partner.

Telcos across the world are on a journey to migrate applications and workloads to the cloud to embrace cloud-native principles and, ultimately, to deliver automated systems. But the sheer complexity of legacy IT systems makes such ambitions extremely challenging. To unleash the potential of new technological advancements to drive growth, it is imperative that they adopt a holistic approach across their operations.

Telcos could seize USD 75 billion of the USD 150 billion NaaS market. However, their investment strategy, business, operational, and go-to-market models are not ready to deliver a competitive NaaS solution, says Reece Hayden, distributed and edge computing analyst at ABI Research. “The market is immature and highly fragmented, but telco market revenue will exceed USD 75 billion by 2030 if they act now and transform technology, culture, and structure to better align with the requirements of the NaaS market.”

Telcos offer NaaS connectivity to enterprises, much in the same way they offer cellular services to individual consumers for wireless connectivity. They like organizations to leverage their bundled WAN interconnection capabilities exclusively. The NaaS offerings can be inclusive of a full-service offering with the managed services component or be extended simply to private VPN services or SD-WAN options so that organizations can continue to manage their local network infrastructure internally.

Telcos need to develop vertical-specific sales strategies, and adopt a consultative process to help bridge the gap between enterprises’ awareness of NaaS, and their understanding of it. To drive short-run sales, suppliers must educate and tailor their sales strategy to focus on first adopters – start-ups and SMEs – and specific verticals. Finally, when it comes to structure, ABI recommends telcos reduce internal fragmentation, focus on cross-business service continuity, and establish strong partnerships across the industry.

And if telcos do not get their collective act together, interconnection providers and hyperscalers will be only too happy to fill the void. ABI notes that the likes of cloud connectivity providers Megaport and Packet Fabric already offer agile NaaS solutions, while Amazon Web Services (AWS), Google Cloud Platform (GCP), and Microsoft Azure have extensive cloud-focused NaaS offerings and huge reach.

The NaaS market is still nascent. Primed for growth, according to ABI Research, by 2030, nearly 90 percent of global enterprises will have migrated at least 25 percent of their network infrastructure to be consumed within a NaaS model. The growth is being fuelled by increasing enterprise demand for cloud-native agility, multi-cloud accessibility, and services that can dynamically scale to support digital transformation. This demand means the NaaS market could be worth as much as USD 150 billion by 2030.

With the expanded interest for the market in the worldwide market, market players from the NaaS market are expected to profit from worthwhile development possibilities later on. Top key players are Cisco Systems, Juniper Networks, IBM Corp, NEC Corp, Vmware, Aryaka Networks, Alcatel Lucent, Brocade Communications Systems, AT&T, and Ciena Corporation.

While, 90 percent of enterprises are expected to adopt NaaS by 2030, the market currently remains stagnant.

2023 so far has seen a normalization of buying patterns amongst cloud and service provider customers, with overall demand moderated in the JFM quarter. This trend is expected to continue through at least the first half of the calendar year. Customers are more closely scrutinizing spending plans and deployment timelines due to the economic uncertainties that are happening around the world.

While the SD-WAN market should expect continued growth this year, its rate of growth has slowed since high demand in the depths of the Covid-19 pandemic, and will start to experience some level of modulation. The spike in demand during that time has left vendors still dealing with backlogs, which might be acting as a buffer in preserving the rate of growth for the SD-WAN market.

Some enterprises are moving to internet services for cost reasons as outdated WAN equipment requires the replacement of traditional branch routers, according to Gartner. Legacy equipment is often being replaced with SD-WAN appliances and solutions, which is causing a relook at the WAN overall. The analyst firm expects that by 2026, 45 percent of enterprise locations will exclusively use internet services for their WAN connectivity.

This tallies with the IDC report that the enterprise segment of the worldwide wireless LAN (WLAN) market rose 47.9 percent year-over-year to USD 3.35 billion in the fourth quarter of 2022 – its highest quarterly revenue to date.

Brandon Butler, research manager at IDC, attributes this growth to the “continued easing of the component shortages and supply chain disruptions, combined with enterprises investing in the newest Wi-Fi standards. The strong performance of the enterprise WLAN industry shows the importance of wireless technology in the network and digital transformation goals of organizations across the globe.”

Enterprises have yet to take advantage of SASE. While, in 2022, only a minority of fully deployed SASE in 2022, they forced their vendors to have a SASE option. Deployment of SASE networking versus security technologies remained extremely disaggregated and on different timelines due to the difficulty of changing too many parts simultaneously.

By contrast, the security side of SASE, security service edge (SSE), is poised for another year of solid growth in 2023. As private sector organizations embrace digital transformation and move their operations to the cloud, they too are looking to zero trust to help provide a robust and secure network infrastructure. SASE has emerged as a cloud-delivered convergence of network access and security services, and is a common approach for enterprise zero-trust adoption. The challenge, however, is that in many organizations, responsibility for networking and security lives in different parts of the organization and these groups often rely on different vendors in their respective areas. Breaking down the silos between security and networking teams and choosing the right tools, products, and vendors to align with desired business outcomes is critical to implement zero trust in larger enterprises.

Outages. With every business, now effectively an online business, we have reached a tipping point. A minor disruption to a single provider or component can have a major domino effect on the wider web. Over the last two years, in a post-pandemic landscape, some of the most significant outages affected major infrastructure and service providers.

There was Meta’s well-publicized outage of 2021 when Facebook, Instagram, WhatsApp, Messenger, and Oculus VR experienced simultaneous and prolonged downtime worldwide. This was followed by AWS that took down major online services such as Amazon, Amazon Prime, Amazon Alexa, Venmo, Disney+, Instacart, Roku, Kindle, and multiple online gaming sites.

Amazon’s recent search failures showed that an outage does not even need to take down the entire website to damage it. For 22 hours, around 20 percent of worldwide users were impacted, with a subset of users unable to use the search function completely the entire time. Imagine the impact on global revenue.

Building internet resilience has become a top priority at the board level for all these reasons. A growing number of companies are augmenting their monitoring capabilities with internet performance monitoring (IPM), built from the ground up to provide comprehensive visibility into the enterprise internet stack. IPM will become an essential tool for organizations that rely heavily on the internet as their network infrastructure, which, after all, is every business out there.

The pandemic exposed weaknesses in the ability of traditional network architectures to support distributed employees at scale, and while organizations managed through the crisis with quick-fix solutions like remote access VPN, fundamental changes to the architecture are needed for long-term success.

There are some core networking priorities that need to be focused on. The first of those priorities is automation. Enterprises are moving into a new phase of SDN. From proprietary benefits of SDN, they are beginning to build their own automation platforms on top of the vendor platforms, and most likely multivendor environments.

With the arrival of holistic digital transformation platforms, offering broader integrated capabilities, leaders are able to focus on business outcomes and customer journeys by offering a staged approach to achieve end-to-end automation without risking increased technical debt. This is because these platforms leverage a customer journey-oriented approach that traditional technology-oriented automation approaches do not. Digital transformation, going forward, is about platforms that bridge the silos operationally as well as technologically.

The enterprise seems to be rapidly losing control of its networks. The shift to the cloud has left the traditional enterprise network stranded, and AI and automation are required if enterprises hope to regain control. Cloud is now very much multi-cloud. Customers benefit from the vast potential of combining the top hyperscalers, which is equaled only by the frustrating management and cost concerns. It is growing at a robust 14.4-percent CAGR and forecast to expand from its 2022 total of USD 483 billion to USD 1.5 trillion in 2030.
Robotic process automation (RPA), combined with holistic digital process automation and content services, is becoming potent beyond task automation to enable 100-percent automation in the end-to-end customer journey.

AI. The stunning debut of ChatGPT in November brought to the fore that AI is growing exponentially and offering a toolset that would have been sci-fi not that long ago.

As enterprise networks become more complicated, distributed, and congested, AI is helping resource-strapped network teams keep up. AI-powered tools are already spreading throughout cloud and enterprise networks, and the number of tools that feature AI will continue to rise in the foreseeable future. AI is currently being used for a wide range of network functions, including performance monitoring, alarm suppression, root-cause analysis, and anomaly detection.

More realistic use cases are emerging where AI and ML can be effectively leveraged in combination with other automation technologies. AI is forecast to increase at a 38-percent CAGR, leaping from 2021’s USD 147 million to USD 1.6 trillion in 2030.

The most recent evolution of AI is the most important development in predictive network technology. AI can detect anomalies quicker than humans and can even analyze the root cause of an anomaly, helping to guide a technician to understand and repair the issue faster than before. AI can correlate alerts and error conditions across many disparate systems, discovering related patterns in minutes or even seconds, something that would take humans hours or days.

The integration of AI tools into predictive network technology also has the potential to be an economic game-changer. In addition to bottom-line economic benefits, AI helps to simplify management, either within an enterprise or across a service provider’s portfolio.

Generative AI, a type of artificial intelligence is poised to have a profound impact on businesses and society.

Clearly, as the technology advances, the capabilities will expand. If generative AI can match or exceed human performance for many tasks, the nature of work – and roles within organizations – will change dramatically. Some roles and job functions will disappear, while new roles will likely replace them. However, this displacement could rival or even exceed past events, such as the Industrial Revolution.
As neural nets and GPUs continue to advance, and AI algorithms become more refined, the ability of machines to perform human tasks will increase. Whether generative AI will lead to singularity, the hypothetical point where AI exceeds human intelligence, remains to be seen. However, it is entirely clear that generative AI is poised to change the nature of business – and the world around us.

Quantum networking, in the early stages of building the underlying infrastructures and validating foundational concepts, is gaining interest from researchers and vendors. Quantum networks are expensive and slow. Some companies are already experimenting with using quantum networks for QKD with the goal of future-proofing communications, and EY is one of them. The professional services firm is using quantum key distribution to secure communications between two offices in London. EY is the first customer in a three-year trial of a quantum-secured metro network, launched last April by British Telecom and Toshiba that connects multiple physical locations over standard fiber optic links using QKD.

Europe is one of the global leaders when it comes to the deployment of quantum networks. The other is China, which set up the world’s largest quantum network in 2021. The network combines 700 optical fibers with two ground-to-satellite links and is able to achieve QKD over a distance of 4600 kilometers, or about 2860 miles. This year, two different US companies, are planning to deploy the first US-based commercial quantum network – EPB Quantum Network in Chattanooga, Tennessee, and GothamQ in New York. Half-a-century ago, when rudimentary classical networks were first being invented, nobody could have predicted how the internet today has evolved. Quantum networking is in a similar phase today.

And last, but not the least is digital twins. Creating a living and breathing network diagram with network change preparation, troubleshooting, and compliance benefits they result in more confidence and increased implementation speed. The technology, still in its infancy, looks promising for enterprise network teams. Future iterations will result in true infrastructure clones that act identically to their production counterparts. This gives teams a virtual playground with which to test new applications and architectures at a rate that was considered impossible using traditional methods. In other words, the digital twin network market is setting itself up to be a tool that will likely revolutionize how networks are managed, monitored, and upgraded, which goes far beyond what is possible today.

2023 is a year when companies pivot innovation efforts from creativity to resilience and efficiency across a range of technology domains. Enterprises cut back on creative innovation efforts last year due to rising costs and economic unease, and this year is the time to double down on strengthening the durability of the core business building around differentiation.