Network Trends
Networks are changing faster than any textbook can print. This lesson covers the trends a modern network professional must know: BYOD, collaboration, cloud, data centres, virtualization, software-defined and programmable networks, automation with AI/ML, and the explosion of IoT — each with its benefits and the new attack surface it creates.
Bring Your Own Device
Bring Your Own Device (BYOD) is a policy that lets employees use their personal laptops, tablets, and phones to access corporate data and applications. The phone in someone’s pocket is often newer and more capable than the company machine, so organisations have largely stopped fighting the trend and started managing it.
| Benefits | Risks and costs |
|---|---|
| Higher productivity and flexibility; employees use the devices they know. | Lost or stolen personal devices can expose corporate data. |
| Lower hardware spending for the organisation. | Uncontrolled, unpatched devices joining the network. |
| Meaningful for remote and hybrid work. | Need for policy, onboarding (for example 802.1X), and mobile-device management. |
| Employees stay reachable outside the office. | Extra network capacity and support burden. |
The technical response to BYOD includes network access control, device compliance checks, separating work data from personal data, and the ability to wipe corporate data remotely.
Online collaboration
Cisco groups the tools people use together under online collaboration. The three broad categories are:
- Video and conferencing — real-time meetings with audio, video, and screen sharing.
- Shared applications and files — documents, calendars, and whiteboards edited by several people at once, typically hosted in the cloud.
- Contact centres — systems that route calls, chats, and messages to agents and track service levels.
Collaboration is what makes remote and distributed work possible, but it also pushes delay-sensitive real-time traffic onto the network — another reason the QoS pillar from Lesson 7 matters.
Cloud computing
Cloud computing means using computing resources — servers, storage, databases, networking, and software — delivered over a network on demand, paid for as you use them, instead of owning and running everything yourself. The deployment models describe who uses the cloud; the service models describe how much the provider manages.
Deployment models
| Model | Description |
|---|---|
| Public | Infrastructure owned by a provider such as a cloud vendor and shared by many customers over the Internet. |
| Private | Cloud infrastructure dedicated to a single organisation, run on-site or by a third party. |
| Hybrid | A combination of public and private clouds with orchestration that lets applications and data move between them. |
| Community | Shared by several organisations with common requirements; a variant of private/public depending on context. |
Data centres
A data centre is a facility packed with servers, storage, and network equipment, designed to run applications and hold data around the clock. A provider data centre is often called a server farm.
Because so much depends on them, data centres are built for high availability: redundant power (including generators and batteries), redundant cooling, multiple network paths, physical security, and backup systems. The large collection of virtualised servers inside a modern data centre is why the next two trends — virtualization and SDN — are so closely tied to it.
Virtualization
Virtualization runs many virtual machines (VMs) on one physical server. Each VM behaves like an independent computer with its own operating system, but it shares the underlying hardware. The software that creates and manages the VMs and allocates CPU, memory, and storage to them is the hypervisor.
Virtualization lets one server do the work of many, improves hardware utilisation, and makes it easy to provision, copy, move, and back up machines. The same idea applied to networking gives us virtual network functions (VNFs): network services such as a router, firewall, or load balancer delivered as software running on a VM instead of a dedicated hardware appliance. Moving network functions from specialised boxes to software is called network functions virtualization (NFV).
Software-defined networking
A traditional router does two different jobs at once: it decides where traffic should go and it forwards the traffic. In SDN these two jobs are separated:
- Control plane — the logic that decides the path (routing, policy).
- Data plane (forwarding plane) — the actual movement of packets based on those decisions.
The controller talks to two sets of things through two interfaces:
- Northbound interface — between the controller and the network applications that express business policy.
- Southbound interface — between the controller and the physical or virtual devices it programs.
SDN also brings the traffic-direction terms north–south and east–west. North–south traffic enters or leaves the data centre (a client talking to a server); east–west traffic moves between servers inside the data centre. Modern data centres carry a great deal of east–west traffic, which shapes how they are designed.
Controller-based programmable networks
A controller-based, programmable network applies SDN principles to the enterprise. Instead of logging in to each switch and router separately, an administrator defines intent — “this group may reach that service” — and the controller translates that policy into device configuration automatically.
Cisco’s enterprise implementation is Cisco DNA (Digital Network Architecture), with DNA Center as the controller that provides a single dashboard for designing, provisioning, monitoring, and automating the network. Because policy lives in software rather than in each device, changes are faster, more consistent, and less error-prone than manual configuration.
Automation, AI/ML and the Internet of Things
Closely linked to programmable networks is network automation: using scripts, APIs, and templates to configure and operate devices instead of typing commands one box at a time. Application programming interfaces (APIs) and model-driven approaches let software talk to network devices, and tools such as configuration management systems push consistent settings to many devices at once.
On top of automation, artificial intelligence (AI) and machine learning (ML) are used in network operations to analyse the enormous volume of telemetry a modern network produces. Typical uses include detecting anomalies, predicting failures before they happen, and recommending or applying fixes — often called network assurance or intent-based networking. The goal is a network that largely manages itself.
The Internet of Things
The Internet of Things (IoT) is the connection to the network of everyday physical objects: sensors, cameras, thermostats, appliances, vehicles, and industrial and medical equipment. These devices collect and exchange data, enabling automation and new insights in homes, cities, factories, and healthcare.
IoT also dramatically expands the attack surface. Many IoT devices have weak default passwords, rarely updated firmware, long service lives, and limited security features, yet they are always online. A compromised IoT device can be recruited into a botnet and used to attack others. This is why securing IoT — changing defaults, segmenting devices, keeping firmware current, and monitoring behaviour — is a major security theme of modern networks.
Trend, benefit, attack surface
Every trend brings a benefit and a new way in. Learn them as pairs.
| Trend | Benefit | New attack surface |
|---|---|---|
| BYOD | Productivity, flexibility, lower hardware cost. | Personal, unpatched, easily lost devices holding corporate data. |
| Online collaboration | Remote working, real-time sharing. | External meeting exposure; sensitive data in shared cloud apps. |
| Cloud computing | Elastic capacity, pay-as-you-go, global reach. | Misconfigured cloud storage; identity and access management risk; data off-site. |
| Virtualization | Better hardware use, easy provisioning. | VM escape and lateral movement between VMs on one host. |
| SDN / programmable networks | Centralised, automated, consistent policy. | The centralised controller becomes a high-value target. |
| Automation & AI/ML | Faster, error-free operations; predictive assurance. | Exposed APIs and automation credentials; poisoned data. |
| IoT | Data, automation, efficiency. | Weak defaults, unpatched firmware, always-on devices in botnets. |
Review
Check what you have learned with these review questions.
Sources & references
- Mell, P. & Grance, T. NIST SP 800-145: The NIST Definition of Cloud Computing (2011). csrc.nist.gov.
- Cisco. CCNA (200-301) Exam Topics, Network Fundamentals and Network Automation. cisco.com.
- Cisco. Cisco Digital Network Architecture (Cisco DNA) documentation. cisco.com.
- ONF. Software-Defined Networking: The New Norm for Networks (Open Networking Foundation). opennetworking.org.
- ETSI. Network Functions Virtualisation (NFV). etsi.org.
Target exams: CCNA 200-301 v1.1 · CyberOps Associate 200-201. Last content verification: 2026-10-11.