Performance Analysis of Slicing on a 10-node 5G Architecture for Networked Music Performances

Networked Music Performances (NMPs), where geographically displaced musicians play together over a data network, represent a challenging application for today's wireless communications. This is due to the stringent constraints on latency, throughput, and reliability that need to be obeyed in order to achieve a satisfactory quality of experience for the musicians. Slicing is a promising feature of 5G networks in the context of NMP applications, as it makes it possible to isolate the networking and computing resources allocated to NMP devices. However, the use of slicing has been scarcely investigated for the NMP context so far. Moreover, previous works focusing on NMPs over 5G involved up to 4 nodes. To bridge these gaps, we study 5G performance in support of NMPs involving an architecture with 10 nodes, both with and without slicing. Specifically, we focused on the assessment of the sole wireless link, as the measurements can be easily transferred to a realistic NMP architecture involv...

Networked Music Performances (NMPs), where geographically displaced musicians play together over a data network, represent a challenging application for today's wireless communications. This is due to the stringent constraints on latency, throughput, and reliability that need to be obeyed in order to achieve a satisfactory quality of experience for the musicians. Slicing is a promising feature of 5G networks in the context of NMP applications, as it makes it possible to isolate the networking and computing resources allocated to NMP devices. However, the use of slicing has been scarcely investigated for the NMP context so far. Moreover, previous works focusing on NMPs over 5G involved up to 4 nodes. To bridge these gaps, we study 5G performance in support of NMPs involving an architecture with 10 nodes, both with and without slicing. Specifically, we focused on the assessment of the sole wireless link, as the measurements can be easily transferred to a realistic NMP architecture involving a wide area network. Our results show that, in the slicing condition, latency slightly increased due to the realistically different computing specifications of the MEC servers compared to those of the Core Network servers, whereas reliability slightly improved as expected.