Tristan Wood, MD of Livewire Digital, argues that while hybrid connectivity promises to seamlessly integrate several network technologies and deliver improved performance, flexibility, and dependability, on closer inspection, it frequently fails to deliver.
The capacity to make use of more than one network is often referred to as ‘hybrid’ – and in a classic networking approach, this simply means an ability to ‘failover’ to one or more backup services. Hybrid in this sense does not sufficiently represent what true hybrid networking is capable of.
A fully hybrid network, also known as a ‘heterogeneous’ network, combines and binds together a number of bearers – such as cellular, satellite and Wi-Fi – into a single ‘pipe’ that provides a speedier and, crucially, a more dependable service in this way.
In order to optimise performance and manage costs, a truly hybrid platform should go a step further and adapt to a variety of other variables, depending on the performance of each bearer and other environmental factors affecting it at any given time, which I’ll touch upon later.
When optimised in this way, a hybrid system can maintain the performance of TCP connections, irrespective of the availability and performance of the underlying networks. In the face of failure and degradation, it will be possible to optimise connectivity even in the most difficult situations, providing users with a genuinely consistent and uninterrupted experience. Hybrid and resilience take on a whole new meaning in critical operations, where connectivity might be the difference between life and death.
On the other hand, failover does not actively utilise the strengths and capabilities of both primary and secondary connections at the same time, instead relying solely on a backup link to take over in the event of an outage. The idea of failover is what actually calls into question the conventional wisdom about what hybrid connectivity actually implies.
The greater reliance on redundancy and resilience
It is more important than ever to be able to ‘connect’ in any location, regardless of what services are available. Voice calls, data transfer, and live video streaming are essential in a wide range of industries.
Defence will always require a variety of communication channels with backup options. Whether it be drone video footage over a forest fire in a remote location, or a rural place where cellular coverage is not an option, emergency services require access to ongoing occurrences in real-time. In order to access patient records or a clinician’s advice while stationary or on the move, the ambulance service depends on reliable connectivity. The technology behind telehealth triage has advanced at a breakneck pace; however, none of it will function without reliable, robust and high performance connectivity.
The marine industry has been an early adopter of the power of hybrid connectivity using cellular for in-shore operations and satellite for deep ocean. Utilities companies make use of machine-to-machine (M2M) and IoT capabilities to control and monitor operations on windfarms, oil rigs, and installations that are traditionally operating in areas that require the use of multiple network technologies.
Likewise, broadcasters need to transmit live news over many different networks. To deliver the highest quality video depends on optimising links, automatically correcting transmission errors and incorporating all available networks to increase bandwidth and therefore, video quality.
Different satellite, Wi-Fi, and cellular communications methods are used by the things we use everyday, including the cars we drive and the appliances we use.
The integration of networks can be hampered by factors including latency, packet loss, and network congestion, which leads to unsatisfactory user experiences. Therefore, the complexity and sacrifices it entails in the traditional approach overshadow the envisioned performance advantages of hybrid connectivity, necessitating new ways of thinking.
Quality of service, and cost
One important factor to keep in mind when it comes to cost management is that real hybrid supports parameters that control how different bearers are used, under what circumstances and by what priority traffic.
These mechanisms allow capacity on the most cost-effective bearers to be used first and to only use more expensive ones to support them, when defined criteria are met. By way of an example, if a Wi-Fi service drops below 10 Mbps, then start supporting it with cellular, and if the combined bandwidth of the Wi-Fi and cellular drop below 5 Mbps, then support this with satellite.
Overlayed on this bearer, hierarchy is logic that dictates what traffic priority can make use of which combination of bearers. This means critical traffic will always be sent ahead of lower priority traffic, but the associated packets will be delivered over the most cost-effective path available at any point in time.
For service providers, being able to offer a ‘faster and more reliable’ Quality of Service (QoS) to their customers is a very important differentiator. However, applying a QoS across managed and un-managed networks is highly complex. Again, a true hybrid approach can address this by applying QoS configurations that can integrate best effort unmanaged services with managed ones and define how and when they are used.
Compliance and security
New security considerations and challenges occur when incorporating new network technologies. Lack of a consistent security architecture may make the entire network infrastructure more susceptible to cyberattacks. Additionally, it becomes increasingly difficult to ensure compliance with business regulations and data protection standards in a fragmented network environment that includes public networks and the Internet.
A true hybrid model considers all wide area network services as ‘untrusted’ whether it is a private or public service. This allows a recognised security policy to be applied to the encryption of data and attack surfaces presented by the technology are minimised.
What makes any of this significant?
There’s an old saying that goes, “If you do what you’ve always done, you’ll get what you’ve always got.” However, things are different with networking. Doing business as usual will actually produce much worse outcomes in a networking environment supporting rapidly developing more data-hungry technology.
Therefore, we must make better use of existing and future networks. A true hybrid approach can achieve this and deliver faster, more resilient connectivity, the classic backup or load balancing model cannot. Mobility adds another dimension to the problem space; network coverage and characteristics change rapidly so relying on technology designed to link offices can lead to intermittent connectivity and poor performance.