Signal Processing - March 2017 - 58

implies that single users occupy resources at
In vehicular environments, especially
A big challenge from
multiple transmission points, causing a
on highways, the wireless channel behaves
a signal processing
reduction of the overall network capacity.
markedly different as compared to other
perspective is efficient
To maintain the efficiency of the network, it
common situations in mobile communicascheduling and
is therefore necessary to dynamically
tions. More specifically, the propagation
coordination of multicast
decide whether the gain of utilizing multiis characterized by shadowing through
transmissions.
connectivity for a user outweighs the overother vehicles, high Doppler shifts with
head caused on network capacity. Hence,
often sparse Doppler spectrum due to few
multiconnectivity comes at the cost of requiring a sophisticated
dominant scatterers (e.g., road signs, highway overpasses),
coordination of multiple transmission points, potentially
and possibly inherent nonstationarity of the channel statistics.
implying a substantial backhaul signaling overhead. Such
Such effects can cause substantial performance degradation
enhanced self-organizing network features can, e.g., be
of common least squares and MMSE channel estimators if
realized by cloud radio access network architectures. In [29],
they are not properly considered in the algorithmic design. In
the authors investigate the tradeoff between radio-link failures
[23], the impact of the shape of the Doppler spectrum on the
and user throughput in dependence of the size of the set of active
performance of channel estimation at the receiver is discussed
multiconnectivity transmission points, showing that both the cellin more detail. Similarly, timing and carrier synchronization
edge user throughput as well as the mobility performance can be
in vehicular scenarios with few dominant scatterers of simiimproved simultaneously.
lar strength that are strongly delay- and Doppler-shifted with
respect to each other can be challenging.

FD MIMO beamforming for high mobility

ICI mitigation
OFDM multicarrier modulation, as employed in LTE, suffers at high mobility from intercarrier interference (ICI) due
to the Doppler spread of the transmit signal. This effect,
however, can, to a large extent, be mitigated through iterative ICI estimation and cancellation at the receivers. In [27],
the authors propose ICI mitigation algorithms that enable
achieving the performance of interference-free transmission.
Alternatively, optimal pulse shaping at the transmitter, as
currently pushed by many research groups and companies
for 5G mobile communications, can be applied to maximize
achievable data rates by trading off residual ICI/ISI for spectral efficiency [28]. Such an approach appears especially
interesting when waveform parameters, such as subcarrier
spacing, prototype pulse shape, and TTI length, can be
adapted to the time/frequency-dispersion characteristics of the
channel. A major challenge then is to optimize sets of compatible parameters that enable efficiently serving users with
strongly different channel properties in parallel, e.g., static and
highly mobile users. Novel multicarrier transmit waveforms,
employing filter banks [filter-bank multicarrier modulation
(FBMC)] or subband filters (universal filtered multicarrier/
OFDM), support the necessary flexibility to adjust waveform
parameters over subbands and, thus, provide the basis for
channel-adaptive modulation.

Multiconnectivity
To enhance the robustness of the wireless transmission link, so
as to support highly reliable communication, macro diversity
can be exploited by extending the dual connectivity concept of
LTE to a multitude of radio network access points. Maintaining
multiple parallel connections to several macro base stations
and/or small cells promises increased data throughput over
multiple parallel data streams, improved reliability due to a
reduced outage probability and enhanced robustness with
respect to mobility, since hard handovers can be avoided. Yet, it
58

FD MIMO refers to wireless transmission systems that support active two-dimensional antenna arrays with a large number of antenna elements. This enables high-resolution adaptive
beamforming in both the elevation and the azimuth domain,
to achieve space-division multiple access gains through spatial separation of users. Within LTE standardization, work is
ongoing to implement FD MIMO within Release 14. Currently, hybrid transceiver architectures are of interest, where part
of the signal processing is performed in base band and part in
the analog domain to limit the number of required radio-frequency chains. Analog beamforming approaches are mostly
based on signal azimuth and elevation arrival/departure
angles. To implement such schemes at high mobility, it is necessary to account for uncertainty in the signal arrival/departure angles due to user movement and estimation errors. In
[30], the authors propose a corresponding robust beamformer
optimization problem and demonstrate improved robustness
at high mobility.

Conclusions
Vehicular communications is an integral part of innovative
transport telematics systems for traffic management and active
road safety. It plays a key role in making public and private
transportation faster, more reliable, more efficient, and safer.
Realizing the necessary information exchange among roadside
infrastructure and vehicles efficiently and reliably can be challenging. In that respect, cellular networks can provide valuable
support to dedicated vehicular communication systems, since
today's cellular base stations are almost ubiquitously accessible
and supply high bandwidth wireless connectivity. In this article,
we have surveyed ongoing efforts and developments within the
3GPP to implement vehicular communications over LTE. We
have discussed research challenges associated with wireless
communications at high mobility, and we have provided an overview of promising signal processing techniques to tackle important hurdles. Even though significant progress in enhancing

IEEE SIgnal ProcESSIng MagazInE

March 2017

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