Experimental Results for the AERO 5G project (Fed4FIRE+)

The corresponding results refer to the experiments conducted during the life of the Fed4FIRE+ project entitled: "AERO 5G (Augmented Reality Tour Guide Architecture for 5G)". The public availability of the results aim to help future experimenters and researchers to obtain some intuition with regard to the benefits of 5G for content-based, bandwitdh consuming, MAR applications.

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Short Description of the Experiments: All values have been rounded to two decimal places. Our team has conducted ten experimental runs for each of the following experimental scenarios.

o 4G SDR srsLTE-to-AWS: A 4G SDR srsLTE network deployed over the Fed4FIRE+ Iris testbed, between a Xiaomi Mi Mix 2S handset and an Amazon EC2 node at Amazon Cloud (AWS) that hosts the AR video content.

o 4G SDR srsLTE-to-IMEC: A 4G SDR srsLTE network deployed over the Fed4FIRE+ Iris testbed, between a Xiaomi Mi Mix 2S handset and a bare metal machine at Fed4FIRE+ IMEC's VirtualWall that hosts the AR video content.

o 4G SDR srsLTE-to-Iris-MEC: A 4G SDR srsLTE network deployed over the Fed4FIRE+ Iris testbed, between a Xiaomi Mi Mix 2S handset and a MEC storage node located at the edge of the network infrastructure at the Iris testbed that hosts the AR video content. Although MEC is considered to be a 5G technology, our aim in this scenario is to explore the benefit of deploying edge storage nodes in wireless mobile telecommunication technologies in general. For this purpose, we assume that the video content has been stored at the MEC storage node a priori to the end-user's requests.

o Commercial Three.ie 4G-to-AWS: A Commercial 4G network deployed by the Three.ie mobile operator in Ireland, between a Xiaomi Mi Mix 2S handset and an Amazon EC2 node that hosts the AR video content. Even though a MEC storage node could not be deployed in this scenario, our intention is to estimate the benefit of deploying edge storage nodes empirically by consulting the results concluded for the 4G LTE-to-MEC scenario.

o 2.4GHz Wi-Fi-to-IMEC: A 2.4GHz Wi-Fi (802.11 n) network deployed over the Fed4FIRE+ Iris testbed, between a Xiaomi Mi Mix 2S handset and a bare metal machine at Fed4FIRE+ IMEC's VirtualWall that hosts 4K AR video. This scenario serves us as a proof-of-concept that a 2.4GHz Wi-Fi network is not capable of supporting the delivery of high quality 4K AR content in areas where there are a lot of 2.4GHz Wi-Fi networks.

o 5GHz Wi-Fi-to-AWS: A 5GHz Wi-Fi (802.11 ac) network deployed over the Fed4FIRE+ Iris testbed, between a Xiaomi Mi Mix 2S handset and an Amazon EC2 node at Amazon Cloud (AWS) that hosts the AR video content.

o 5GHz Wi-Fi-to-IMEC: A 5GHz Wi-Fi (802.11 ac) network deployed over the Fed4FIRE+ Iris testbed, between a Xiaomi Mi Mix 2S handset and a bare metal machine at Fed4FIRE+ IMEC's VirtualWall that hosts the AR video content.

o 5GHz Wi-Fi-to-Iris-MEC: A 5GHz Wi-Fi (802.11 ac) network deployed over the Fed4FIRE+ Iris testbed, between a Xiaomi Mi Mix 2S handset and a MEC storage node located at the edge of the network infrastructure that hosts the AR video content. Similar to the 4G LTE-to-MEC scenario, we assume that the video content has been stored at the MEC storage node a priori to the end-user's requests.