Use Case #1: Sensing enabled Services
Storyline #1: Exploiting sensing information to improve communication services (sensing-aided communication)
Current status, Problem statement – limitations of today’s situation.
Nowadays, MNOs are essentially the dominant stakeholders operating mobile networks serving large numbers of devices. 5G solutions usually aim at exploiting 5G network information captured by gNBs or UEs and processed by networks planning applications, towards optimising network deployment following time-consuming planning- configuration – replanning- reconfiguration cycles.
In micro-planning cases where link quality is significant, radio network solutions automatically optimising link quality and performance are focusing on exploiting closed-loop beamforming to focus the antenna beams towards communicating devices (user devices, vehicles, etc.). Also, in this case, this incurs significant planning overhead as it is not fully automated. Moreover, the communication channel quality can be affected by obstacles, i.e., objects blocking or reflecting the propagation of the communication signal; cases which need to be detected to avoid steering beams towards blocked directions. Further considering existing technical challenges, beamforming is used based on exhaustive or hierarchical beam search to optimize the communication services. Beam prediction is leveraged to lessen the measurement overhead and latency, primarily based on supervised learning using neural networks. However, these face scalability issues due to the complexity of training and deployment. Therefore, solutions for beam prediction exploiting another type of sensor have been proposed, such as radar, camera or combination of camera and GPS, which can lead to improved performance but at the cost of increasing the complexity and the cost of the deployment.
In both cases it becomes apparent that automated sensing of the environment (in terms of traffic demand, or physical structures) is crucial in optimising network performance towards delivering high quality services.
From a third perspective, device positioning services (required both for network optimisation and for service provisioning) are already available in traditional wireless communication systems. However, the accuracy is limited by the location, availability of Line-of-Sight (LoS) and number of APs. Compared with deploying APs, deploying RIS is more flexible and the cost is lower, but still the integration needs to be performed to reap the higher spatial resolution and positioning accuracy. The combination of ISAC and RIS can offer a solution to this problem.
6G-SENSES concept
Considering the 6G vision, a multitude of basic and advanced services will be delivered to verticals or individuals as end users in versatile environments. A precondition – and, at the same time, an enabling capability for this – is the provisioning of Ubiquitous coverage with the required service performance. At the same time, the advent of RIS and ISAC introduce new capabilities for achieving this goal, while enabling new roles of the 6G ecosystem. This use case revolves around the development of ISAC capabilities in multi-WAT radio units (RUs), extended to RIS, with the purpose of delivering a solution for network coverage and performance optimisation based on environment sensing.
More specifically, we consider the case of a 6G ecosystem as detailed in Chapter 3. Multiple RAN Resource providers/ RAN Operators undertake the role of deploying RAN elements (of various technologies) -including RIS- that enable ISAC. The sensing data/information captured by RAN resource providing elements/roles is either processed and used for RAN self-optimisation (e.g. via closed loop beamforming, RIS configuration for communication enhancement) or sent to a RAN Operator or NOP for deciding on the network reconfiguration to optimise performance based on environment and traffic assumptions.
In more technical terms, 6G-SENSES leverages the work on ISAC to demonstrate – in an E2E small-scale proof-of-concept (PoC) prototype –ISAC performance in a form of an operated RU, DU, and CU/UPF pools with a RT control fabric providing sub-millisecond control loop over each network component. This solution will benefit from the availability of sensing information stemming from other non-3GPP technologies, such as Wi-Fi, and mmWave. This use case will deliver 6G communication links and sensing of the RAN environment and simultaneous access to the channel by multiple users.
Optionally, this use case can involve RIS panels, in a way that a LoS link can be established to provide the sensing service for the Non-Line-of-Sight (NLoS) areas and an extra LoS reflected link can be provided by the RIS to sense the target from a different angle, thus the sensing performance of the RIS-assisted networks can be significantly improved. Wireless communication will be optimized given the availability of sensing information stemming from a multi- WAT RAN. This sensing information is extracted from static and mobile environments and the information will be harnessed for the optimization of beamforming, allocation of resources and even for prediction purposes.
Storyline #2: Enabling Active Sensing with Wi-Fi system and Wi-Fi sensing standardization design
Current status, Problem statement – limitations of today’s situation.
The current SoTA primarily utilizes passive Wi-Fi signals for sensing, but the active approach proposed here is a novel concept that has yet to be fully explored. By integrating active sensing, we can expect to see a new wave of innovation in ISAC, leading to more immersive AR/VR/XR experiences and a transformative impact on user interaction with laptops and other smart devices. This proposal not only aligns with the current trajectory of technological advancement but also charts a course for future innovations that will enrich our interaction with the digital world. An additional significant impact of active sensing lies in motion presence detection. This capability plays a crucial role in managing network resources and optimizing the EE of the network based on sensing information. By intelligently detecting the human presence, the system can optimize the energy and resources of the network based on this information.
6G-SENSES Concept
Considering the 6G vision, a multitude of basic and advanced services will be delivered to verticals or individuals as end users in versatile environments. Key enabler for the support of “Immersive Experience/ Seamless Immersive Reality” along with “Collaborative Robots/ Cooperating Mobile Robots” services is the incorporation of sensing capabilities.
More specifically, we consider the case of a 6G ecosystem as detailed in Chapter 3. Multiple RAN Resource providers/RAN Operators undertake the role of deploying RAN elements (of various technologies) -including Wi-Fi- that enable ISAC. The sensing data/information captured by RAN (Wi-Fi) resource providing elements/roles is either processed and for network optimisation (see Storyline #1) or sent to provided as a service to service providers/ verticals, to enable Immersive Communication and Collaborative Robots service elements for relevant applications.
In more technical terms, considering the afore mentioned technical limitations of current solutions, the algorithms developed for this purpose are designed to refine the process of delay and Doppler estimation, which are critical parameters in determining the relative motion and distance of objects in the vicinity of the sensor. These advancements pave the way for a host of applications, particularly in the burgeoning fields of AR, VR, and XR. The implications for these technologies are profound, as they rely heavily on seamless and intuitive interaction with virtual environments.
Furthermore, active Wi-Fi sensing has the potential to revolutionize power management and network resources. By intelligently detecting the absence of human presence, for example, the system can transition into a low-power state, thereby conserving battery life and extending the usability of the device.
This proposal has the potential to revolutionize how we interact with Wi-Fi networks, making them not only data conduits but also valuable sensors for understanding human behavior and environmental dynamics. The expected outcomes are the following:
- A comprehensive evaluation of Wi-Fi-based activity sensing techniques.
- Clear requirements and standardization guidelines for incorporating activity sensing into Wi-Fi protocols.
- Innovative applications leveraging Wi-Fi intrinsic ability to sense the surrounding environment.
- Develop active sensing ability that works simultaneously with regular communications.
Use Case #2 — Ubiquitous Connectivity & Immersive Services
Current status, Problem statement – limitations of today’s situation
CF-mMIMO is currently in the focus of 6G research due to the promising capacity and link performance capabilities, in view of delivering high capacity/high reliability immersive services. Experimental work related to CF-mMIMO is carried out in many SNS JU research projects (e.g. MARSAL.
The main challenges associated to such technology are listed below, with some in the context of deployments of CF-mMIMO in large outdoor service areas.
- Synchronization (Phase coherence). In the coherent transmission mode, the APs coherently precode and broadcast the identical data symbol to every user and function as a distinct antenna array. However, that requires phase synchronization among APs, which is challenging when APs are managed under different CPUs owing to the difficulty in synchronizing the transmission channel.
- The impact of a large Doppler on the wireless channel. Effective estimation and evaluation of CF-mMIMO systems in mobile situations are rendered more challenging owing to the fluctuations in the mobile users’ velocity in RT.
- Given the large number of deployed APs, selecting the optimal APs (dynamic AP cluster per user) for handover can be challenging.
- Coordination of the beamforming vectors among the distant APs can be challenging and may introduce latency.
6G-SENSES Concept
Considering the 6G vision, a multitude of basic and advanced services will be delivered to verticals or individuals as end users in versatile environments. Especially for the support of Immersive Services Usage scenarios, a precondition – and at the same time an enabling capability for this is the provisioning of Ubiquitous coverage with the required service performance. At the same time, the advent of CF-mMIMO and RIS introduce new capabilities for achieving this goal, while enabling new roles of the 6G ecosystem. This use case revolves around the development of CF-mMIMO capabilities (also RIS-assisted), with the purpose of delivering a solution for augmented network coverage with high channel quality.
More specifically, we consider the case of a 6G ecosystem. Multiple RAN Operators undertake the role of deploying RAN elements and provide these to a CN/Network OP. These elements feature CF-mMIMO capabilities (being also RIS-assisted) enabling high-capacity network deployments, and connectivity for massive number of devices.
In more technical terms, the CF-mMIMO system consists of many distributed APs, with multiple APs serving one user (which is known as user-centric). It has been shown that a distributed CF-mMIMO system performs better than a centralized system under certain conditions. In recent years, the CF-mMIMO systems have been evaluated for data rate, outage probability, and EE. Topics such as pilot contamination, scalability, and system integration have also been studied theoretically.
The major challenge in the practical implementation of CF-mMIMO systems is the overhead in terms of signaling and coordination for pre- and de-coding of the transmitting and receive data, the data itself, and the required control signals for synchronization or scheduling. The scheduling and UE to AP assignment is a difficult combinatorial problem which could be solved by machine learning, e.g., by graph neural networks. Different splits of the transceiver chains are available and the so-called functional split can be selected and optimized. Different options to perform rate splitting in the CF-mMIMO can lower the demand on the front-haul links. Depending on which calculations and functions are running on the APUs and on the virtual RANs, the corresponding interfaces and signaling via the fronthaul connections must be defined. Depending on the architecture of the physical network (fronthaul network) between the APs and the (one or more) Central Units (or CPUs) and, depending on the functional split, an efficient signaling of the data (user plane) and the control signals (control plane) must be found.
Consider a multi-WAT deployment case where some of the nodes (Sub-6, mmWave) feature CF-mMIMO capabilities. The APs and UE CF-mMIMO PHY layer are able to establish a data plane and expose interfaces to be monitored and controlled. A RAN RT control fabric will enable the RT control of the 6G PHY prototype, e.g. instructing each AP to be active or passive when serving a user.
Use Case #3 — Network DT – Network Optimization
Current status, Problem statement – limitations of today’s situation
A network digital twin (NDT) allows the assessment of configurations, methods and algorithms prior to their application in the network under evaluation, and allows predicting its performance under different conditions. It resembles a zero-risk environment where the network can undergo diagnose and emulation without impacting the real network. This concept enables the training of ML solutions to optimize the RAN PHY processing, e.g. using Reinforcement Learning as a decision-making tool.
In the context of 5G/6G offline and online network digital twin, the project will leverage a variety of WATs to ingest cross-technology sensing, telemetry, and control into the evolved 6G RIC. The goal is to enable efficient management and optimization of heterogeneous networks with multiple WATs, each with its own unique characteristics and performance capabilities.
To achieve this, the project will focus on the injection of native 3GPP LCS/Positioning from the LMF into the 5GNR OAI-based RIC, and potentially direct Near-RT injection into Near-RT RIC via E2+. The O-RAN RIC will also be extended to support cross-domain integration of Non-3GPP RAT control-plane and injection of Wi-Fi sensing for Sub-6 and mmWave networks. Wi-Fi sensing using active sensing technology can be used for indoor sensing of human proximity and approximate location. This context can be injected into the sensing system, and will help extend the reach of ISAC systems. With these enhancements, the project aims to enable real-time network monitoring, control, and optimization to achieve superior network performance and user experience.
Moreover, the project will explore the potential of multi-RAT ISAC to further enhance network performance and reliability. By leveraging advanced sensing capabilities across different wireless domains, including non-3GPP RATs such as Wi-Fi 7/8, the project aims to improve the accuracy and reliability of network sensing and monitoring. This will enable proactive network management and optimization to mitigate potential issues and ensure seamless service delivery.
Nowadays, networks are designed based on initial information/estimation about the environment and the traffic demand. Based on these, network elements are placed so to optimize the network performance and deployment efficiency. However, the surrounding environment and traffic demand may change after the network deployment is ready, in ways that are not foreseen at initial network planning phases, e.g., due to an incident or the evolution of services to be provided. In legacy network deployments this is performed via manual network reconfiguration, or by the deployment of additional network elements, being not instantaneous and requiring from hours to days timedays. The vision of “Ubiquitous coverage” (in terms of place and time) implies that 6G networks should continuously monitor the environment and optimize their configuration to optimise network performance for all service provisioning stakeholders.
Moreover, a wide set of vertical services envisioned for 6G require knowledge of the environment. This can be captured by tailored sensors detecting, e.g., movement/obstacle/carbon/air/ noise, etc., at application level or by the implementation of ISAC as inherent network capabilities. 6G vision is steered towards ISAC and AI capabilities as means to enable 6G services. Such types of services, as indicated in Chapter 3, include Immersive Communication and/or Collaborative Robots service elements.
From a vertical service perspective, digital twining involves creating a virtual representation of a physical system. The virtual entity is synchronized in real-time with its physical counterpart, allowing for monitoring, analysis and optimization throughout the physical system lifecycle. In contrast, Massive IoT refers to the large-scale deployment of interconnected IoT devices collecting and exchanging vast amounts of data from a multitude of sensors and systems. Nowadays, this plethora of IoT devices enables comprehensive monitoring, automation, and data-driven decision-making across various industries. This data-driven approach improves decision-making, lifecycle management, and customer experiences by offering detailed insights and remote monitoring capabilities. Latest efforts focus on combining IoT data with DTs to support sustainability and energy management by identifying and implementing energy-saving measures, ultimately driving efficiency and innovation in various industries.
An exemplary storyline of such vertical use case is the deployment of various sensors in a historical monument, providing temperature, humidity, air quality, motion detection, and movement information, which are used to simulate potential issues, provide predictive maintenance schedules, and ensure the optimal preservation of historical artefacts and architecture. However, sensor-based sensing requires the deployment of a massive number of sensors along with radio network coverage for communication.
6G-SENSES Concept
Considering the 6G vision, a multitude of basic and advanced services will be delivered to Verticals or individuals as end users in versatile environments. A precondition and, at the same time, an enabling capability for this is the provisioning of Ubiquitous coverage with the required service performance. Simultaneously, the advent of O-RAN and ISAC introduce new capabilities for achieving this goal, while enabling new roles of the 6G ecosystem. This use case revolves around the joint exploitation and development of O-RAN and ISAC capabilities – coupled with AI and digital twinning functionalities – with the purpose of delivering a solution for network coverage and performance optimisation based on environment sensing.
More specifically, we consider the case of a 6G ecosystem as detailed in Chapter 3. Multiple RAN Resource providers/RAN Operators undertake the role of deploying RAN elements (of various technologies) that enable ISAC. At the same time, the O-RAN Operator maintains the network configuration data and fuses it to a relevant xApp that shapes the network DT. The sensing data/ information captured by RAN resource providing roles is sent to the O-RAN Operator who fuses this information in relevant xApp. The latter suggests changes to the network configuration with the aim to optimise network performance based on environment and traffic assumptions.
Based on the xApps/rApps, the O-RAN Operator can undertake the task to:
- Self-serve network optimization.
- Provide optimal network configuration suggestions to the NOPs.
- Help NOPs in better network architecture planning. NOPs based on suggestions can decide whether to accept the suggestion, to delay it, or even to reject the suggestion.
- Provide sensing data and enable Immersive Communication and Collaborative Robots service elements for applications to service providers/ verticals.
Augmenting the envisioned Use Case with the vertical service perspective, we consider the sensing data captured by the O-RAN segment being provided to the Service provisioning layer under specific terms and conditions; in order to further enable a number of Vertical services (the Vertical being both the site owner and/or the site visitors). Indicatively, such data can be exploited by digital twinning applications similarly with the aforementioned IoT/sensor captured data (especially in crowd/ human activity sensitive sites/environments), so as to optimize comfort, EE, and efforts to ensure the optimal preservation of the site (e.g. historical artefacts and architecture). These data, and enabled services, can also improve crowd management, event planning, and overall site visitor experiences along with site infrastructure management services (e.g. smart lightning and heat energy savings).
