High speed, high reliability and low latency are the key benefits that Telcos expect from 5G. While high speed (targeted at 20 Gbps) helps to upload and download video-based content faster and in larger volumes, high reliability (always 100%) supports mission-critical services especially in the IoT world. Low latency (sub-millisecond) cuts down the time to deliver mission-critical applications. Never before has Service Quality Management (SQM) in mobile networks been as important as now. This is because IoT has opened up a new world of differentiated services that offer new business opportunities for the IoT players as well as the Communications Service Providers(CSPs).
How is 5G helping with this?
While 5G offers a recourse to the unprecedented use of video services, fuelled by social media, Virtual Reality (VR) and widespread availability of smartphones, it is the sudden growth of IoT devices flooding the global market that helps the most.
IoT is revolutionizing the human lifestyle, with billions of hyper-connected devices communicating with humans and with each other. With the increase in wearable technology, motion-based sensors, voice command and eye movement sensors, 5G use cases are being driven by low latency and high-reliability requirements of such IoT devices.
5G offers a unique capability to cut the RAN and Core networks into suitable slices so that they can support different IoT services according to an associated SLA. Never before have such differentiated services been imposed upon the mobile network. Let’s take a few examples of what the service categories supported by 5G networks are:
- Enhanced mobile broadband (eMBB) that delivers gigabytes of bandwidth on demand for UHD video, VR, etc.
- Massive machine-type communication (mMTC) that connects billions of sensors and machines to the tune of 1 million per square kilometer to support smart cities
- Ultra-reliable, low latency communication (uRLLC) that allows immediate feedback with high reliability and enables real-time remote control over robots in manufacturing and autonomous driving
5G slices and associated complexity
5G network operators need to address the needs of new customer groups: those that deliver, amongst other services, hyper-sensored automotives, connected factories, distant healthcare and smart cities. The consumers of digital network services have high expectations of quality and are pushing for stringent SLAs and policy control. Slices of the 5G network are required to be dedicated to such customers along with cloud-based automation techniques to deliver on the demands of speed and reliability. The slices then are assigned QoS capabilities to suit a dynamically allocated IoT service. So, irrespective of who the IoT customer is (connected factory, smart energy, smart home, etc.) or where the mobile IoT customer goes (connected car, fleet management, etc.), the network slice stays dedicated to it.
The challenge lies in managing the complexity associated with creation and management of the slices. For example, mapping an IoT service to the most suitable network slice or mapping a bundle of IoT services to a network slice. IoT users can also be assigned a primary slice and a secondary slice for backup. In some cases, there can be the existence of sub-slices within a network slice. There is also a possibility that network slices can be leased to other mobile networks.
As the service and the required QoS characteristics are assigned to a slice or sub-slice as per the above configurations, it needs to be continuously monitored for the offered QoS/SLA. And there needs to be a closed loop system to orchestrate the network slice in real time.
Role of Service Quality Management (SQM) in managing the slices
By virtue of its design, SQM incorporates the capability of managing a service and orchestrating changes based on QoS policies/SLAs. This is critical to the constant negotiation of the network slice to support the service supported. In certain cases, a new network slice would need to be configured on the fly.
In such a high-demand dynamic service environment of 5G, an SQM especially designed for digital/IoT services will enable the CSPs to provide the following:
- Service monitoring: IoT networks are currently being overlaid on existing communication networks (and soon 5G) and therefore require CSPs to manage service degradations, root cause analysis and resolution to be carried out constantly. An SQM can help with:
- IoT services managed as per industry vertical
- Network/slice analytics and associated service analytics
- Network/slice capacity and congestion KPIs
- Real-time network/slice KPIs: As IoT data traverses 5G networks, it needs to be monitored in real-time (down to a few ms of delay). Some key network/slice KPIs would be capacity, latency, jitter, throughput and bandwidth allocation. As latency can range from 1-10 ms for different IoT services, jitter from 10-100 microseconds, bandwidth from 100 Kbps (small sensored devices) to several hundred Mbps (robotic cameras), varying levels of these metrics per slice means constant manoeuvring for scaling-in, scaling out of slice resources
- Assuring reliability: The acceptable level of service assurance for most time-critical services is 100% and many operator processes need proactive SQM with closed-loop automation to reduce delays and increase network reliability
- Predictive maintenance: In order to protect the IoT service uptime, the SQM offers predictive and pre-emptive maintenance for services prone to failure. For example, 5G network slices can carry out dynamic route optimization especially for services that are mobile (cars, drones, etc.). In order to predict problems accurately and then follow up with pre-emptive maintenance, use of analytics-led SQM will be key
- Analysis and resolution: Root cause analysis for identification of service issues and use of dynamic orchestration to resolve issues to improve the latency, jitter, bandwidth and other critical characteristics of the network slice, including clearing of alarms
- Service dashboards: For end-to-end monitoring for each industry vertical and for each end user, the 5G CSP can provide aggregated service performance dashboards and visualization/drill-down to individual network slices
The IoT service providers and the 5G CSPs understand that mission-critical and time-critical IoT services are highly dependent on the success of the 5G network slicing concept. It requires renewed effort on the 5G operator’s part to analyze and action performance data quickly using predictive models and automated root cause analysis to extract the most from the 5G slices. The 5G CSPs’ promise to the IoT service provider will be realized when CSPs slice their networks and tie them to IoT services. And when they introduce new techniques of listening and reacting to the network slice and the associated service, using automation, orchestration and policy-driven Service Quality Management.
Note: This is an edited version of the original MYCOM OSI article first published in IoT Now.
