The Patent Pathway from Research to Reality

I graduated as a physicist and did my PhD in electrical engineering. Around my master’s thesis, I developed an interest in optical communication and related technologies, for example, lasers. Optics is the technology that enables the realization and deployment of the largest capacity parts of global telecommunication systems. Here, we are talking about potentially several terabits every second over a single optical fiber. At the same time, optical technologies operating with single photons allow unbeatable security of communication systems. This topic is strongly related to the 2022 Nobel Prize in physics, indicating its importance. 

But I might be yarning too much, so following my PhD this whole topic was so interesting for me that I spent a couple of years in various academic postdoc positions. Then I joined Sigma in 2021 where I had the opportunity to dig deep and extend my knowledge towards mobile network architectures and work on problems with relevance for the telecommunication industry. 

In academic research, it was an everyday task and actually an obligation to regularly publish our results in scientific journals and conferences. I personally really like this activity and find it motivating, this way we stay on track of finding novelties and elaborating on ideas and methods within a field. It also generates opportunities to discuss our work with peers. 

On my new customer project in Sigma, we quickly and naturally found each other with colleagues at the customer site who work on parallel, sometimes more research-focused subtasks of the same topic, or on related topics. Then, it gradually became one of my main activities to elaborate on novelties and publish them, which was done in addition to meeting the short-term project goals of the customer. The format and focus changed, however: in the academy, I wrote research papers, and now I have switched to patents. 

It is important to note that a supportive environment, leaders, and encouraging technical and frontier discussions in Sigma and at the customer site were all essential for this activity. 

A patent is the protection of an intellectual property which excludes others from commercial usage of the invention for a specified period of time. The patented property can be commercialized by others, however, with the consent of and royalty paid to the patent owner. 

Any field of technology can be patented. This can be a product, for example a chemical, or a process itself to produce a product. Further potential examples are processes used by software or hardware, such as a novel process or flow anywhere in a telecommunication network, should it be in the core, in Radio Access Network, or over an interface. This is of special importance for Sigma. 

Please note that although pure mathematical methods and formulas on their own cannot be protected, their application to solve a specific use case is potentially patentable.

Sure, they cover various interesting topics. One of them is related to quantum key distribution, a technology potentially offering unconditionally secure communication between parties. Maintaining precise temporal synchronization of the communicating terminals is of vital importance for these systems. In this patent we investigate a specific quantum key distribution architecture and propose an automated adaptation of the system synchronization to varying environmental and weather conditions. So, the proposed solution facilitates practical implementation of a specific quantum key distribution system in the future. 

Two more filed patents are related to AI use cases in 5G mobile communication systems. The field is under intensive development and practical issues regularly arise where AI-based algorithms offer potential solutions. 

One of these issues is the large number of network exposure APIs introduced by various vendors and organizations, the functions and parameter lists of which usually overlap. The excessive number of these APIs and their overlaps potentially make it difficult to manually find optimal API sets to set up a complex network configuration. 

To mitigate this, we proposed an approach based on reinforcement learning that finds the optimal set of the APIs to obtain a pre-defined network configuration, according to various key performance indicators monitored, improving network configurability. 

Another field where we found AI approaches beneficial was the automation of 3GPP Vertical Application Layer Group configuration. Member devices of a group communicate privately within a 5G system, and the configuration of these groups has been done manually to date. Here, we proposed an algorithm that consumes historical Group configuration data as well as various historical and predicted network performance metrics, for example, packet delay or packet loss. It is based on an interesting novel concept called Convolutional Graph Neural Network with Generative Adversarial Networks. It generates the policies necessary to find the optimal Group configuration in a 5G system. 

At a personal level my main advice is keep focusing on collaboration. Research and innovation are not a one man show and good things can come from frontier areas. Listening to novelties in the market and in the literature is crucial, similar to any kind of research activity. 

I found it also important in industrial innovation to link your work with ongoing product development. This can be a collaboration with one of the development units. 

Take care of the patenting process from the beginning to the end. In case there is a review process within your organization for potentially patentable material, you should take special care of that. Reviewers are different, properly addressing their concerns and properly communicating with them might mean the difference between a filed patent and wasted weeks or months of work. 

At managerial and organizational level, I believe a supporting environment is crucial and a prerequisite of any innovation, together with smooth communication within the organization. I also noticed that sometimes there is a tendency that only distilled and highly task-specific information is forwarded to the engineer or developer during a project. This might be effective in terms of project completion but will probably not result in many inventive ideas. 

I do hope that, as there are still ideas in the pipeline in the above-mentioned areas of mobile and secure communication. Discussing and working on one aspect of a topic might trigger other aspects we can think about. There is no shortage of ideas, time must be allocated to them!