Internship position 5G Software solutions

-
Machine Learning for Wireless Channel: In 5G radio network, channel is playing a central role in the algorithm design and system evaluation. Even if this has been studied extensively before for simpler cases, higher requirements have been posed to the 5G system to have a more accurate characterization for the wireless channel. The objective of this project is to apply machine learning to capture the complex structure in the channel, which can be further used to predict the behaviour of channel and hence improve network performance.

-
Machine Learning for Distributed Wireless: Extra-large-scale multi-antenna systems via collaboration among multiple antenna arrays distributed in different geographical locations is regarded as a promising evolution of today’s massive MIMO system for 5.5G and beyond radio networks. However, there are two major problems related to the scalability of the system. First, the large dimensionality makes the processing delay exceed the signal processing time budget. Second, the distributed nature of the baseband processing on different nodes requires minimal communication between sub-processes. Most of the conventional distributed schemes, such as ADMM, BCD, or message-passing algorithms, rely on a large number of iterations among different computing nodes in order to achieve a satisfactory performance, which is not affordable in practical systems. Our intention is to apply advanced machine learning techniques in designing more efficient distributed optimization algorithms to realize parallel and distributed beamforming/precoding/scheduling schemes with extremely low latency, which are also robust against channel state information errors caused by, for example, noise, pilot contamination, and user mobility.

-
HPC(High Performance Computing) for Wireless: The development of 5G and the next generation wireless standard requires more connections and traffic, and the adoption of machine learning in wireless puts forward higher requirements on the computing capacity for base station. On the other hand, with the failure of Moore's Law, it is difficult to further improve the computing capacity by advanced semiconductor manufacture. Therefore, the object is to study innovative HPC design methods to further improve computing efficiency for baseband chipset.

Qualifications

- Background in Computer Science, Wireless communication or equivalent;

- Excellent programming skills in Python/C++;

Following experience will give you a plus:

- Ability to quickly understand a new technology area;

- Hands-on experience on Machine learning;

- High performance computing (HPC) background, or experience in designing chipset specific acceleration algorithms;

Other skills and capabilities

- Excellent written and oral communication skills
- Strong problem-solving and analytical skills
- Excellent team player: cooperative and consultative behavior; ability to work independently and as a member of various teams.

Minimum education and experience requirements

- Bachelor degree or equivalent in Computer Science or Electrical Engineering is required.

What do we offer?

- A mentorship program with 5G experts
- Global vision, platforms, and resources
- A competitive salary and benefits package