Dharmendra Kumar,Rabyte- Director IoT

Edge computing, a strategy for computing on location where data is collected or used, allows IoT data to be gathered and processed at the edge, rather than sending the data back to a cloud. Even IoT edge computing is the practice of using data processing at the network’s edge to speed up the performance of an IoT system. Moving data processing physically closer to IoT devices offers a line of benefits to enterprise connected device…………

The massive growth of connected devices including sensors and machines is revolutionizing every aspect of human life. The centralized cloud architecture adopted in the early days of mobile internet was designed primarily to allow access to data stored on the world wide web. Humans and devices are now producing most of the data consumed on the internet. So the existing centralized cloud infrastructure is no longer efficient and sustainable. There is significant waste of network bandwidth to send terabytes of data to server farms that may be hundreds of kilometers away from the source and/or destination of data.

The prevalent centralized cloud architecture does not adequately leverage massive amounts of computing resources on smart devices which are idle most of the time.

Hybrid Edge Cloud combines the benefits of new network technologies such as 5G in private and public clouds to leverage computing resources on smart devices to build a sustainable decentralized infrastructure for the Hyper Connected World.

Cloud decentralization has several advantages. As mentioned in the Introduction, in the current centralized cloud model, as more devices are added or more content is generated by these devices, more servers in data centers must be added to support them. Using a decentralized cloud, we can create a cloud fabric that scales with the number of smart devices. This reduces the need for additional servers and the upgrade cycle of these servers in datacenters. In effect, we increase the “cloud” capacity as the number of smart devices increases. In addition,, given that most of the data are produced on smart devices, we minimize the transport costs and latencies for applications. In this new model, much of the processing is performed on devices, communication is kept as local as possible, and heterogeneous smart devices from different vendors and operating systems can collaborate and share computing and other resources. The central cloud remains a valuable resource because it may be indispensable for many applications that require global management, central storage.

The software component referred to by the Edge SDK enables any smart device to act as a cloud server. It is a collection of software libraries and their corresponding APIs. Edge SDK can run on any MCU/DSP/OS or Linux node.Once the Edge SDK is loaded the device becomes an HEC node.

Edge SDK resides between the operating system and the end-user application. Runtime environment for microservices is also provided by the Edge SDK.

Developers develop microservices on the device using Edge SDK container manager.

Node and service discovery, Auto-discovery and auto-routing for all nodes with Edge SDK in local and global networks. Node and service connection: ad-hoc edge cloud of nodes forming a self_x0002_organizing cluster. Light container, Microservice runtime environment to allow remote/local load of microservice images, start, stop of microservices. Sidecar pattern, Enable frontend application decomposition to abstract Networking, Security, Authentication function. The Hybrid Edge Cloud node can dynamically discover each other independent from the OS and/or the network. It can expose available capabilities and functionalities via API to each other. Also form and organize into clusters and can communicate within a cluster even with no Internet availability

There are several challenges to building a Hybrid Edge cloud platform.

• Fragmentation in operating systems and networks
• The availability of network resources
• Unlike servers in data centers
• Management of distributionin data centers
• Consequently network device security

The Hybrid Edge Cloud node can dynamically discover each other independent from the OS and/or the network. It can expose available capabilities and functionalities via API to each other. Also form and organize into clusters and can communicate within a cluster even with no Internet availability

Principle of Cloud Decentralization

1. Meritocracy: Meritocracy is a key principle in ensuring an efficient system design in which the use of network bandwidth and central resources is minimized. All nodes should have equal opportunities to participate.

2. Decentralized Discovery: A node must discover other nodes based on its scope. Some scopes are attached to the same network, owned by the same account holder or within the proximity of one another.

3. Clustering: Human and machine communication occurs mostly in clusters.

4. Microservice to Microservice Communications: Once a decentralized cloud fabric is formed, applications on devices can utilize it to communicate directly without a pre-assigned third-party trust element.

5. Dynamic Resource Instantiation: Signaling and data resources should be deployed dynamically based on the network conditions and demand from nodes within clusters.

6. Collaboration: To leverage their collective power, all nodes, must collaborate and share resources. The sharing of decentralized cloud resources should be seamless, as it is with server nodes in the central cloud.

7. Infrastructure Independence: The decentralized cloud platform must be agnostic to the operating systems, central cloud platforms, networks, and locations.

8. Zero Trust Security: Building a zero-trust network, encrypting all communications, and authenticating every device is feasible a limited number of devices are active within every edge cloud cluster of devices.

With a decentralized cloud, all nodes, including the severe farms in data centers can act as cloud servers, and

there is no designated permanent central element. Nodes can communicate, collaborate,  and share resources directly, generally without resorting to a central element, unless  necessary. With this approach, the central cloud resources are used only when required. For  instance, when there is a need for global storage, archiving, updating of centralized  databases, centralized registration, and so on. Any other functions that can be handled by  smart devices at the edge can be assigned to them. For instance, messaging between devices, handshaking control signals between machines, and transmitting data between devices within a small cluster.

Edge and 5G Impact

5G Infrastructures can connect many wireless devices and enable M2M communication. 5G enables data prpcessing close to the connected device, using ultra low latency and ultra high bandwidth.

Local cloud and Colo Data (Equipment/Space/Bandwith Data Centre) serve as Edge gateway.Edge compute in endpoint networks enables real time analysis and insights.

Benefit of HEC (Hybrid Edge Computing) for Cloud to make 5G efficient and economically viable, communication service providers must install computing resources on their deployment sites to minimize service latency and reduce capex and opex costs.

• Offload unnecessary traffic over the network by enabling data processing, and caching on edge devices.
• Reduce energy consumption and the carbon footprint of the MEC/3GPP infrastructure in their base stations, Edge devices no longer need to constantly send traffic over the network.
• Manage higher-value data instead of raw data, which may have significant duplication/noise, and as a result, improve the value/cost ratio of their MEC/3GPP infrastructure.
• Provide cross-industry solutions to their customers and ecosystem synergy across their customer base and seamlessly connect highly fragmented MEC/3GPP silos.
• Increase revenue by offering a differentiated private enterprise network as a service with higher data privacy and security.
• Provide a foundational approach for GDPR (General Data Protection Regulation) for their enterprise customers.
• Support GDPR compliance with built-in privacy and reduced complexity.
• Support environmentally friendly solutions and reduce carbon footprint.
• Avoid heavy device orchestration by relegating many functions to furthest edge of the network.
• Empower native microservice developers to build applications instead of being limited to embedded systems for specific industry specialists.
• The resiliency and robustness of their solutions are increased by minimizing the points of failure through decentralization and localization.
• Reduce the necessary upgrade cycle of computing, storage, memory, and bandwidth resources in the infrastructure.

Hybrid Cloud model is powerful today using traditional broadband networks, introducing 5G will increase efficiency. Customers will be able to stream, review and access video from the cloud much faster. While 5G does offer lower latency for video transfer, in reality, this claim does not hold up without leveraging edge computing. Thus, for streamlined video recording, analysis and storage, customers should consider using a hybrid cloud model where video and data move from endpoints to servers at the edge and lastly to the cloud.

This article ” Role of Edge Computing for 5G IoT Deployment” Architect and Methodology reference taken from MIMIK IEEE Communication.