| Popis: |
This document, 5G-CLARITY D3.2, aims to provide evaluation results and refinements on the initially designed 5G-CLARITY user and control plane architecture that is introduced in 5G-CLARITY D3.1 [1]. This document is also aligned with the "network function and application stratum" that covers not only user- and controlplane but also application plane functionality as presented in 5G-CLARITY D2.2 [2]. In essence, 5G-CLARITY D3.2 provides the performance evaluations and refinements for: Multi-WAT aggregation: Including 5GNR CU/DU/RU integration, integration of Wi-Fi and LiFi networks as a single non-3GPP network, integration of 3GPP and non-3GPP wireless access technologies (WATs) and assignment of traffic flows via MPTCP; 5G-CLARITY eAT3S framework: Including operational flows, initial enhanced access traffic steering, switching and splitting (enhanced AT3S / eAT3S) algorithm design and control plane aspects of the custom MPTCP scheduler; Scheduling and resource management: Including Wi-Fi and LiFi airtime-based schedulers and utilitybased scheduler to manage different service types; Positioning: Including WAT-specific positioning scheme and its performance evaluations, as well as the fusion approach; Integrated 5G/Wi-Fi/LiFi network performance evaluation: Including possible access point (AP)/gNB deployment options, achievable communication bandwidths, technology-specific areacapacity achievements and integrated network area-capacity performance. Details for the 5G-CLARITY multi-connectivity framework evaluation are presented in Section 2. The 5GCLARITY multi-connectivity design includes, i) the multi-WAT aggregation, integrating 3GPP (5GNR) and non3GPP (Wi-Fi and LiFi) access networks, and ii) an enhancement on the AT3S scheme to improve the (multiaccess based) multi-connectivity functionalities. The details of design and validation of these features for the 5G-CLARITY user- and control-plane are provided. Section 3 delivers discussions on AP level and service level (traffic routing) resource scheduling techniques. Primarily, the corresponding telemetry and performance measurements are used to route the traffic across 3GPP/non-3GPP networks in near real-time (near-RT) using 5G-CLARITY eAT3S introduced to ensure qualityof-service (QoS), and as a following step, the AP level resource scheduling is performed by the gNB and/or Wi-Fi/LiFi AP. In this respect, a Linux-kernel based airtime management evaluation framework is discussed which can be used to segregate multi-WAT resources for a given 5G-CLARITY slice. Due to LiFi’s different channel and link reliability characteristics, the airtime scheduling for the LiFi technology is specifically discussed and slicing the attocellular network resources is researched. Section 4 is focused on 5G-CLARITY multi-WAT positioning solution. The associated technologies are 60 GHz mWave, sub-6 GHz, LiFi and Optical Camera Communications (OCC) based positioning. A localisation server obtains the position information from these WATs, and provides the position estimate, by fusing all the relevant data, to the entities requiring position services. Details of the overall architecture, each technology ranging/positioning scheme, and the fusion approach are provided. The simulation architecture to evaluate the integration and performance of 5G-CLARITY multi-WAT scheme, including the corresponding user- and control-plane functionalities are presented in Section 5. Results for a dense deployment of multi-WAT AP/gNB in contrast to the generic scenario, using both conservative (based on the available technologies) and opportunistic (assuming greedy usage of available bandwidth), are presented. The achievable system area capacity in each scenario is discussed and the limiting factors are introduced. Overall, this document, 5G-CLARITY D3.2, presents the achievable KPIs of the main components of the 5GCLARITY integrated 5G/Wi-Fi/LiFi network user- and control-plane architecture. The integration of these components and the evaluation of the overall 5G-CLARITY user- and control-plane will be reported in 5GCLARITY D3.3. |
