nismod2

National Infrastructure Systems Model setup, configuration and tests

View the Project on GitHub nismod/nismod2

Digital Communications

Cambridge Digital Communications Assessment Model (cdcam)

Model code: nismod/cdcam

Model documentation: cdcam.readthedocs.io

Key reference: Oughton, E.J., Russell, T., 2020. The importance of spatio-temporal infrastructure assessment: Evidence for 5G from the Oxford–Cambridge Arc. Computers, Environment and Urban Systems 83, 101515.DOI: 10.1016/j.compenvurbsys.2020.101515

For a full set of inputs and outputs:

Longley-Rice Irregular Terrain Model (itmlogic)

Model code: edwardoughton/itmlogic

Model documentation: itmlogic.readthedocs.io

Key reference: Oughton et al., (2020). itmlogic: The Irregular Terrain Model by Longley and Rice. Journal of Open Source Software, 5(51), 2266, DOI: 10.21105/joss.02266

Description

The 5G assessment model developed here can undertake system-level evaluation of wireless networks, to help quantify the capacity, coverage and cost of different 5G deployment strategies. The capacity of a wireless network in a local area is estimated using the density of existing cellular sites, the spectrum portfolio deployed and the current technologies being used (either 4G or 5G for mass data transfer). When supply- side infrastructure changes are made, such as building new cellular sites or adding new spectrum bands, the incremental enhancement of such decisions can be quantified in terms of the improved cellular capacity and coverage, as well as in terms of the required investment.

The model used is a high-resolution spatially-explicit implementation of a telecommunication Long Run Incremental Cost (LRIC) model. The model code, cdcam, [1] is made available under an open-source license, unit-tested and thoroughly documented online.

For 5G assessment, an infrastructure planning simulation model is developed which consists of a set of interconnected software modules. The model represents the key rollout period from 2020 to 2030, across spatial zones in the Arc, as illustrated in the figure below.

Digital communications system-level evaluation framework.

Necessary data inputs include spatially disaggregated demographic forecasts, taken from SIMIM in this study, as well as forecasts on how per user data demand will evolve in the future. Geospatial information is also required for site locations, as well as data on the available spectrum portfolio by carrier frequency, bandwidth and technology generation.

Demand assessment

The mobile demand assessment module takes into account the two main drivers of demand for cellular capacity: (i) the per user throughput rate and (ii) the number of users in an area. The total number of active users accessing the cellular network in an area is estimated and multiplied by the average user data rate to obtain the total data demand being placed on the radio access network.

Per user data demand is taken from the widely-used Cisco traffic forecast. [2] The adoption of unlimited data plans is likely to have a substantial impact on data growth, with UK mobile traffic expected to grow at 38.5% Compound Annual Growth Rate (CAGR) over the coming years.

Population scenarios at Local Authority District level are disaggregated to 9,000 Postcode Sectors using weights based on shares of 2011 census population. We model a hypothetical operator with a market share of 25% of users, in line with the UK’s Mobile Call Termination Market Review. [3] It is reasonable to expect that not all users will access the network at once, and therefore an overbooking factor (OBF) of 50 is used, which is standard practice for network dimensioning traffic throughput. [4] Smartphone penetration in Britain is 80%, so only this proportion of the population is assumed to access high capacity wireless services such as 4G LTE or 5G.

Capacity assessment

The capacity assessment module is capable of quantifying cellular capacity expansion using three methods, including improving spectral efficiency via new technology generation, the provision of new spectrum bands and the deployment of new cells to densify the network.

The mean spectral efficiency is obtained using a stochastic geometry approach via the open-source python simulator for integrated modelling of 5G, pysim5G. [5] First, pysim5G estimates the Signal to Interference plus Noise Ratio in different urban and rural environment using industry-standard statistical propagation models. Next, a spectral efficiency is allocated for the level of received signal at the user, based on the ETSI coding and modulation lookup tables for 5G. [6] The estimated cellular capacity can then be obtained for an area by multiplying the spectral efficiency by the bandwidth of the carrier frequency. To ensure a specific Quality of Service, the stochastic approach allows the 10th percentile value to be extracted from the distribution of simulation results for each frequency. This means that the network will be upgraded to meet a desired user capacity at the cell edge with 90% reliability.

Physical sites data are taken from Ofcom’s Sitefinder data and updated to be consistent with existing 4G coverage statistics released by Ofcom’s Connected Nation report. In recent years, passive infrastructure sharing agreements have essentially created two physical networks in the UK, the first between Vodafone and O2 Telefonica (‘Cornerstone’) and the second between BT/EE and Hutchinson Three. We consider the Vodafone and O2 Telefonica (‘Cornerstone’) sites as the key supply-side input for (predominantly Macro Cell) sites. Representative site locations are obtained by taking latitude and longitude coordinates for individual cell assets, buffered by 80m, with the polygon centroid of touching buffers forming an accurate location approximation. This results in approximately 20,000 sites.

The statistics are disaggregated by ranking the revenue potential of each postcode sector and calculating the cumulative geographic area covered using the expectation that mobile networks operators (MNOs) rationally deliver 4G coverage to the highest revenue sites first. This approach is consistent with how MNOs deploy new cellular generations.

This assessment considers a hypothetical operator, representing a set of average operator characteristics. A set of representative 4G LTE and 5G New Radio (NR) carrier frequencies and bandwidths are tested in Frequency Division Duplex mode. These frequencies consist of 10 MHz bandwidth for each of the 700 MHz, 800 MHz, 1.8 GHz and 2.6 GHz bands, 40 MHz bandwidth for 3.5 GHz, and 100 MHz bandwidth for 26 GHz. The Total Cost of Ownership is estimated for each asset by calculating the Net Present Value of the initial capital expenditure required in the first year of deployment as a one-off cost, combined with the ongoing operating expenditure over the lifetime of the asset (with opex being 10% of the initial capex value for all active components, annually). A discount rate of 3.5% is used over a period of 10 years. This calculation does not consider price trend changes and assumes a 10-year lifespan of Macro Cells. The total cost per square kilometre for different network configurations can then be calculated based on the density of assets by area. The costs per asset item are based on the Mobile Call Termination model. [7]

Fibre-to-the-Premises

The fixed broadband modelling assesses the cost of Fibre-to-the-Premises based on density of premises in Output Areas under different urban development scenarios. Openreach do not make detailed fixed broadband network data publicly available to use for modelling. Therefore, the approach taken here is to use network cost information from the report produced by Tactis & Prism for the National Infrastructure Commission, as the analysts had access to the necessary Openreach network data.

With this information, a cost modelling ‘geotype’ approach is used which is based on the Office for National Statistics’ (ONS) urban-rural local authority categories. A geotype is a group of geographical areas which have similar cost properties. The six geotypes are based on a categorisation which ranges from the densest urban conurbation, to remote rural areas.

To provide a geographically granular analysis, and to take the Arc scenarios into account, premises estimates for 2050 are taken from the Urban Development Model (UDM) outputs, where each hectare grid cell is either undeveloped or developed at a given density. These results are then aggregated to the 11,085 ONS Output Areas within the Arc. Density of premises defines the geotype, and therefore the cost per premises, for each Output Area under each urban development scenario. The total cost estimates follow from number of premises and cost per premises in each area.

The analysis script for this process is available within the nismod/digital_comms repository at arc_fixed.py.

References

  1. Oughton, E.J., and T. Russell (2019). The Cambridge Digital Communications Assessment Model v1.0.2. Available online: https://github.com/nismod/cdcam doi: 10.5281/zenodo.3583132.
  2. Cisco (2017). VNI Mobile Forecast Highlights Tool.
  3. Ofcom (2018). ‘Mobile Call Termination Market Review 2018-21: Final Statement – Annexes 1-15’.
  4. Holma, Harri, and Antti Toskala, eds. (2012). LTE-Advanced: 3GPP Solution for IMT-Advanced. Chichester, UK: John Wiley & Sons, Ltd.
  5. Oughton, E.J. (2019). ‘Python Simulator for Integrated Modelling of 5G (Pysim5g)’ Available online at https:// www.github.com/edwardoughton/pysim5g
  6. European Telecommunications Standards Institute (2018). ‘ETSI TR 138 901 V15.0.0 (2018-07). 5G; Study on Channel Model for Frequencies from 0.5 to 100 GHz (3GPP TR 38.901 Version 15.0.0 Release 15)’.
  7. Ofcom (2018). Mobile Call Termination Market Review 2018-21.
  8. Oughton, E., Russell, T. 2020. The importance of spatio-temporal infrastructure assessment: Evidence for 5G from the Oxford–Cambridge Arc. Computers, Environment and Urban Systems, 83, 101515, DOI 10.1016/j.compenvurbsys.2020.101515.

cdcam data pack

DAFNI dataset

DOI 10.5281/zenodo.3525286

Contains population scenarios at LAD scale created for ITRC studies from Arc Scenarios workflow defined at https://github.com/nismod/arc-scenarios

Baseline population per postcode sector is derived from ONS/NOMIS (2011) Table KS101EW Usual Resident Population http://www.nomisweb.co.uk/census/2011/ks101ew

Boswarva, Owen. (2017). Sitefinder Mobile Phone Base Station Database, [Dataset]. https://doi.org/10.7488/ds/1975.

Capacity look-up table by frequency, output from radio propagation model. See itmlogic

Local authority district, postcode sector area definitions, covered by the Open Government License Contains Ordnance Survey data © Crown copyright and database right 2012.

OFCOM Connected Nations (2018) Mobile local and unitary authority - Coverage and performance data https://www.ofcom.org.uk/__data/assets/file/0009/131040/201809_mobile_laua_r02.zip

Mobile data traffic forecasts from VNI Mobile forecast highlights tool Cisco (2017) https://www.cisco.com/c/m/en_us/solutions/service-provider/forecast-highlights-mobile.html

Population density bands to define geotypes following OFCOM Mobile call termination market review 2018–21 https://www.ofcom.org.uk/consultations-and-statements/category-1/mobile-call-termination-market-review