Loading...

MSc Thesis - ML for Enhancing URLLC DRL scheduler

Posted date: Nov 10, 2020

Location: Stockholm, AB, SE

Company: Ericsson

Our Exciting Opportunity


Ericsson Research


Ericsson Research is the central research organization within the Ericsson group. We are about 600 researchers in many countries with the central site in Kista, Stockholm. Our mission is to provide Ericsson with system concepts, technology, and methodology, to secure long-term competitive product provisioning. World-class innovation is achieved through cooperation within Ericsson and with partners, customers, universities, and research institutes.
For more information, see https://www.ericsson.com/en/tech-innovation/research.


Background


The fourth industry revolution (Industry 4.0) depends on the incorporation of cyber-physical systems, Internet of Things, cloud computing, and cognitive computing. The main pillar of realizing full industrial automation is the wireless connectivity among industrial robots and nodes. Industrial automation, including factory automation, is considered the main driver for time-sensitive networks (TSN). One of the main requirements for industrial automation is high reliability (e.g., 10 years without application fail down) and availability (e.g., 99.9999%). Such requirements mean that an error is quite rare to occur in the real deployment of the system. That is, from a simulation perspective, we need to run the ultra-reliable low latency communication (URLLC) simulator for a long time and exhaust a large pool of resources to trigger and understand such rare failure events. Optimizing a URLLC scheduler will be difficult due to not having enough rare event samples.
The recent development of reinforcement learning enabled the deployment of RL agents as a communication scheduler. However, such an application of RL on a rare event environment (URLLC like) might not result in a logical performance due to the under-representation of the rare event samples. Teaching a real scheduler using a simulated experience (in a feedback loop) is one way to enhance such RL scheduler. However, we will still face the issue of huge simulation time requires to generate a rare event. Several works aimed at speeding up simulators of such rare event, via utilizing importance sampling and sequential Monte Carlo methods. Other methods aim at speeding up rare event simulators by synthetization of a rare sample. The process of synthetization of samples could take several directions, one example is to focus only on the rare samples, another is to focus on segmenting the distribution.


Purpose


The aim of this thesis is to utilize GANs (Generative Adversarial Networks)  to improve the training of DRL (as a wireless communication scheduler).


You will


•    Study GANs’ application to improve DRL scheduler
•    Investigate KPIs for the evaluation of trained GANs
•    Run an example of wireless communication protocol, produce data out of the simulator
•    Train a single or multiple GANs per simulation distribution
•    Evaluate the performance of multiple GANs against a real simulator
•    Train DRL using the real simulator (R-DRL), and train DRL using the segmented GANs models (GAN-DRL). Then, evaluate R-DL against GAN-DRL (optional but recommended)

•    Report to the Research Manager


Required background


•    ML & GANs 
•    Statistical Simulation
•    Wireless networks or equivalent (optional).
•    Good background in Python and Java
•    Course requirements which must have been fulfilled for the degree project
•    Good spoken and written English
•    Ability to perform studies and reports
•    Ability to do innovation
•    Being self-driven and able to initiate action
•    Ability to work independently, as well as collaborating in teams
•    Good presentation and communication skills


Application


Applications should include a brief personal statement, CV, and a list of grades. In the application, make sure to mention previous activities or other projects that you consider relevant for the position. Suitable applicants will be interviewed as applications are received.


What´s in it for you?


Here at Ericsson, our culture is built on over a century of courageous decisions. With us, you will no longer be dreaming of what the future holds – you will be redefining it. You won’t develop for the status quo, but will build what replaces it. Joining us is a way to move your career in any direction you want; with hundreds of career opportunities in locations all over the world, in a place where co-creation and collaboration are embedded into the walls. You will find yourself in a speak-up environment where empathy and humanness serve as cornerstones for how we work, and where work-life balance is a priority. Welcome to an inclusive, global company where your opportunity to make an impact is endless.


What happens once you apply?


To prepare yourself for next steps, please explore here: https://www.ericsson.com/en/careers/job-opportunities/hiring-process

 

Kindly note that we cannot process applications sent via email.

 

Please submit your application in English.

 

Location for this role: Stockholm, Sweden

Supervisor: Abdulrahman Alabbasi, abdulrahman.alabbasi@ericsson.com

Recruiter:  Dorota Baran, dorota.baran@ericsson.com

Last day to apply: 2020/11/25

 

Curious to know more about the life at Ericsson? Meet some of your future colleagues and watch our People film.

 

Do you believe that an organization fostering an environment of cooperation and collaboration to execute with speed creates better business value? Do you value a culture of humanness, where fact based decisions are important and our people are encouraged to speak up? Do you believe that diverse, inclusive teams drive performance and innovation? At Ericsson, we do. 

We provide equal employment opportunities without regard to race, color, gender, sexual orientation, transgender status, gender identity and/or expression, marital status, pregnancy, parental status, religion, political opinion, nationality, ethnic background, social origin, social status, indigenous status, disability, age, union membership or employee representation and any other characteristic protected by local law or Ericsson’s Code of Business Ethics.

 

Primary country and city: Sweden (SE) || || Stockholm || Stud&YP


Job Segment: Network, Scheduler, Telecom, Telecommunications, Technology, Administrative, Customer Service