Skip to main content

Ultra-Dense Unsupervised Heterogeneous Wireless Cloud Coded Networks for 5G/B5G

Periodic Reporting for period 1 - RECENT (Ultra-Dense Unsupervised Heterogeneous Wireless Cloud Coded Networks for 5G/B5G)

Reporting period: 2018-11-01 to 2020-10-31

Beyond 5G (B5G) wireless networks will undoubtedly have greatly increased density and scale compared to current networks, resulting in massive interaction between nodes. Moreover, the topology may be a priori unknown and rapidly time-variant, and according to legacy networking paradigms, it will be impractical to provide a global coordinating authority and thus these networks will be essentially unsupervised. In this context, the conventional networking paradigm will be severely limited by interference in these scenarios, greatly reducing efficiency.
It has already been shown in both theory and practice that wireless physical-layer network coding (WPNC) is nevertheless capable of resolving this situation because it can allow relay nodes to extract useful information from all combined received signals, rather than treating them as deleterious interference. However, previous work on WPNC have not addressed the case of large-scale, unsupervised, secure, dynamic networks. Moreover, Systems should preferably be secure at the physical layer, thus not easily revealing their structure and waveforms, e.g. to avoid denial-of-service (DoS) attacks and the disclosure of node locations.
In this context, RECENT addresses such networks starting with fundamental theory and technology, including information theory, network coded modulation (NCM) for WPNC, stochastic network theory, and physical layer security. The developed technologies will be deployed and validated in the system level simulator and hardware-in-the-loop (HIL) platform. Running in parallel to the technical activities, a rigorously crafted innovation management programme will assess business opportunities of RECENT technologies, taking into consideration standards and regulation, and hence ensure that they achieve their full industrial and societal impact.

The project vision is implemented by the following objectives:
- OB1: Develop PHY-centric theory for unsupervised, asynchronous, heterogeneous, wireless cloud communication networks with sub-millisecond latency in dynamic environments.
- OB2: Develop secure, energy-efficient and reliable communication for unsupervised, asynchronous, heterogeneous, wireless cloud networks.
- OB3: System implementation, validation, and hardware-in-loop (HIL) experiments: thmain purpose of this objective are to validate and implement the theoretical results from previous objectives on practical environments and also to train practical research by educating them how to test, validate and calibrate proposed algorithms/protocols on software tools and hardware platforms so that their proposed research will be closer to a marketable product. The proposed HIL platform includes both hardware and software elements.
RECENT use-cases were defined: RECENT builds on 5G scenarios to propose four specific use cases that are potentially market relevant to the 5G and beyond technology roadmap. The methodologies for the selection of the use cases for practical experimentation are also described.

The specific use-cases include:
- Cell-free Massive MIMO network with NURA (Non-Uniform Rectangular Array) by using PNC: distributes the precoding/detection among a group of APs and users located around a CPU (central processing unit) where the APs utilize multiple antennas where the elements are arranged in an unstructured and non-uniform array form. PNC is then employed through the transmission between APs and CPU in order to reduce the signalling whilst maximizing the spectral efficiency of the network.
- Secure Physical Layer Coding for Massive MIMO networks: this use case will combine both Physical Layer Security (PLS) and Physical Layer Network Coding (PNC) and exploit these in massive MIMO network. The large number of antennas in massive MIMO provides spatial multiplexing gains and improve on the link reliability through spatial diversity, and with beamforming, transmissions can be localized in beams to targeted nodes.
- Ultra-Reliable Low Latency Communications with PNC targets small range hotspots delivering high-speed data to demanding applications such as augmented or virtual reality applications such as remote surgery, holographic communication, virtual meeting rooms and networked games. Simulated realities require massive amount of real-time data transfer over the air and sub-ms end-to-end latency requirements, using cloud services to jointly optimize both tasks offloading and resource allocation.
- Adaptive Multi-Connectivity using Hybrid Data Duplication and PNC: considers multi-connectivity where a user is connected to more than one BS in order to increase the reliability of the system whilst satisfying other stringent QoS (Quality of Service) requirement (e.g. throughput and latency) for 5G/B5G networks. The use case proposes a Multi-connectivity Decision Function (MDF) which determines whether the multi-connectivity user will be served by data duplication (same data from multiple BSs) or network coding (different data from multiple BSs) in an adaptive manner. This framework is expected to significantly improve the system performance in the ultra-dense and heterogeneous 5G/B5G networks by guaranteeing seamless connectivity and improved reliability, throughput and latency.

Architecture Defined: A novel RECENT functional architecture was developed as a collaborative effort between all the secondees. This builds on the current 5G architecture to cater for PNC and the notion of unsupervised networks, the latter concept based on collaboration between handsets.

Performance Metrics Defined: Out of a wide range of available metrics, there are some Key Performance Indicators (KPIs) that RECENT is interested in monitoring. These KPIs are common to the 5G/NGMN (Next Generation Mobile Networks) alliance, and other industry standards. The metrics for each use-case are defined, include latency, energy consumption, and reliability, among others; these are also quantified in order to provide the best possible end-user service.
The project is highly innovative in its scientific concepts and provides a number of distinguishing features. The concept of WPNC replaces much of the classical “layered” networking approach by extending the operational scope of the PHY, which overcomes the inevitable performance bottlenecks of the classical solution. The concept is further extended to a number of previously unexplored areas:
- Large-scale, stochastically defined dynamic wireless environments, with
- A heterogeneous, unsupervised, and weakly synchronous wireless cloud, while
- Paying attention to security and reliability.

All these issues will be directly resolved by PHY-centric solutions, which will allow completely novel B5G wireless cloud architectures, use cases and business models to be explored.

The research and industry innovation potential of the project is centered around the following core areas:
- Advanced WPNC cloud communication research creates a disruptive new paradigm for dense unsupervised radio networks. This has the potential to create completely new scientific disciplines and industries, strengthening Europe’s R&D sites and scientific potential.
- Simulation results and HIL verification of the project directly boost the competitiveness of European industry, both through major companies and their influence in forming future global standards and regulations and for SMEs rapidly exploiting first-hand results, especially in small specialized markets.