ALSABAAN et al.: VEHICULAR NETWORKS FOR A GREENER ENVIRONMENT: A SURVEY 1375
Fig. 2. The impact of speed and acceleration on vehicle CO
2
emissions.
acceleration levels as shown in the following equation.
ln(MOE
e
)=
3
i=0
3
j=0
(L
e
i,j
× s
i
× a
j
), for a 0
3
i=0
3
j=0
(M
e
i,j
× s
i
× a
j
), for a<0
(1)
where s is the instantaneous speed (km/h), a is th e instan-
taneous acceleration (km/h/s), the measures of effectiveness
(MOE) is the instantaneous fuel consumption or emission
rate (L/s or mg/s), e is an index denoting fuel consumption or
emission type (e.g., CO
2
,HC,andNO
x
), and L
e
i,j
and M
e
i,j
represent model regression coefficients for MOE
e
at speed
power i and acceleration power j for positive and negative
accelerations, respectively. The function ln() is the natural
logarithm function.
Example of using the VT-Micro Model: Sample model
coefficients for estimating CO
2
emission for a composite
vehicle are introduced in Table II. The vehicle was derived
as an average across eight light duty vehicles ( LDVs). The
required input parameters of the model are s, a, L
e
i,j
, and M
e
i,j
.
In this example, the effect of speed and acceleration on
the vehicle CO
2
emissions is studied. To study the impact of
vehicle speed, the vehicle acceleration is set to a constant value
(say 0 kph/s). After that, the CO
2
emissions are computed
with different values of speed using the VT-Micro model.
Figure 2(a) shows that CO
2
emissions increase with high
speeds. Similarly, to show the effect of vehicle acceleration,
the vehicle speed is set to 30 kph. Then, the CO
2
emissions
are calculated with different values of accelerations. Figure
2(b) demonstrates that negative accelerations do not affect the
CO
2
emissions much because vehicles do not exert power in
negative accelerations. On the other hand, the amount of CO
2
emissions increases with high acceleration.
B. Electricity Consumption
In wireless communications, the importance of the power
consumption level is considered system dependent. For exam-
ple, in a cellular system, the base station has no restricted
power constraints, whereas a cellular phone does. Instead,
power control is d efinitely used to control interference between
nodes. The ad hoc and wireless sensor networks comprise a
special class of networks that requires power saving. Tremen-
dous work has been proposed for this class of network in the
literature [50] [5 1]. However, such networks are not applicable
TABLE II
VT-M
ICRO MODEL COEFFICIENTS FOR ESTIMATING CO
2
EMISSION
Coefficients s
0
s
1
s
2
s
3
Positive a
a
0
6.91 2.75E-02 -2.07E-04 9.80E-07
a
1
0.22 9.68E-03 -1.01E-04 3.66E-07
a
2
2.35E-04 -1.75E-03 1.97E-05 -1.08E-07
a
3
-3.64E-04 8.35E-05 -1.02E-06 8.50E-09
Negative a
a
0
6.91 2.84E-02 -2.27E-04 1.11E-06
a
1
-3.20E-02 8.53E-03 -6.59E-05 3.20E-07
a
2
-9.17E-03 1.15E-03 -1.29E-05 7.56E-08
a
3
-2.89E-04 -3.06E-06 -2.68E-07 2.95E-09
in vehicular networks as the power consumption has not been
considered an issue from a communication perspective.
One case in which the electricity power consumption is very
crucial measure is the case o f EVs. In the past decade, EVs
and Hybrid EVs have been introduced into th e market by some
car manufacturers [19] [20] [21], and their penetration rate is
increasing consistently worldwide [52]. Consequently, forming
a network of EVs would require higher consideration of the
issue of the power consumed in the designed communication
protocols. Applying protocols similar to the protocols that
were applied for fuel consumption, however, would yield to
promising results in the overall power consumption of the
vehicles. To the best of our knowledge, neither research nor
implementation of a system that enhances power consumption
in EVs has yet used VANETs.
One thing that can be optimized regard ing power is the
placement of road-side units (RSUs) [53]. RSUs are used to
enhance the communication b etween vehicles in a centralized,
distributed or hybrid architectures. An RSU can be optimally
placed in order to enh a nce probability of messag e delivery,
message latency, and communication signal range. In [53],
an RSU is optimally deployed in order to save energy. In this
case, similar optimization methodology from cellular networks
can be applied.
C. Communication Cost
In wireless communications, energy consumption is con-
sidered an important performance metric. Energy-efficient
wireless protocols were extensively studied in the literature