Contact Persons
Gunnar Bark +46730435104
gunnar.bark@ericsson.com
Master Thesis – Ultra Reliable Low Latency Communication in 5G
Background
Ericsson, and other mobile network vendors, are now standardizing 5G in the global standardization body 3GPP. 5G should provide in the order of 10 times better performance in terms of user bit rates, end-to-end delays, system capacity, energy performance etc. With respect to the delays, we have aggressive 5G targets on below 1 ms delay in the Radio Network part of the communication chain between terminal and server.
The applications for this low latency communication could be controlling robots or Intelligent Traffic Systems applications. In both these scenarios, not only the delay needs to be short, the reliability of the signaling needs to be high as well. The control signaling needs to reach the robot, or the car, within 1 ms with a probability of more than 99,99%. This is denoted Ultra Reliable Low Latency Communication (URLLC) within 5G.
However, there is a natural contradiction, or balance, between being fast and reliable in wireless communication, where multi-path fading and interference are inherent in the radio channel. Radio transmission and reception diversity over the frequency, code and spatial (i.e. antenna) dimensions are all beneficial to solve this challenge.
Thesis Description
This thesis work will propose and evaluate algorithms and signaling for Ultra Reliable Low Latency Communication in 5G.
More specifically, this thesis work will focus on scheduling and link adaptation for URLLC uplink (meaning the radio link from the terminal to the base station). Scheduling is the function where the base station selects which terminals that are allowed to transmit in next transmission slots. Link adaptation is the function that selects modulation and coding for the transmission. If scheduling and radio link adaption is too conservative the capacity in terms of the load the system can handle, may be too low, while if, scheduling and link adaptation is too aggressive the robustness requirements cannot be met.
The so-called Hybrid Automatic Repeat Request (HARQ) protocol will play a central role in the problem. In a standard HARQ method, error detection bits are added to the data that is transmitted by the mobile. The base station can detect if the transmission was incorrectly decoded and tell/schedule the mobile to make a re-transmission of the data. This means that URLLC scheduling and link adaptation can target a far higher BLER than e.g. 0.01% for first transmission since together with re-transmissions the target reliability e.g. 99.99% can be met for the data.
Scheduling mechanisms and improvements to the HARQ protocol for URLLC is currently standardized by 3GPP and algortihms for how to best utilize them will be studied in the thesis.
Evaluations of the proposed methods will be performed in Java-based radio network simulators and the results from those will inspire to further refine the proposed algorithms together with the advisors
Contact Persons
Gunnar Bark +46730435104
gunnar.bark@ericsson.com
Qualifications
This project aims at Master of Science students in mathematics/statistics, computer science or electrical engineering.
Java and Matlab are our primary programming languages for simulation so a background in both of these is preferred. A successful candidate has a solid knowledge in radio communication and has an average grade above B/4.0.
Extent
This position is for one student. Scope is for 30 university credits (Swedish högskolepoäng)
Location
Ericsson AB Mjärdevi, Linköping
Preferred Starting Date
January 2018
