Master Thesis: Lightweight software containerization for the 5G edge
5G systems are being deployed in 2020. One of the advantages of the 5G technology will be lower latencies compared to previous 2/3/4G technologies for both users and devices. Devices become more useful if they can communicate in the lower latency range, typically 5-10 milliseconds. However, the distance of the devices to the computing resources, such as those found in cloud infrastructures, can be costly in terms of real-time performance. Edge Computing (EC) is an architectural approach to reducing the distance between endpoint devices and cloud computing resources. The core idea is to bring software that normally would be in a data center closer to the devices that utilise it. Edge computing implies not only narrowing the networking, but also the computation and associated storage. Therefore, this thesis will help bring computation closer to the devices.
The technical setting
Within cloud computing containerisation has become the operations norm. Lightweight virtual machines, such as Kubernetes, Docker, etc. are self-contained Linux systems that can support a number of services. Containerisation allows packaging, deployment, run/suspend, upgrading and migration to alternative compute locations. In an edge setting, where the Compute-Storage-Comms is remote, current cloud practices may not work due to resource constraints: 1-2U units, 100-1 sec latency, disconnected operations. However, the appearance of a single manageable cloud-like infrastructure would be very desirable.
Brief outline of work
- Assess the given software alternatives for Edge Compute
- Rank their worthiness with respect to EC
- Identify missing software components for the following two Edge uses:
- Vulnerable (cyclists & scooters) road users1
- 5G Robotics2
- Implement part of a 5G Edge demonstrator project (as part of a team)
What this thesis will not involve
It will not delve into microservices-orientated systems or Infrastructures as Code (IaaC), like Ansible, Terraform etc. It will not be very theoretical or involve large amounts of reading.
- Scope: 30 hp, start Autumn 2020.
- Supervisor: Ian Marsh, Ph.D.
- Placement: Kista, Stockholm.
- Skills: Excellent UNIX knowledge + good system coding skills, a must (pref. Rust, Go, C, Java).
- Academic advisor: TBD.
- Academic Examiner: TBD.
Welcome with your application
If you want to know more about this thesis project, please contact supervisor Ian Marsh, Ph.D, +46 10 228 43 59. The last day to apply is August 31st, 2020.
The State-of-the-Art in Container Technologies: Application, Orchestration and Security, E. Casalicchio and S. Iannucci, Jan. 2020, doi https://doi.org/10.1002/cpe.5668 , paper pdf .
- A KTH prototypes project between RISE and KTH.
- The ANIARA project includes joint work between RISE and Ericsson, Kista.
Our union representatives are: Linda Ikatti, Unionen, +46 10-516 51 61 and Ingemar Petermann, Sveriges Ingenjörer, +46 10-228 41 22.
Calls from external recruitment companies and sales persons are kindly declined as we rule under the Public Procurement Act.
Visstidsanställning 3-6 månader
Student - examensarbete/praktik
+46 10 228 43 59
2020-08-31Skicka in din ansökan