Article Preview
TopIntroduction
According to the IoT Analytics market study published in August 2020 (IoT Analytics Market Update, n.d.), one of the major factors driving the growth of the IoT Analytics industry is the increasing number of connected devices. By 2022, according to Ericson's accessibility report, 1.5 billion IoT systems will be wired to cellular networks.
According to Forbes (2018) IoT returns will double by 2025 compared to 2017 revenue, confirming the market's future promise. Even though trend remains high, the number of IoT implementations projected for 2020 is smaller than anticipated. According to Forbes, the biggest hurdles to IoT adoption are “security, alignment with existing technology, and unpredictable returns on investment.”
The broad heterogeneity (GSMA Network Slicing, 2018) of the IoT ecosystem is one of the most significant integration challenges that IoT technologies will face. The variety of use cases in which these technologies can be applied, the numerous types of devices that can be used, the various communications methods that can be used (especially wireless), and the number of protocols and IoT platforms all contribute to this heterogeneity. In order to satisfy end users' Quality of Service needs, interoperability and integration are crucial factors for these existing technologies, not just from the perspective of the IoT network, but also from the standpoint of managing data, resources, and processes applications across IoT, Edge, and Cloud platforms (QoS).
One crucial aspect of existing commercial IoT applications, given the broad variety of IoT usage cases, is the potential for latency introduced by their cloud-based main components. As a consequence, having the flexibility to run multiple IoT functions (e.g., gateways, databases, or analytics servers) in separate network locations is important depending on the requirements of the use case in question. An IoT gateway, for example, may be implemented at the edge domain to allow for rapid inspection and transformation of the high-volume data generated by IoT devices, as well as performing computational tasks near to the devices. Other cloud-based features, such as databases and IoT analytics servers, translate data from edge computing systems and IoT applications into formats that are easier to store and analyze. Any commercial platform now has the capability to operate certain components at the edge, but resource management remains essential. The stability provided by osmotic provisioning is important for taking advantage of the benefits of osmotic computing.
The elasticity offered by osmotic computing in scaling up and down network resources to leverage network resources across various IoT solutions or even other applications is a significant factor. Current IoT systems support multi-tenancy, allowing many users to use the same computing power, applications, or even facilities in a safe manner. Current systems, on the other hand, have a number of resource utilization limitations that must be overcome in order for different forms of use cases to coexist. As a result, osmotic computing facilitates seamlessly coordinated distributed networks across the entire network, allowing for the entire solution to be provisioned in real time.