5G edge computing enhanced mobile augmented reality

Augmented reality applications work better when the heavy computing is moved from the user's device over a 5G network to servers located nearby. The results of the doctoral research indicate that 5G edge computing can bring benefits such as reduced system latency, which improves the user experience.

Augmented reality combines the physical world around the user with virtual elements drawn on top of it. In mobile augmented reality, these elements are shown on devices that travel with the user, such as smartphones or head-mounted displays. For an application to offer content that fits the user's situation, it needs to know its surroundings. Information about the surroundings comes from the device's own sensors, such as the camera, or from external Internet of Things sensors. Processing this information, for example analysing camera images, can however be computationally demanding, and doing it on the device itself causes delay and consumes the device's resources. One solution is to move the augmented reality tasks from the device to nearby edge servers, which have significantly more computing power. When this is combined with a fast 5G connection, data moves between the device and the server even faster.

The thesis studied how 5G and edge computing can improve the quality of service of mobile augmented reality applications. Using two example applications, it examined how augmented reality tasks can be distributed and scaled across edge computing infrastructure by means of service orchestration. It also evaluated the transfer of image and video data from an augmented reality application to an edge server over 5G and other network connections, and made an initial investigation into how the buffer size of transport protocols affects performance.

Finally, the thesis considers the limitations and significance of the research, as well as possible topics for future work.