Software-defined networking (SDN) is gaining prominence among technologists for its disruptive quality, vis-à-vis the traditional network. SDN is an approach to networking that enables network nodes to be managed through programming, rather than traditional system administration methods.
SDN will become important as IoT matures and its demands on the network increase. If you are wondering about how critical SDN will become, well, experts predict that IoT could comprise a whopping 21 billion devices by 2020. In order to better support these devices, the SDN market will burgeon to $132.9 billion by 2022, predicts a report by Allied Market Research, a global market research firm.
For SDN to become the norm, the growing number of mobile and other connected devices must begin to process the data they create between them. Additionally, the sectors interested in accessing data from these connected devices will need to create security related processes to transfer and store this data. Experts have a term to describe the changes the network will undergo to cope with IoT, and it is called elasticity. That’s industry jargon for how a network responds in an agile way to data transfer.
SDN’s elasticity becomes significant as we enter an era dominated by Big Data. Traditional networks are not equipped and certainly weren’t designed to deal with the flow of Big Data into the average enterprise. Big Data must be parsed to help, not hinder, an organization. That’s why SDN, with its flexibility, allows for the right kind of data management in this new age.
For an example, let’s turn to the healthcare sector. In the near future, hospitals are likely to have a large presence of machinery and patient end-point devices. All of these are also expected to be IoT-enabled and connected. Thus, one may come across a patient hooked up to a medical imaging device that is bandwidth-heavy on the hospital’s network. To cope with this real-time load on the network, hospitals and other medical enterprises will need to have flexible bandwidth.
Adding to this need, is that hospitals are dealing with personal data that is strictly regulated, and will thus have to factor in end-to-end data security. If the hospital machinery which in this case is a portable imaging device, needs to be moved around the premises, the instrument will have to be connected in the new spot immediately and the same network policies and controls it had in the earlier network will need to be pushed dynamically to the new location as well.
This is a flexible requirement and needs the network to be elastic and respond with agility. The need for elasticity around reams of Big Data is what has necessitated SDN; now you can see why this new way of dealing with an ultra-connected world is expected to have tremendous opportunity for growth in the near future.
I see SDN becoming significantly advantageous in a number of ways:
First, it will make the network aware of its applications. An IT specialist at the hospital in the above example can utilize a remote monitoring mechanism to realize the level of awareness of appropriate applications via SDN. The specialist leverages technology to deal with situations in a particular and specialized way.
If the network traffic is heavy during a certain time of day, the specialist can ensure he routes applications in ways that will deal with that heightened traffic level. What makes SDN so attractive is that it allows the specialist to develop a policy to make apps central and optimized during such times of stress on the network.
The bandwidth that a network demands becomes much easier for an enterprise to govern when using SDN because of its elastic nature. Suppose your organization has both a field office and a switching device as part of the IT scenario. By using SDN, you can deal with these varied demands in an elastic manner, providing the bandwidth to the entity or office that needs it most at a certain moment.
There are many aspects of SDN that are dynamic, including QoS (quality of service). Specialty switching instruments can respond to QoS by making requests and creating preferential treatments of packets within the network. Better still is that the SDN controller can add the extremely advantageous application-awareness feature. Doing so lets the application become aware of the preferential treatment that is provided.
Coming back to the connected medical scenario and applying some of the capabilities outlined above. Application awareness enables the network to identify that the heavy data is coming from a critical medical device. If the network is running low on bandwidth at that point in time, either a preferential, application-aware routing or QoS remarking and treatment can be accomplished dynamically.
With self-learning the hospital will know that there are particular days of the week or hours of the day when there is heavy usage of the data exchange (such as remote patient consultancy). During these times, it can adopt an elastic bandwidth based on requirements through a simple-to-use interface.
Finally, to secure end-to-end data, SDN-enabled adapters are made available to close the end devices and dynamic policies can be injected in real-time. These adaptors can make the portable medical devices connect to the network very fast and download the policies automatically.
So there you have it: The IoT and SDN are two technologies that very much depend on each other. SDN technology can better prepare a network for a successful and robust IoT. It provides the agility and elasticity, which IoT demands. Moreover, it provides an open environment for application developers to develop innovative tools and software connecting the IoT more effectively. Both of these technologies supplement each other in bringing us a better — and vastly more connected — world.
Author
Sreekanth. S.S is a Principal Technology Architect at Infosys. For more posts, visit InfyTalk
