Abstract
There are 28 million homes in the UK, with a requirement for 53 million Smart: gas and
electricity meters. Out of these, 20 million homes will rely on Telefonica’s mobile network
to connect them to energy supply companies, raising concerns over their connectivity.
These ‘connected’ gas and electric meters are more commonly and collectively known
as ‘smart meters’ and are typically connected to a Wide Area Network via a
Communication Hub. It is vitally important that wide scale ubiquitous connectivity is
provided to deliver service to 99% of the UK population, while avoiding any non-essential
investment.
The radio prediction models used by Mobile Network Operators typically use a statistical
approach, and therefore to increase model accuracy at a particular point location you
simply increase the prediction error margin of the link budget. However, this reduces the
coverage area by the inverse square of the range reduction, meaning that improving the
accuracy of the model will result in a significant reduction in the predicted coverage
footprint, resulting in large numbers of additional base stations being required. Making it
difficult to meet contractual obligations in a financially viable way, and significantly
delaying rollout.
Based on the signal strength measurements conducted at the Building Research
Establishment determining building penetration losses in both 900 and 2,100 MHz band,
this thesis presents building penetration loss models using these measurements that are
practical and cost effective when compared to traditional statistical propagation loss
models. Building on this model, the thesis then provides an alternative method of
calculating reliable mobile network coverage for wide area networks serving smart meters,
without requiring a significant prediction ‘error margin’.
The aim of the research documented in this thesis is to develop a prediction system that
accurately predicts in-building mobile coverage, whilst ensuring that the predicted range of
the cell was not reduced. The provided framework then provides a highly accurate
estimation of coverage, enabling a binary decision to install a Smart Meter or not. This
thesis also provides alternative solutions for hard-to-reach locations such as enhancing the
coverage range of antennas with a mesh radio network technology. Extensive
measurements are made in Newbury in both suburban and open environments for
validation and delivery of a simple statistical model for the 868 MHz band in United
Kingdom conurbations.
Results presented in this thesis will analyse the effectiveness of a Mesh radio network's
ability to provide an effective coverage extension method at 868 MHz. As it is expected
that Mesh radio will contribute to a significant increase in the number of connected smart
meters. Finally, a propagation model derived from the measurement campaign data will
then provide a method to estimate the likely coverage extension range of a mesh radio
gateway.