Wide Area Networking is shifting towards AI-driven, edge-based architectures. In many cases this may be to mitigate latency and packet loss to gain higher performance, but they bring with them new challenges related to complexity, security and reliability. Data Centre Solutions asks David Trossell, CEO & CTO of Bridgeworks to explain how this shift protects data and business continuity.
March 20, 2026
While artificial intelligence is increasingly being used to manage them, it’s often an afterthought and burgeoning AI workloads must cope with a rapid, complex shift from legacy infrastructure to SD-WANs and SASE. Adding to the issues are the risks associated with relying on automated network agents.
The editor of CSE Icon wrote in December 2023, ‘Top 5 Challenges in Edge Computing and How to Overcome Them.’ He wrote that edge networks suffer from limited computational resources because “edge devices work with limited processing power, memory and storage capabilities.”
The solutions to this issue require the use of lightweight algorithms to reduce code size and the memory consumed by edge devices. Data filtering can be used to “leverage algorithms that filter and prioritise data at the edge so you can reduce the size of the information that needs to be processed,” and then there is edge-cloud collaboration. This is a hybrid approach, whereby tasks that require more resources for edge devices are loaded into the cloud.
Challenges at the edge
While the proximity of edge devices to data sources reduces the volume of data transmitted to centralised servers, there can still be challenges with managing local network conditions. They can still experience congestion and interference, which may limit their performance due to localised network latency.
The solution is to locate edge device closer to the data source to reduce the distance the data has to travel to minimise latency. Content Delivery Networks (CDNs) can also be deployed to optimise the distribution of data closer to edge devices. Then there is edge caching, which is used to store frequently used in information at the edge. This also means that less critical data can be sent to centralised servers.
The three other critical issues include data security concerns, such as surface attacks; communications issues between devices – including their interoperability; and efficient data management and storage. Edge data volumes are increasing daily, and so the data could be lost if it isn’t managed and stored appropriately. This may then lead to delayed decision-making and create significant security risks.
The answer to the latter is tiered storage, which requires data to be categorised based on its importance and access frequency. It also necessitates edge analytics and filtering to ensure that only the data you need as an organisation is stored. Lastly, local edge caching enables frequently accessed data at the edge to be stored with local caching to reduce time required to access the data. So, while there are benefits to edge networking, some significant challenges remain.
Proximity is a risk
A key danger is their proximity to each other. In summary, Google AI explains: “Having edge networks in close proximity—specifically, deploying numerous edge nodes, IoT devices, and computing resources near one another—increases the complexity and risk associated with managing these systems. Key risks include intensified security vulnerabilities, heightened potential for physical tampering, signal interference and increased management complexity.”
However, a shift to the edge makes sense when the aim is to avoid sending large amounts of data over a long distance. This is in contrast to WAN Acceleration.
Rather than being in close proximity to the source of the data, WAN Acceleration uses in-built artificial intelligence, machine learning and data parallelisation to mitigate latency and packet loss. It also mitigates the need to store data within the same circles of disruption, allowing data centres and disaster recovery sites to store data at a distance from each other – including on the other side of the world. This technology aims to accelerate data transfers by mitigating the effects of latency and packet loss, while strengthening bandwidth utilisation.
WAN Acceleration can work in conjunction with edge networks when large volumes of data need to be backed up, restored, sent, received and analysed in the cloud. Its primary purpose is to enable the accelerated transfer of large and encrypted file transfers, data replication, remote backups and cloud migration, where a high level of latency exists. Edge networks are best for real-time applications such as for IoT, and autonomous vehicles to reduce latency for immediate feedback. However, both technologies can be used to improve application performance.
WAN data traffic volumes
A Google AI summary of WAN data traffic volumes for 2026 from sources such as Ericsson, Computer Weekly and Gartner says, “Global Wide Area Network (WAN) and overall IP data volumes are projected to reach significant milestones in 2026, driven by a surge in AI workloads, 5G adoption, and cloud-intensive applications.” WAN data volumes are expected to reach “602.1 exabytes (EB) per month, a 15.4% increase from previous levels.”
In comparison, the Edge Data Center Market Report 2026, published by The Business Research Company says: “The edge data centre market size has grown exponentially in recent years. It will grow from $16.55 billion in 2025 to $21.17 billion in 2026 at a compound annual growth rate (CAGR) of 27.9%. The growth in the historic period can be attributed to growth in mobile data traffic, expansion of cloud computing adoption, increasing use of content delivery networks, rising demand for real-time data processing and growth of distributed IT infrastructure.”
A need for WAN Acceleration
While industry reports expect edge networks to be dominant this year as they don’t require any centralisation, there will still be a need for WAN Acceleration, which can even improve the performance of another increasingly popular technology – SD-WANs. They are a great technology, but their performance can be boosted with a WAN Acceleration overlay. WAN Acceleration also can also place data out of reach to ensure regulatory compliance and service continuity, since data centres and disaster don’t need to be located in the same circles of disruption.
So, yes, there may be a shift to the edge, but WAN Acceleration is still the right tool for when your organisation needs to shift large volumes of data to the other side of the world. In contrast, and while each technology has their own specific use, the danger with any data that is situated in close proximity, is that when a disaster strikes, there is a risk of significant disruption.
To conclude, if large volumes of that data is properly backed up, and stored elsewhere, WAN Acceleration can help to restore it quickly, and more securely than WAN Optimisation can. This includes data that wasn’t originally at the edge, or isn’t at it, as WAN Acceleration is a completely data agnostic technology. Therefore, not all WAN networking needs to shift to the edge – particularly as edge networking isn’t always the right answer to mitigating latency, packet loss and poor bandwidth utilisation.
