Directing network traffic with telecommunications policy

Picture this: You’re on a crowded New York City street, crowded with Ubers — yes, yellow cabs — in front of you, cyclists weaving in and out of the crack, and pizza delivery men cursing because they promise free delivery unless it arrives in 30 minutes.

Suddenly, an ambulance approached the crowd with a frantic alarm. Instantly, the traffic slowed down, and the crowd moved spontaneously to both sides, opening up a road. This magic scene only happened because of policy: ambulances have the right of way.

If some transit systems do not know right-of-way rules, they will not be able to make decisions to determine when to slow down and turn, which will cause confusion for other cars and pedestrians, prevent ambulances from moving smoothly, and potentially bring traffic to a standstill.

The same goes for telecoms. A network policy is a set of rules (such as the right of way) that govern the behavior of network devices. Policies use predefined rules and real-time network data to allocate limited resources, such as bandwidth, that determine how traffic flows through the network. Like traffic lights, the telecommunications network policy is also used to restrict limited access to network capacity, and we should not break it.

5G carriers, like 4G carriers, provide services to individuals and businesses. However, 5G is more complex because it is expected to be extremely fast and reliable with millisecond latency, and it provides broadband mobile services. As operators seize these opportunities to develop new services and IoT use cases, and as network slices proliferate, telecommunications policy will need to evolve in line with 5G needs and user expectations.

The digital transformation requires a data platform

5G networks have huge capacity, and network slicing provides an end-to-end method of segmenting the network for specific use cases (such as pedestrians and sidewalks) and the capacity requirements of those use cases.

However, the capacity requirement is only one aspect, we also need to provide quality of service assurance and delay assurance. And these requirements will vary from use case to use scenario, which is why modern telecommunications companies need dynamic policy control, rules and conditions will change over time, or applications will need to adapt and adjust according to the actual scenario.

The most critical factor in making these policy decisions is the extremely low latency. As a result, we are faced with the conundrum that these systems need to make complex decisions, and they need to make them in a very low latency manner. Decisions need to take into account all of these aspects:

  • Data: account balance data, rating data, SLA data, network data;
  • Business logic: QoS based on rating, QoS based on network data management, slice-based policies, etc.
  • Machine learning on network data.

Typically, 4G and older systems treat each element as a separate technology. In the past, data was accessed from the data storage unit and then processed in the policy control function module. This is a very expensive process based on the following perspectives:

  • Too long latency: it takes more time to transfer data over the network;
  • Network: moving data over the network rather than real-time processing, where the database artificially limits the number of policy calls for a given infrastructure;
  • Infrastructure: So, while keeping latency low, to ensure scalability might mean adding more infrastructure, which might be great for software and hardware vendors, but not so easy for telecom carriers.

Digital networks such as 5G require a smooth relationship between the network functional layer and the data layer. Modern policy systems rely on a data platform that shares business logic with network functionality while allowing business rules to be modified en route. By tightly integrating data and data-driven decisions, operators can achieve predictable scalability while maintaining extremely low latency. Only modern telecom data platforms can support real-time data processing and decision-making within the 5G latency and 10-millisecond window required by modern telecom policies.

Modern data platform — born for modern telecommunication service

The digital transformation heralds a new era of real-time, event-driven, data-driven intelligence that will require processing complex decisions to automate operational optimizations at the speed, scale, and precision required by 5G.

Operators need a supercombustible data platform that can operate quickly and seamlessly throughout the data life cycle. This enables real-time decisions to prioritize high-performance applications while effectively and efficiently managing network traffic. This platform can strip away the redundant stack layer and support advanced AI tools and distributed computing, which will be our main business direction.

VoltDB can help telecom companies and communication service providers run their digital and 5G businesses using real-time data. Real-time analytics and instant control of stream data to manage networks and applications to gain more opportunities to interact with new customers and increase revenue through 5G monetization. We have verified and tested in the telecom space of BSS, and our database has the best architecture for BSS in the 5G era.

Do you like VoltDB? Welcome private message, and more friends to discuss.

VoltDB supports strong ACID and real-time intelligent decision making applications to enable a connected world. There is no other database product like VoltDB, which can fuel applications that require a combination of low latency, large scale, high concurrency, and accuracy. VoltDB was created by 2014 Tering Award winner Dr Mike Stonebraker, who has redesigned the relational database to meet today’s growing challenges of real-time operations and machine learning. Dr. Stonebraker has been studying database technology for more than 40 years and has brought many innovative ideas in the areas of fast data, streaming data and in-memory databases. During the development of VoltDB, he realized the full potential of using in-memory transaction database technology to tap into streaming data, not only to meet the latency and concurrency requirements of processing data, but also to provide real-time analysis and decision making. VoltDB is a reliable name in the industry and has been used in practical scenarios in cooperation with leading organizations such as Nokia, Financial Times, Mitsubishi Electric, HPE, Barclays, Huawei, etc. *