Signal Processing - March 2017 - 52

cellular networks can easily be overloaded by CAM distribution, especially when robust MCSs with low spectral efficiency are employed. In such overload situations,
reliable distribution of C-ITS messages cannot be guaranteed and, thus, they must be avoided. One possible method
could be to adapt CAM generation periodicity according
to user mobility; i.e., at low mobility, less frequent vehicle
status updates are required than at high mobility. Also, the
spatial area in which CAMs are distributed, i.e., the number of base stations involved in CAM exchange, should be
carefully chosen and potentially adapted according to traffic load.

Future enhancements and challenges
In this section, we present promising improvements for
wireless vehicular communications in cellular networks and
highlight associated signal processing challenges. Our focus
is on two topics: in the section "CSI Feedback Enhancement," we discuss CSI feedback enhancements to improve
throughput performance of mobile users, employing channel
predictive approaches to partly compensate for feedback
delay and temporal channel variation. As we demonstrated
in the section "Dual Connectivity-Enhanced V2I Transmission," inaccurate and outdated CSIT can be the limiting factor for the achievable efficiency and reliability of wireless
data transmission. Hence, enhancing CSI feedback for highmobility situations is an essential prerequisite of wireless
vehicular communications. In the section "Multicast
Enhancements," we consider techniques for efficient multicast transmission, putting scope not only on advanced multiple-input, multiple-output (MIMO) and coordinated
multipoint (CoMP) transmission schemes, but also highlighting issues related to scheduling and resource assignment.
Such techniques are important for vehicular scenarios to
avoid the network overload problem discussed in the section
"MBSFN-Based V2V Communication" by improving the
capacity of multicast transmissions. The presented methods,
again, require accurate CSIT for beamformer/precoder calculation and, hence, rely on efficient CSI feedback techniques. Finally, we present further important topics for
progressing wireless vehicular communications in the section "Further Research Topics to Enhance Wireless Vehicular Communications."

TransmitSignal
Adaptation

Transmitter

CSI feedback enhancement
CSI is useful for achieving the highest performance in multiple antenna wireless communications by enabling transmission rate adaptation, adaptive MIMO beamforming, and
spatial multiplexing. While CSI at the receiver (CSIR) is comparatively easy to obtain through pilot-aided channel estimation, CSIT in frequency-division duplex systems is only
available if the receiver provides feedback information to the
transmitter; Figure 4 illustrates the situation. (In time-division
duplex systems, channel reciprocity can be exploited to estimate the CSI on both sides of the link.) Since the feedback
link between the transmitter and receiver is of limited capacity, quantization of CSI is necessary to enable signaling with a
finite number of bits; this is known as limited feedback operation. In general, we distinguish between providing explicit
and implicit CSI feedback. In the former case, the channel, as
estimated by the receiver, is directly quantized and fed back,
while in the latter case derived information, such as the optimal MIMO precoder and the supported transmission rate, is
provided as side information. In LTE, the second type of
feedback information is currently supported with the standard-defined CQI, the precoding matrix indicator (PMI), and
the rank indicator (RI). More advanced multiuser MIMO and
CoMP schemes, however, mostly rely on explicit CSIT.

Implicit CSI feedback
In the section "Dual Connectivity-Enhanced V2I Transmission," we observed that CQI feedback, as employed for transmission rate adaptation in LTE, is sensitive to feedback delay
in high-mobility scenarios, since outdated CQI feedback can
cause signal outages, thereby reducing transmission efficiency and increasing latency. A similar behavior exists for the
MIMO-specific feedback information PMI and RI, yet with
reduced sensitivity since their mismatch does not directly
cause signal outages.
To reduce the impact of feedback delay, either the transmitter or the receiver should attempt to compensate for
it. At the receiver side, one natural approach is to apply
channel prediction. The simplest method for OFDM-based
LTE is a subcarrier-wise linear extrapolation of the channel transfer function between pairs of transmit and receive
antennas. If channel statistics are known, the prediction performance can be improved by linear minimum mean squared

Receiver

Channel

Feedback Link

Channel
Estimation

Q
Delay

CSI Quantization

Figure 4. An illustration of limited feedback operation in wireless communications assuming error-free but delayed feedback.
52

IEEE SIgnal ProcESSIng MagazInE

March 2017

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