Master Thesis: Brain-Inspired Algorithms for Radio Access Networks

Stack

Responsibilities

• ▹Review brain-inspired algorithms for RAN workloads
• ▹Select a representative RAN use case, research question and evaluation metrics
• ▹Prepare data and experimental scenarios suitable for the selected capabilities
• ▹Implement conventional baseline methods and obtain reference KPIs
• ▹Develop and validate a brain-inspired algorithm, focusing on a capability such as robust inference, low-latency processing, multi-timescale temporal processing, sparse representations or adaptation
• ▹Evaluate the algorithm against the baselines using the defined metrics
• ▹Assess hardware suitability by analysing operation count, memory requirements, sparsity and parallelism
• ▹Document the solution and evaluation results

Requirements

• ▹Pursuing a Master's degree in Machine Learning, Mathematics, Engineering Physics, Computer Science, Embedded Systems or a related field
• ▹Strong foundation in probability theory, deep neural networks, spiking neural networks or coupled dynamical systems
• ▹Understanding of neuromorphic computing hardware such as Loihi 2 or SpiNNaker 2, and software simulation frameworks
• ▹Good programming skills and knowledge of C++, Python and Linux
• ▹Strong analytical and problem-solving skills
• ▹Good technical writing and communication skills

Nice to have

• ▹Experience with brain-inspired algorithms or neuromorphic computing
• ▹Familiarity with RAN workloads, wireless communication systems or signal processing
• ▹Experience with experimental evaluation, KPI analysis or hardware-oriented modelling

Soft skills

About the company

Ericsson is a team of diverse innovators and an equal opportunity employer that welcomes applicants from all backgrounds.