Aerospace and Electronic Systems Magazine May 2018 - 48

Enabling RF Data Analytics Services and Applications via Cloudification

Figure 2.

RF spectrum data representing Allen County, Indiana.

the time dimension and one-time measurement campaigns were
largely adequate. Due to the insatiable demand for higher data
rates and more spectrum, the spectrum environment has become
increasingly complex. Future spectrum allocations will require
RF devices to share spectrum. Information about how efficiently
is spectrum being used is increasingly required, and one-time
measurements of spectrum occupancy are clearly inadequate [17].
Spectrum monitoring technology has evolved from one-time to
long-term, and now to permanent capability conducted not by
spectrum analyzers, but by specialized platforms as the one described in this paper.
One example of mining the results of spectrum monitoring
is shown in Figure 2, capturing activity on Fort Wayne, Indiana's
public safety P25 radio system that includes fire safety, emergency
medical services, and law enforcement. This communication is
distributed in a 2-MHz band at around 853 MHz. This RF data analytics application is based on pairing the spectrum data with metadata provided by the RF client, such as location, duration, channel,
and time instant, and can use metadata available on the Internet
about the P25 radio system in Fort Wayne, Indiana. As a result,
it is capable of identifying emergency situations in the local area.
Another example of a descriptive RF data analytics application
is spectrum or channel occupancy. Using the architecture shown in
Figure 1, novel RF data analytics services and applications can be
introduced, helping to determine how much wireless systems are
using the spectrum that they have been given or how effective is
the utilization of a given frequency band. This is a valuable service
48

for cellular service providers. We illustrate this with an example
of the practical estimation of the load of the downlink (DL) LTE
channels. The estimation is done indirectly, on the basis of measurements and analysis of the change of the power of the resource
blocks (RBs).
An RB is the basic frequency-time unit in LTE. It occupies a
bandwidth of 180 kHz (12 subchannels) and has a time duration
of 0.5 ms (six or seven orthogonal frequency-division multiplexing symbols). Depending on the traffic requirements, the mobile
terminals are assigned a given number of RBs. One of the main
tasks of the power control mechanism in LTE is to maintain the
SINR within certain limits, compensating the path loss and different types of interference. The power of the allocated RBs for every
mobile terminal depends on the distance between the base station
(BS; also known as eNodeB) and the terminal. Consequently, an
indirect estimation of the load in an LTE cell is possible through
analysis and clusterization of the measured and stored RF data
[18].
The results of the measurements and analysis of the power
spectrum variation for one LTE channel for a cell in the city of Sofia, Bulgaria, within 1 day are shown in Figures 3-6. The channel
frequency is 1,827.5 MHz, the channel bandwidth is 15 MHz, and
the number of RBs is 75. For this particular example, the settings
of the SDR-based RF client are controlled appropriately so that a
measurement result of the average power spectrum in every 0.1
second, with resolution bandwidth = 15 KHz (one LTE subchannel), could be obtained.

IEEE A&E SYSTEMS MAGAZINE

MAY - JUNE 2018

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine May 2018