5G Cloud-based NWDAF on AWS
Amazon Elasticsearch Service (Amazon ES) is a fully managed service that makes it easy for you to deploy, secure, and operate Elasticsearch in AWS at scale. It is a widely popular service and different customers integrate it in their applications for different search use cases.
Zen Networks developed a big data analysis platform for mobile networks. The platform takes root in 5G releases 15+ for Network Data Analysis Function (NWDAF) as well as general 5G guidelines for Cloud-native and Service Based Architectures. As network automation becomes more and more present, NWDAF plays a central role in mobile networks in order to provide data-driven optimizations.
In this post, we will also cover how Amazon Managed Streaming for Kafka played a key role in the architecture.
Overview of Amazon ES
Amazon ES makes it easy to deploy, operate, and scale Elasticsearch for log analytics, application monitoring, interactive search, and more. It is a fully managed service that delivers the easy-to-use APIs and real-time capabilities of Elasticsearch along with the availability, scalability, and security required by real-world applications. It offers built-in integrations with other AWS services, including Amazon Kinesis, AWS Lambda, and Amazon CloudWatch, and third-party tools like Logstash and Kibana, so you can go from raw data to actionable insights quickly.
Amazon ES also has the following benefits:
Fully managed : Launch production-ready clusters in minutes. No more patching, versioning, and backups.
Access to all data : Capture, retain, correlate, and analyse your data all in one place.
Scalable : Resize your cluster with a few clicks or a single API call.
Secure : Deploy into your VPC and restrict access using security groups and AWS Identity and Access Management (IAM) policies.
Highly available : Replicate across Availability Zones, with monitoring and automated self-healing.
Cost-effective : Deploy automatically Elasticsearch without need for a team to manage it and resize it on demand as per your usage.
Overview of Amazon MSK
Amazon MSK is a fully managed service that makes it easy for you to build and run applications that use Apache Kafka to process streaming data. Apache Kafka is an open-source platform for building real-time streaming data pipelines and applications. With Amazon MSK, you can use native Apache Kafka APIs to populate data lakes, stream changes to and from databases, and power machine learning and analytics applications.
Overview of Zen Networks
Zen Networks offers innovative solutions for Communications Service Providers (MNO & MVNO) by bridging cloud and big data technologies with mobile networks expertise.
Our mission is to provide full 360-degree visibility to mobile networks with open technologies, leveraged to build observability solutions as well as value-added services on top.
We also bridge these technologies with our SIM Over-The-Air server and other steering services to build new capabilities and optimize OPEX.
How Zen Networks Innovates for the telecommunications Industry using AWS
5G NWDAF Solution
Release-15 and Release-16 define the framework for data analytics in 5G by introducing the Network Data Analytics Function (NWDAF). See figure below. This entity is key to better network automation using AI and ML capabilities powered by extensive network data events. In fact, it is a central part in zero-touch network management.
The NWDAF usages are very broad and cover (non-exhaustive list):
Customized mobility management per mobility pattern.
5G QoS enhancement.
Dynamic traffic steering based on UE service usage .
NWDAF consumes data from different NF and AF sources to analyse it then provide it to AF, 5GC NF and OAM. The NWDAF answers use cases in different domains such as QoS, steering, security and dimensioning. At the same time, the ingested data mixes between a wide range of sources.
Using AWS managed services, we were able to spin up an NWDAF bringing high value to the CSP. Below is a simplified architecture of the build.
Telco-centric data analysis
Network events and EDRs flow through enrichment services to give them more context from BSS and other dimensioning data before being centralized in AWS Elasticsearch Service.
Using this method, we build a highly reliable and fast data platform that can be used for real-time analysis. Below some of the key use cases:
Troubleshooting and support: Enriched and correlated data empowered by an efficient query language are key to drill-down and find out abnormal patterns related a specific customer or SIM card usage. These capabilities reduce by a lot the MTTR and enhance by far customer experience
Network Operations Center and monitoring: NOC requires synthetic and real-time dashboards to monitor the health of the network. By enriching streamed data, dashboards become more meaningful. Also, by using AWS Elasticsearch Service, we can drill-down to the actual impacting network events to find out the incident pattern.
Market analysis: Marketing decisions should be data-driven. The platform permits advanced data analysis queries to take product and marketing decisions as well as to evaluate the impact of product choices.
Security and signalling optimization: Over-signalling is a long-lived issue that requires continuous improvements to reduce the amount of useless signalling. In fact, the latter can be quite costly in a roaming environment. AWS Elasticsearch Service shows abnormal behaviour and overly verbose modules that can be targeted for optimization using statistical algorithms.
Data-aware real-time services
AWS MSK service is key to build real-time services. In fact, by combining it with AWS serverless technologies like Lambda or Fargate, network automation becomes a quickly grasped reality. In fact, network steering and automated provisioning decisions can be automatically taken by ingesting network events in real-time and building AI-based or explicitly defined rules.
Using AWS serverless technologies, we build highly reliable telecom workloads and have them scale on-demand using AWS auto-scaling mechanisms. In fact, at Zen Networks, we have found AWS cloud offering to be very compatible with the current 5G Service Based Architecture trends.
The platform helped us answer key use cases and while new opportunities show up to leverage it better, we already prepare for the next steps using it. Some on them are:
Leveraging AI/ML capabilities for better aberrant behaviour detection. For this, we are benchmarking AWS EMR (SparkML) and AWS Sagemaker
Add newer integrations towards NFV for enhanced automation
In this post, we explained how AWS Elasticsearch Service helped us bring data analysis capabilities to a mobile network with low effort. In fact, managing an Elasticsearch cluster can be daunting when done on premise. The same goes for Kafka where AWS MSK was a game-changer.
AWS services allowed us to focus on the business and development parts instead of the underlying infrastructure. This proved priceless and ended up with a very low time-to-market for the services we provided to the CSP.