Radio frequency systems performance in Smart Cities

Posted on

< Project Overview >

Introduction

The densification of the population in urban environments and the use of wireless technologies and systems means that stress is being placed on the underlying phenomena of radio frequency (RF) transmission and reception. Hence, measuring and interpreting the use of the RF spectrum to deliver connectivity in the emerging Internet of Things (IoT) networks, as well as cellular mobile and Wi-Fi communications, in dense urban environments is crucial to ensure that critical messaging takes place without problems. Equally, gaining an understanding of how the ever-changing urban landscape affects the operation of wireless networks is essential to the planning of connectivity systems. To meet these needs, research needs to be conducted with equipment for RF surveys being installed on-board an electric vehicle, ‘drive-test’, complemented by the ‘walk-test’ equipment.

Project aims

As the IoT technologies become more prevalent and systems based on these technologies become increasingly pervasive, the efficacy of wireless connectivity supporting critical infrastructure increases in importance.

Working with TheThingsNetwork Sheffield and the Urban Flows Observatory at the University of Sheffield, we have sought to gain a greater understanding of the operation of wireless networking in a changing urban environment. This involved using the parts of the city of Sheffield as a case study and we have exposed the work undertaken to organisations, such as Sheffield City Council, Barnsley Metropolitan Borough Council via the Digital Media Centre and SmartSheffield. While the intention was to conduct extensive surveys of the network present, COVID-19 restrictions resulted in a shift towards developing a comprehensive modelling capability with more limited measurement activity. However, we expect this to prove extremely useful for future work where models can be compared with empirical measurement. The measurement systems are based on an autonomous mobile network scanner that allows us to survey the RF spectrum in terms of frequency occupancy, signal strength, and signal source direction. Our goal with this is to obtain better insights into the above-mentioned phenomena, thereby influencing the design of cities and IoT platforms to effectively unlock the potential of the IoT.

What was done?

A spectral survey of networks supporting IoT, mobile cellular, and Wi-Fi communications was conducted with an autonomous mobile network scanner, Rodhe & Schwarz TSMA6, installed in our Smart ForFour electric vehicle called ‘Morca’. The ‘drive-test’ surveys involved frequency band occupancy, signal strength, and signal source direction, and identified network name association (where appropriate). Besides, the surveys investigated the current use of the prospective frequencies intended for future 5G networks and explored the performance of low-power IoT alternatives, such as NB-IoT and LoraWAN. In addition to the network scanner, portable GPS devices, e.g., TTGO T-Beam, Adeunis, and Zane, were also used to map the network coverage by periodically connecting to the gateways when in range, fulfilling the ‘walk-test’ survey together with the TSMA6. In addition to empirical survey work, a few university buildings (initially three) around the cellular base stations, Wi-Fi access points, and IoT gateways were selected and modelled in RANPLAN Wireless (indoor & outdoor wireless planning software) to simulate the same parameters of interest, the results of which were compared with those of spectral surveys.

The findings revealed the factors that lead to signal loss and degradation in urban environments.

Results

  • A thorough analysis of the scientific and grey literature on the performance of wireless networks in urban environments
  • A highly granular IoT network (TheThingsNetwork) coverage map via the data collected on highways and streets of the city using the urban sensing electric vehicle and pedestrian surveys via GPS trackers
  • A better understanding of the issues associated with network mapping (while using mobile scanners in ‘walk-test’ trials)
  • Potential solutions addressing the shortcomings of current wireless architectures that fail to offer reliability and efficiency
  • Surveys identifying the causes of fading in urban system operations to be used before deploying any particular wireless system, e.g., IoT, Wi-Fi, and mobile cellular
  • A guidance note on how to make IoT friendly cities that addresses technical, regulatory, and policy challenges

Deliverables and other tangible outputs

  • A document listing 47 stakeholders and their potential impact and influence
  • A white paper summarising the grey literature on the factors that lead to signal degradation in cities, the impact of the ever-changing urban landscape, and the building materials on the performance of wireless networks
  • A guidance note on both IoT and urban design to build IoT-friendly cities that addresses technical, regulatory and policy challenges
  • 3D models of the selected University buildings and the city of Sheffield built in Ranplan Wireless to simulate wireless networks (indoor and outdoor) and analyse their performances, leading to the redeployment of the associated equipment to achieve the optimal coverage
  • A journal paper (submitted to IEEE Internet of Things Magazine) proposing a novel network architecture comprising energy harvesting low-cost mobile sensors to achieve the same or better coverage with fewer units, hence lowering the cost of implementation and better use the allocated spectrum. The effectiveness of the proposed architecture is validated through a case study that took place in two cities of the UK (Sheffield and Southampton)
  • A workshop/conference paper (to be submitted) on indoor modelling and optimization

Impact

  • With this project, we became able to model buildings and terrains in the 3D space using dedicated software, which allowed us to assess indoor and outdoor wireless network performance across the whole city. The outcomes of this effort were supported by empirical measurements, leading to a fine-grained network map showing the current status of the existing wireless infrastructure. The surveys also revealed the impact of the ever-evolving urban landscape on the network operation. By providing insights that are more relevant and meaningful to city dwellers (one of the stakeholders), we could helped them to understand how the position of wireless components and the materials surrounding them affect the signal reception, enabling them to take initiatives for improving the wireless coverage and hence the operation of their devices.
  • Using the outcomes of the project, we applied for UKRI’s QR Strategic Properties Fund and secured £12k of additional funding. The new project will last two months (February to March 2021), during which we will extend the outcomes of our Pitch-In project in the directions mentioned under “Next Steps” with a particular focus on policy-making.
  • We have been invited to give a talk around our outcomes in the next Sheffield IoT Meetup event, which will be held in late-March. IoT Meetup is a forum bringing relevant bodies, such as practitioners, stakeholders, and city dwellers, to share experiences and develop best practices.

By providing connectivity through the deployment of wireless access points, we have plans to extend this project to the neighbouring rural areas, e.g., Peak District National Park. This effort would focus particularly on healthcare monitoring and remote surveillance of hikers, cyclists, and people doing other outdoor activities to assist first-responders to accurately and effectively attend to emergencies.

Furthermore, indoor modelling/analysis would be conducted in collaboration with Barnsley DMC02 to determine the optimal number, location, and operation conditions of wireless components, e.g., hotspots and femtocells, achieving the optimal coverage in buildings.

In addition, if possible, we would like to investigate drone-based applications, in which the drones operate as service providers, i.e., flying base stations/gateways for wireless networks, ensuring ground-to-air and air-to-ground connectivity. Since the surveys done revealed the poorly-covered areas in cities, i.e., the dead zones, the drones could be deployed to these places on-demand to ensure that critical messaging takes place without problems.

Since the ultra-reliable low-latency communications (URLLC) are of utmost importance for both 5G and connected cars, in-car wireless coverage during driving in the city could be assessed through experiments and simulations, the results helping to improve the wireless infrastructure ensuring URLLC in cities.

Lastly, the intelligent reflecting surfaces could be modelled in Ranplan Wireless, and their optimal locations in cities determined to provide enhanced network performance.

The datasets generated related to named applications will be made publicly available to help researchers to train their models and build several top-tier applications.

Lessons learned

  • Data collection: We collected enough data to (~20k measurements) have a better understanding of the performance of the existing wireless infrastructure.
  • 3D modelling and simulation of mediums: Using a dedicated software called Ranplan Wireless, we realistically modelled buildings and terrains in the 3D space. In this software, we positioned the existing wireless components (base stations, access points, and hotspots) into their original locations (sourced from MastData) and ran simulations to analyse their performance.
  • Better coverage: Using the surveys made through experiments and by simulations, we determined the factors leading to signal fading, i.e., poor coverage, hence we suggested the ‘right’ locations for the wireless components to provide the optimal service quality.

The approval of the Risk Assessment application required for the experimental study took much longer than expected due to the Covid-19 pandemic, i.e., the altered working conditions/habits and/or reduced/affected personnel. The restrictions also postponed the planned work and the corresponding outputs and led to cancelling the mapping study that was proposed in Newcastle and our collaboration with Newcastle University.

The timely approval of the Risk Assessment application would have been useful to realize the planned collaborations with the external bodies.

TheThingsNetwork gateway roll-out, undertaken by the Urban Observatory was delayed and its earlier deployment and completion would have led to a richer data set with respect to network coverage.

Access to deliverables, resources and media content

The data collected during this project relates to network performance metrics, such as frequency occupancy, signal strength, and source direction, as well as spectrum-related information, e.g., network name association, of mobile cellular, WiFi, IoT (LoRaWAN, NB-IoT), and 5G networks. The datasets have been uploaded to the University’s data repository (ORDA) and also been visualized using a platform called TTNMapper (https://ttnmapper.org/advanced-maps/, dataset file names indicate the device names).

The pre-print versions of the technical articles, the white paper (grey literature report), and the guidance note on IoT and city design can be also found on ORDA: https://figshare.com/projects/RF_systems_performance_in_Smart_Cities_Automated_NeTwork_EvaluatioN_iN_urbAn_Settings_ANTENNAS/100940

Furthermore, we produced two blog posts, which are as follows:

1) RF Spectrum – An unseen, under-valued, much-used resource… | Source: https://urbanflows.ac.uk/morca/

2) LoRaWAN infrastructure in Sheffield | Source: https://abc-rp.com/lorawan-infrastructure-in-sheffield/

What has Pitch-In done for you?

The subject of this project has been something of interest to the researchers for some time however, it wasn’t until this funding became available that the work could proceed.

In conjunction with the Urban Observatory, the data collected provides valuable insight when considering the deployment of urban sensing systems for initiatives such as Smart Cities.

Project lead:

Professor Martin Mayfield, the Unviersity of Sheffield
Professor Tim O’Farrell, the University of Sheffield
Steve Jubb, the University of Sheffield

Partners:

The University of Sheffield Urban Flows Observatory
Newcastle University Urban Observatory

Further public and private sector partners to be confirmed.

< Related Content >