Exploring Edge Computing and the Internet of Things from a Data Center Perspective

Along with the digital transformation of various industry organizations, many changes have taken place. Digitalization is not just a marketing term, the enterprise becomes a digital entity, which means that they can support and run advanced use cases, mobile users, and new types of use cases.

Enterprise users and its platforms are now more widely distributed. And, many of them are associated with network-connected devices, users, and the Internet of things. In addition, the market will continue to grow. New research by Accenture, a research firm, found that the medical internet of Things (IOHT) has achieved cost savings, but the sustained investment is essential. The report notes that by introducing more IoT connectivity, remote monitoring and information gathering, the medical Internet of Things (IOHT) can better facilitate the use of medical resources, help people make smarter decisions, reduce inefficiencies or waste of costs, and contribute to patient health and recovery.

The report estimates that the market value of the medical Internet of Things (IOHT) will reach $163 billion by 2020, with a compound annual growth rate (CAGR) of 38.1% percent from 2015 to 2020. Growing markets are an important reason for data center operators and business managers to actively invest in IoT solutions. In addition, they invest in solutions for all users who can support IoT devices and access data. And this is where edge computing works. Ofcom’s recently released data center industry study found that 81% of respondents said edge computing was used to support and enable the Internet of things.

40% of respondents have deployed or plan to deploy edge calculations. Businesses need to think hard and understand themselves and their edge computing requirements before investing or using edge computing.

Needs assessment in infrastructure design for edge data centers

Edge Computing solutions are not just a data center site, they are smaller, use case-specific, and use an IT environment-intensive design to help organizations process more service and user data.

Use case definitions

Businesses need to look at their organization’s long-term strategy. For example, is the business growing? Will remote users be supported? Are you trying to provide a new connection service? If the enterprise believes that it has these advantages and is suitable for edge computing, then follow-up steps can be taken to prepare a good business plan and technical strategy to support the application of Edge computing. Businesses can clearly define their use cases without becoming edge computing experts. In addition, there are many vendors who can help companies complete this process. However, it was important to restructure infrastructure and operations to ensure that the strategy of the enterprise could be implemented. Recruiting the right people who can turn this vision into reality is the key to business success.

Delayed budget

For end users, network latency is the main reason why it takes a long time to download a movie. But for content vendors, the number of milliseconds required to complete this feature needs to be measured by customer dissatisfaction and cost.

In addition, for an enterprise, network latency can also mean a business loss or a competitive advantage. Even a round trip from a central data center (such as a data center facility located in a primary market) at the speed of light can mean a cumulative transfer cost. According to a study conducted by ACG Research, locally cached content processed in metropolitan data centers can save about $110 million in costs over five years.

If you apply this logic to an enterprise running an industrial IoT (IIoT) component tracking application, you can evaluate the hard cost of the transfer, but the costs associated with the degradation of application performance will be incalculable.

Security

Security is a big problem, adding a critical and complex feature to deploying Edge computing. Businesses will have to spend some extra time defining data requirements and management policies. Businesses may ask the following questions: “What happens to my data? is the data transient transmission or is it stored at the edge? What is the data being processed? What is the connection control method around the data?” Again, all of this needs to be defined and integrated into the enterprise’s own edge solution. That is, organizations can still build compliance and rules into the Edge computing architecture. However, additional precautions need to be taken to ensure data security and control. Enterprises also need to consider the location of Edge computing, storage systems, how data is processed, and access rights. The most important part is that software-defined solutions allow enterprises to integrate with core data center systems and support powerful data location policies.

This could have a significant impact on industries such as medicine, health care, and other regulated organizations.

Delay Index

Typically, people talk about “slow levels” or delays without really understanding what this means for the business or the network. So add some numbers behind the term delay. Latency is the time it takes to transfer packets over the network. Delays can be measured in many different ways: round-trip, one-way transmission, and so on. Latency may then be affected by any element of the link used to transmit data: workstations, WAN links, routers, local area networks (LANs), servers, and, in the case of very large networks, may be limited by the speed of light.

There, it will have an impact on the amount of data sent/received by throughput or unit time, as well as packet loss, which reflects the number of packets lost per 100 packets sent by the host. Therefore, when the latency is high, this means that the sender needs to spend more time (without sending any new packets), which slows down throughput growth. A recent study shows that latency has a profound impact on Transmission Control Protocol (TCP) bandwidth. Unlike the User Datagram Protocol (UDP), TCP has a direct inverse relationship between latency and throughput. As end-to-end latency increases, TCP throughput decreases. The following table shows what happens to TCP throughput when the round-trip latency increases. This data is generated by a delay generator between two computers that are connected using a Fast Ethernet (full duplex) connection. Note that TCP throughput drops dramatically as latency increases.

In addition, there is another serious problem. The packet is missing.

Packet loss will have two serious effects on data transfer speed:

  • Packets need to be retransmitted (even if only the confirmation packet is missing and the packet is routed)
  • TCP Congestion Window size does not allow optimal throughput
  • When packet drop rate is 2%, TCP throughput is 6 to 25 times lower than no packet loss.

While some packet drops are unavoidable, user performance and access permissions will be affected when this continues to occur. Regardless of the situation, keep in mind that packet loss and latency can have a profound negative impact on TCP bandwidth and should be minimized. This is where edge computing solutions play a key role. They help eliminate most of the latency by bringing critical data points and resources closer to the user.

What Edge Computing can do

So, in addition to addressing the most important latency challenges, what can edge computing do for users? Edge Computing solutions are specifically deployed around use cases.

For example, is the user trying to deliver the application or the entire virtual desktop? Or are you trying to provide data that needs to be analyzed near the user or its system? In this regard, the use of Edge computing systems can include:

  • Software-defined solutions that can be configured according to application requirements
  • branch offices and microdata centers
  • Hybrid Cloud Connectivity
  • IoT processing (e.g. Azure IoT Edge)
  • Firewall and network security
  • Internet-supported devices and sensors for the collection and analysis of real-time data
  • Connect the entire device network
  • Asset Tracking
  • Simplified Research
  • Reduce latency for specific services
  • Support delivery requirements for latency-sensitive data points and applications

What does this mean for edge data centers?

The introduction of 5G technology will accelerate the development of edge data center networks and bring them closer to end users.

The development of Edge computing and the advancement of wireless networks, the upcoming 5G to efficient mobile connectivity and data center solutions, coupled with the use of smart mobile and wearable devices, contribute to the development and growth of next-generation solutions and technologies that enrich the environment. Looking ahead, Edge computing facilities will accommodate applications that can be easily defined as “critical tasks.” With the advent of technologies such as 5G, the distance from the edge calculation to the user group can usually be measured in feet rather than miles. The combination of 5G technology and edge computing close to devices and users can provide some powerful experiences and create an amazing competitive advantage for the business. Therefore, it is important to properly plan and design the ecosystem of the Edge Computing data center.