How can you use network slicing to provide custom services for IoT and M2M customers?

Machine to Machine (M2M) was an acronym of choice for describing networks where there are no human end users. For example a network of connected vending machines or electrically submersible pumps. However M2M as a concept fails to cover the full span of system requirements. Today in a digitized and digitalized infrastructure there is a combination of connected equipment and people, including compute, storage and network elements for application processing. Both M2M and IoT are outdated terms and should be replaced with another label. We refer to digitization (linking physical activities with their digital representation) or digitalization (powering a physical process through data).

IoT is like a framework which includes integration of sensors, actuators and communication technologies which allows "things" to communicate and share information which would be helpful in creating an interconnected and a responsive environment. M2M represents the automated exchange of the information between devices or machines. This communication between machines provides real time data sharing and coordination which in turn increases the efficiency in various applications.

Machine to Machine (M2M) e Internet das Coisas (IoT) se distinguem na forma em que obtêm a conectividade, no que podem conectar, na escalabilidade que oferecem e na maneira como os dados são utilizados. Ambas, no entanto, têm o objetivo de levar a indústria a um patamar mais conectado e inteligente.

For most practitioners IoT as a concept infers the following: - Communication is from a small embedded device with limited functionality for interacting with the physical world (sensing & actuating). - Communication is to a centralized application platform, where data from things is presented in the form of a dashboard. Further data processing is performed in the platform. - Communication is over the public Internet using NAT friendly protocols. Devices always initiate the connection to a centrally hosted service. This concept is outdated and has resulted in poor system architecture choices. In a modern digitized infrastructure, processing, storage and networking are much more distributed and layered.

Simply put, they don't. Network slicing is like a very large hammer that makes anything look like a nail. Network slicing from a telecom operator point of view is where radio spectrum, channels and capacity is dedicated to an application. In most industrial and utility use cases this is overkill. An alternative to network slicing inside the telco network is to build "closed user groups" with private APNs and VPNs to segregate traffic. QoS parameters for packet networking in the telco network can also be applied if there is a requirement to prioritize certain protocols. However we need to remember that prioritizing one protocol will require deprioritizing another and this approach is subpar compared to overprovisioning capacity.

To understand the benefits, let’s look at two very recent network slicing examples: Couple of days ago, Verizon showcased the ability to maintain performance for mission-critical functions while transmitting video data over a network slice in a commercial 5G environment. This highlights the potential of network slicing to enhance situational awareness for public safety personnel. Last month, Deutsche Telekom demonstrated the application of 5G network slicing to bolster safety and security in smart cities. Utilizing their Standalone 5G Network, they achieved on-demand, mission-critical surveillance video streaming in a congested urban setting. This underscores the promise of network slicing for optimizing urban infrastructure and security.