NOKIA is a global leader in the technologies that connect people and things. With a complete end-to-end portfolio, NOKIA is uniquely positioned to provide novel solutions to all sorts of customers, including mobile network operators, governments, and large enterprises.
Bell Labs is the innovation engine of NOKIA and a global research organization with sites in the US, Europe, and Asia.
The department of Radio Access Architecture in NOKIA Bell Labs, Munich, is offering the opportunity for a student to pursue a Ph.D. degree in communication engineering in collaboration with the Ilmenau University of Technology, Communications Research Laboratory headed by Prof. Dr. Martin Haardt. The thesis work will be conducted in Munich and is expected to be completed within 3 years.
Scope of the Ph.D. thesis
5G mobile radio system research and standardization is ongoing and in 3GPP the first so called new radio (NR) Release 15 standard is available. Further performance gains are expected from the combination of massive MIMO (mMIMO) with joint transmission cooperative multi point (JT CoMP). Related important enablers are efficient channel estimation and reporting of channel state information (CSI) for a high number of relevant channel components. Competing proposals for efficient CSI feedback exist, like either implicit codebook-based solutions or using explicit time domain reporting that rely on compressed sensing gains.
Explicit time domain CSI opens the door to channel prediction based on an estimation of the multipath component parameters comprising a channel impulse response (CIR). JT CoMP and the mentioned enablers are powerful, but complex and machine learning in combination with deep neural networks (DNN) looks promising. DNNs have become very popular, but the goal of this work would be to carefully evaluate the benefits and limits of DNNs for the intended PHY (physical) layer applications like the already mentioned channel prediction.
This PhD thesis includes the following topics.
- Integrating machine learning based on deep neural networks into the overall JT CoMP framework and evaluation of benefits, limitations and drawbacks
- Channel estimation, reporting and prediction for massive MIMO and JT CoMP
- System level analysis of massive MIMO in combination with JT CoMP
- Influencing 3GPP NR standardization to support the identified solutions
Qualification and skills
We are looking for a highly-motivated candidate who is enthusiastic about developing breakthrough innovations and participate in cutting-edge research in NOKIA Bell Labs.
A diploma or master degree in communications, information theory or computer science with focus on distributed systems with a grade of at least 2.0 (applicable in the German grade system) is required, as well as a good knowledge in communication networks, programming tools for performance evaluation (e.g., Matlab). Most importantly, the applicant should be willing and able to get engaged in novel disruptive research areas, which will require the applicant to learn new skills.
The applicant should be highly proficient in spoken and written English, and he/she should have the ability to conduct independent research while also collaborating in team in a multi-cultural environment.
Application and contact
Applications shall include: a cover letter, curriculum vitae, bachelor and master (or diploma) degrees, academic transcript of courses and grades, any other related documents such as certificates and reference letters.
The selection process will start on December 01st, 2018, and will finish as soon as a qualified applicant has been selected.
For more information about Bell Labs, visit www.bell-labs.com.