Signal Processing - September 2017 - 113

new paradigm to overcome the limitations of GNSS-based
navigation has emerged over the past decade [5]. Examples of
signals of opportunity include AM/FM radio [6], [7], iridium
satellites [8], [9], cellular [10], [11], digital television [12], [13],
and Wi-Fi [14], [15]. Figure 1 illustrates various radio navigation transmitters over the past century.

Cellular-based navigation
Among the different signals of opportunity types, cellular signals are particularly attractive due to their following qualities:
■ Abundance: cellular base transceiver stations (BTSs) are
plentiful due to the ubiquity of cellular and smartphones.
■ Geometric diversity: the cell configuration by construction
yields favorable BTS geometry-unlike certain terrestrial
transmitters, which tend to be colocated, e.g., digital
television.
■ High carrier frequency: cellular carrier frequency ranges
800-1,900 MHz, which yields precise carrier phase navigation observables.
■ Large bandwidth: cellular signals have a bandwidth up to
20 MHz (as discussed in the section "LTE"), which yields
accurate time-of-arrival (TOA) estimation.

High received power: cellular signals are often available
and usable in GNSS-challenged environments-the received
carrier-to-noise ratio C/N 0 from nearby cellular BTSs is
more than 20 dB-Hz higher than GPS SVs.
Besides the aforementioned advantages, there is no deployment cost associated with using cellular signals for positioning
and navigation-the signals are practically free to use. Specifically, the user equipment (UE) could "eavesdrop" on the transmitted cellular signals without communicating with the BTS, extract
necessary positioning and timing information from received
signals, and calculate the navigation solution locally. While
other navigation approaches requiring two-way communication
between the UE and BTS (i.e., network based) exist, this article
focuses on explaining how UE-based navigation can be achieved
with cellular CDMA and LTE signals, presenting receiver architectures that are suitable for software-based implementation (see
"Software-Defined Receivers for Navigation") along with ground
and aerial vehicle navigation results achieved with these receivers.
■

CDMA
Cellular CDMA systems employ orthogonal and maximallength sequences to enable multiplexing over the same channel.

Software-Defined Receivers for Navigation
Software-defined receivers (SDRs) offer many advantages
over their hardware-based counterparts, such as 1) flexibility: designs are hardware-independent; 2) modularity:
different functions can be implemented independently;
and 3) upgradability: minimal changes are needed to
improve designs. Signal processing algorithms in SDRs
are typically implemented on general-purpose digital signal processors (DSPs), with only minimal dedicated hardware components to the radio-frequency (RF) front end.
Traditionally, baseband operations in GNSS receivers
have been implemented using dedicated hardware due to
cost, power, and speed. Until recently, GNSS SDRs were
limited to postprocessing applications operating on raw
samples recorded from an RF front end. However, with
modern DSPs, real-time GNSS SDRs are becoming more
prevalent [16], [17]. Such SDRs are typically implemented
in high-level, textual-based languages, such as C/C++.
Processor-specific optimization techniques are often utilized for computationally expensive baseband operations.
Graphical programming languages, such as LabVIEW
and Simulink, are attractive choices for implementing
navigation SDRs, whether for GNSS or signals of opportunity, for a number of reasons. First, while the optimized
C/C++ SDR implementations are often portable and reusable on multicore DSPs, the optimizations required for
each processor in real-time applications could slow
development and introduce platform-specific errors.

Graphical programming languages offer tools that often
generate optimized implementations for multiple platforms-desktop, DSP, and field-programmable gate
arrays (FPGAs)-without code modifications. Second,
navigation SDRs are conceptualized as block diagrams,
enabling a one-to-one correspondence between the
architectural conceptualization and software implementation. Third, graphical optimized routines are abundant,
which could be readily exploitable by navigation SDR
designers. Fourth, data-flow-based graphical implementations are easier to understand and debug, and they offer
rapid access to all internal signals. Finally, graphical
tools provide attractive graphical user interfaces, allowing designers to develop interactive panels that have the
look and feel of hardware-based navigation receivers.
Graphical implementations of GNSS SDRs [18], [19]
and cellular SDRs [20], [21] have been the subject of a
number of recent publications.
While SDRs offer many advantages over hardwarebased receivers, they suffer from a number of shortcomings: larger size and weight, increased power
consumption, and higher cost. This is due to the fact that,
unlike hardware-based receivers, SDRs are not optimized
for a particular application and they could utilize high-level, general-purpose scripting tools to translate a graphical
SDR design into code that gets deployed onto DSPs and
FPGAs.

IEEE SIGNAL PROCESSING MAGAZINE

September 2017

113

Table of Contents for the Digital Edition of Signal Processing - September 2017