The MultiX project aims to revolutionize the 3GPP Radio Access Network (RAN) design and operation by developing a pioneering MultiX fusion Perceptive 6G-RAN system (MP6R) that will support an integrated multi-sensor, multi-static, multi-band, and multi-technology paradigm to enable multi-sensorial perception for future 6G sensing applications. The MP6R builds on top of three innovation pillars: i) MultiX Perception System (MPS) that introduces 3 levels of sensing functions into the RAN stack to support multi-sensor, multi-band, multi-static, and multi-technology Integrated Sensing and Communication (ISAC), following a streamlined functional split architecture to enable a fully flexible ISAC deployment in 6G-RAN and to facilitate vendors to extend their RAN stack to support sensing in a plug & play manner; ii) MP6R controller (MP6RC) that extends the RAN control plane functionalities to coordinate and control multi-technology integration (including 3GPP, non-3GPP, and other sensor technologies such as Radar, LiDAR, camera, etc.), while considering new connectivity approaches as well as mobility challenges for sensing and localization services; and iii) Data Access and Security Hub (DASH) designed as a novel RAN data plane entity that aggregates multi-sensor data of diverse technologies, providing secure data access, processing, storage, and exposure, ensuring data privacy and trustworthiness, and that can be fully distributed throughout the data plane wherever needed in the 6G-RAN. To fully address the 6G Sustainability goals, MultiX will explore AI’s full potential across all layers of the MP6R and develop an energy efficient AI architecture for adaptive ISAC transmitter/receiver and distributed learning with a novel low power AI engine design. The proposed MP6R design, and a set of other selected innovations, will be validated and demonstrated in two specific Proof-of-Concepts (PoCs) targeting TRL 4-5: PoC 1) Multi-layer Network Digital Twin for Industrial Manufacturing and PoC 2) Contact-free eHealth Monitoring at Home Environment. In addition, MultiX also aims to shape 6G standardization for achieving maximum sustainability and impact by contributing to relevant standard developing organizations (SDOs), including 3GPP, IEEE, ETSI – especially its ISAC Industry Specification Group (ISG), and Open-RAN (O-RAN), to pave the way for adopting the MultiX design and innovations in different SDOs both during and after the lifetime of the project.
The vision of 6G is to integrate sensing with communication in a single system, which will serve as a distributed neural network (NN) for the future Intelligence of Everything. Network sensing and native Artificial Intelligence (AI) operations are the two key aspects to build the connected intelligence in 6G. For network sensing, the use of higher frequency bands – from millimeter wave (mmWave) up to THz, wider bandwidth, and massive antenna arrays will enable high accuracy and high-resolution sensing, which can help implement the ISAC in a single system for mutual benefit. First, the entire communication network can serve as a sensor. The radio signals transmitted and received by network elements and the radio wave transmissions, reflections, and scattering can be used to sense and better perceive the physical world. The capabilities to obtain range, velocity, and angle information from the radio signals can enable a broad range of new services, such as high accuracy localization, gesture capturing and activity recognition, passive object detection and tracking, as well as imaging and environment reconstruction [TZ21]. This is the so-called “network as a sensor” paradigm. Second, the sensed information provided by these new services on the localization, detected objects, and environment knowledge via network sensing can help improve communication such as more accurate beamforming, faster beam failure recovery, and less overhead when tracking the channel state information (CSI) [KL21]. This is the so-called “sensing-assisted communication”.
Last, but not least, sensing will be served as a “new channel” that observes, samples, and connects the physical world to the cyber world. Real-time (RT) sensing will become an essential enabler to make the concept of the digital twin (DT) — a true and RT replica of the physical world — a reality in the future. The capabilities of ISAC and relevant technologies are currently being explored in a broad range of studies, including channel modeling and co-design of the hardware by sharing the spectrum for both communication and sensing, research on different waveform and signal processing design or a joint design of both, and some early work on exploring multi-band and multi-static ISAC. But most the current state-of-the-art (SotA) ISAC solutions primarily focus on a single wireless technology domain; i.e., researchers are mostly focused to work in either 3GPP or non-3GPP (like IEEE 802.11) environments, without a real integration of both worlds and, especially, without any coordination of multi-technologies in the RAN domain. However, the future 6G-RAN systems will not only across multi-band (e.g., mmWave, sub-6 GHz, THz, new frequency ranges), multi-radio interface (e.g., OTFS or OFDM cellular radio interface) and multi-static (multiple non-collocated transmitters and receivers), but also need to across multi-RAN technologies (3GPP, non-3GPP including different Wi-Fi versions). Besides, in addition to the network sensing, there are also many other diverse sensor technologies such as LiDAR, cameras, radars, which have already been widely used for traditional sensing applications and services. To this end, a huge variety of heterogeneous sensing and network technologies will all coexist in the future 6G RAN systems.
The biggest challenge is how to manage such heterogeneity and fully integrate various sensing sources and ISAC technologies throughout the RAN up to the User Equipment (UE) across different layers of the RAN stack and thus explore the full potential of each sensor, each node, each band, each technology as well as their combination to achieve the maximum spectrum and system efficiency, energy efficiency and resource usage across the entire RAN system.
We believe that the true benefits of Sensing for 6G can be significantly enhanced by enabling a real integration of different sensing sources and technologies, not only including network-based sensors leveraging on ISAC (to realize the “network as a sensor” concept) that can be based on either 3GPP or non-3GPP technologies, but also integrating other wide range of sensors. A fully integrated system with multi-sensor and multi-technology data fusion and integration and information sharing will greatly improve the performance and efficiency of both sensing and communication in a joint way, as well as reduce the overall costs and power consumption of the network system.
We also believe that such integration for the future 6G ISAC system should take place in the RAN domain to allow near real time (RT) or even RT processing and synchronization across technologies within the RAN, to meet the critical latency requirements of future sensing applications. This feature is missing in the current 3GPP specifications, and, particularly, there is not yet any integration of multi-sensor data and coordination of the sensing across different technologies, such as selection of sensors and their control within the RAN.
In addition, we envision that new connectivity and topology paradigms to perform sensing in different channels therefore freeing resources for communication are worthy to be explored for the RAN to gain maximum system efficiency, which have not yet been considered for ISAC. This will also change the way of handover (HO), from traditional mobility based to perception-driven HO and network selection, where the handover operation will be triggered upon the perception of sensing elements such as moving sensing objects.
The last, but not the least, it is a key challenge to maintain low power and energy consumption in the future 6G RAN system, considering a large number of devices, sensor, and various network nodes and the huge amount of data that is generated and processed, especially the native integration of AI in the 6G networks requires high computing and network resources and thus results in high energy consumption on transmitting and processing data as well as training AI models. To overcome this, how the data to be collected and processed, as well as how the AI models to be designed, placed and distributed needs to be fundamentally resigned for the RAN across both data plane and control plane, so as to meet the sustainability goal of 6G.
To address the above ambitions and challenges, we will develop a pioneering MultiX fusion perceptive 6G-RAN system (MP6R) in this project by integrating the MultiX concept across all system levels. The perceptive network created by MultiX will offer sensing and location information to UEs and authorized entities, catering to diverse applications and vertical sectors like presence detection, healthcare, and autonomous vehicles. Moreover, the network will also utilize sensing data for network optimization such as resource management, edge computing, energy efficiency, and beamforming. Notably, the ability of a 6G-RAN to perceive its environment significantly impacts handover (HO) operations, necessitating continuous perception for various applications. MP6R will prioritize maintaining UE sensing requirements during HOs and detecting moving objects. Collaborative sensing plays a crucial role in network architecture, aligning and extending the ongoing work in the 3GPP on Proximity Services (ProSe) integration with sensor selection. As shown in the next figure, the vision of MultiX is to create MP6R, a redesign of the 3GPP RAN system that can aggregate a set of diverse design concepts in a seamless way (defining the so-called MultiX concept) to create an integrated multi-sensor, multi-band, multi-static, and multi-technology paradigm to enable multi-sensorial perception for future 6G sensing applications.
- Multi-sensor: LiDARS, cameras, and other radio-based sensors.
- Multi-band: UE and other transmitting devices that use different frequency bands (e.g., cellular Sub-6 GHz, mmWave, new frequency ranges such as the 7-24 GHz band), denoted by inter-band, as well as different spectrum portions within the same frequency band (e.g., different channels in the mmWave band), denoted by intra-band.
- Multi-static: configuration of ISAC transceivers in which the signal transmitted by one or multiple nodes, and scattered by objects in the environment, is processed by multiple receivers that may be either synchronized or not.
- Multi-Radio Access Technology (RAT): the different sources of data that can connect to a 3GPP network, either directly via next-generation Node B (gNBs), if using cellular technologies, or via the Non-3GPP Interworking Function (N3IWF), if using non-3GPP networks like Wi-Fi.
The different sources of data are managed by the newly proposed data plane entity, the Data Access and Security Hub (DASH), which will be distributed throughout the RAN data plane where needed to store, aggregate, process the MultiX data as well as exposure of data through secure access and authentication APIs towards any third-party applications or external network functions. Given the sensitive nature of sensing data, privacy and security considerations are paramount in its design, a key design aspect of DASH is to ensure security, data privacy and trustworthiness. Thus, new security methods will be developed to ensure data privacy by addressing inherent privacy to every step of data collection, aggregation and processing as well as the interface to transmit and exposure the data to external functions or third parties. MP6R necessitates new data exposure models and interface analysis. Timely delivery and processing of sensor data impact the transport networks and radio transmission methods, potentially influencing Centralized Unit (CU) / Distributed Unit (DU) split design. Adding to the above, advanced AI techniques for sensing optimization and edge computing will play a pivotal role in enhancing network performance.
To greatly expand the perception capabilities of the 6G-RAN, MultiX proposes key wireless communication and signal processing enhancements, expressing the MultiX concept in the so-called MultiX Perception System (MPS). The MPS focuses on sustainable radio system implementations with ISAC-capable wireless devices across various frequency bands (multi-band). The MPS incorporates novel ISAC strategies that exploit the potential of new 6G capabilities such as Cell-Free massive Multiple-Input Multiple-Out (CF-mMIMO), RIS, and that are based on Orthogonal Time Frequency Space (OTFS) modulation-based sensing to optimize spectrum utilization and increase energy efficiency. Advanced RIS architectures and access schemes improve coverage, interference management, and energy efficiency while enabling control options from 3GPP/O-RAN systems. The development of a joint signal processing framework for ISAC includes both mono-static and multi-static approaches, non-line-of-sight (NLoS) identification, near-field ISAC utilization, channel sparsity exploitation, and dynamic ISAC compensation. All these new ISAC innovations will be embedded into 3 levels of MultiX Sensing Functions (SFs), including MultiX LowPHY, High-PHY and high-level sensing functions, which will be placed at the different layers of the RAN stack at the RU, DU, or a fully-fledged BS, and up to the UE device across the RAN system. MultiX SFs will be supported by AI and/or ML algorithms. Specifically, MultiX will design a novel Energy-efficient AI architectures for adaptive, event-based ISAC receivers and distributed learning from multi-static sensing devices. To this aim, a new low-power AI engine will be developed to be used at the UE level or DU or base station levels, to aid or replace traditional signal processing functions for ISAC to bring intelligent perception capabilities throughout all RAN nodes in a fully distributed manner. The low-power AI engine will feature distributed and optimized learning architectures including mixed model-based/data-driven algorithms and event-based Spiking Neural Networks (SNNs). MultiX will also explores GenAI’s potential in ISAC, developing energy-efficient models, ensuring seamless interaction with existing ISAC protocols, and optimizing data integration.
To seamless coordinate and control multi-technology integration as well as to address new connectivity options and mobility related challenges for sensing and location services, we introduce the extension of the RAN control plane entity, MP6R Controller (MP6RC) as the brain of the MP6R system, to i) coordinate and control multi-technology integration and mobility management of the sensing objects inside the RAN; and ii) interface and manage the distributed DASH nodes throughout the RAN for handling multi-sensor, multi-connectivity, and multi-technology data integration and processing.
MultiX will strongly impact standardization work across major SDOs, especially related to ISAC, involving funding members of the recently created ETSI ISAC ISG. Prototyping efforts are well underway in Open Labs like IMDEA Network, KU Leuven, IHP, and 5TONIC for physical layer (PHY) testing and system-level implementation and integration. The commitment of the project to open-source Software (SW) and the participation of the Open Labs in currently ongoing Stream C and D projects, ensures replicability and integration of PoCs in future projects within the SNS WP2025-26 Stream C/D initiatives and, more in general, by all EU-funded other related calls, e.g., future projects funded under the HEU cluster 4 on DATA or on HUMAN topics.