key: cord-0453792-5qf9opgj authors: Du, Jianhe; Ahn, Kyoungho; Farag, Mohamed; Rakha, Hesham title: Environmental and Safety Impacts of Vehicle-to-Everything Enabled Applications: A Review of State-of-the-Art Studies date: 2021-12-07 journal: nan DOI: nan sha: 50c93a96df2f87e455cfc9988de16274028af84c doc_id: 453792 cord_uid: 5qf9opgj With the rapid development of communication technology, connected vehicles (CV) have the potential, through the sharing of data, to enhance vehicle safety and reduce vehicle energy consumption and emissions. Numerous research efforts are quantifying the impacts of CV applications, assuming instant and accurate communication among vehicles, devices, pedestrians, infrastructure, the network, the cloud, and the grid, collectively known as V2X (vehicle-to-everything). The use of cellular vehicle-to-everything (C-V2X), to share data is emerging as an efficient means to achieve this objective. C-V2X releases 14 and 15 utilize the 4G LTE technology and release 16 utilizes the new 5G new radio (NR) technology. C-V2X can function without network infrastructure coverage and has a better communication range, improved latency, and greater data rates compared to older technologies. Such highly efficient interchange of information among all participating parts in a CV environment will not only provide timely data to enhance the capacity of the transportation system but can also be used to develop applications that enhance vehicle safety and minimize negative environmental impacts. However, before the full benefits of CV can be achieved, there is a need to thoroughly investigate the effectiveness, strengths, and weaknesses of different CV applications, the communication protocols, the varied results with different CV market penetration rates (MPRs), the interaction of CVs and human driven vehicles, the integration of multiple applications, and the errors and latencies associated with data communication. This paper reviews existing literature on the environmental, mobility and safety impacts of CV applications, identifies the gaps in our current research of CVs and recommends future research directions. he transportation system is growing rapidly, with the current number of registered vehicles at 276 million, a total of 4.17 million miles of highways, and 3.23 trillion vehicle miles of travel (VMT) 1 . This gigantic system operates with a significant impact on the environment and traffic safety due to the huge number of vehicles and associated delays in traffic. According to Forbes, Americans lost nearly 99 hours, or an average of $1,377 per person in 2019 due to traffic crashes 2 . In addition, at least 38,800 people were killed in motor vehicle collisions 3 . The majority of these crashes were caused by human errors, such as distraction, driver inexperience, drowsiness, or speeding, all of which may be mostly avoided if warnings could be provided to drivers ahead of time. A report released by INRIX found that the economic impact of traffic congestion is both broad and complicated. Congestion cost New York $11 billion dollars, Los Angeles $8.2 billion dollars, and Chicago $7.6 billion dollars in 2019 4 . Safety and efficiency are not the only concerns regarding the traffic system; environmental impact has also become a major issue following the realization that climate change is caused by human activity. The U.S. Environmental Protection Agency (EPA) published the Inventory of U.S. Greenhouse Gas Emissions and Sinks in April 2021. This annual report summarizes the latest information on U.S. greenhouse gas emission trends from 1990 through 2019 [1] . The report identified the primary sources of greenhouse gas emissions in the U.S. and demonstrated that the transportation sector generated the largest share of greenhouse gas emissions, at 29%. The report also mentions that petroleum-based transportation fuel, including gasoline and diesel, is responsible Hesham A. Rakha is with the Virginia Tech Transportation Institute; 3500 Transportation Research Plaza, Blacksburg, VA 24061 USA (e-mail: hrakha@vt.edu) > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < This section summarizes the various efforts reported in the literature quantifying the network-wide impacts of CV applications. As summarized in Table I , we found a limited number of studies focusing on network-wide impacts of CV applications. All studies entailed evaluating these applications in a virtual traffic simulation environment given that actual field implementations are cost prohibitive. A research study by Ahn et al. developed and evaluated an Eco-Cooperative Automated Control System (Eco-CAC) system that integrates vehicle control strategies with connected automated vehicle (CAV) applications. The developed system was tested on a large-scale network in downtown Los Angeles, CA, considering a combination of internal combustion engine vehicles (ICEVs), hybrid electric vehicles (HEVs), and battery electric vehicles (BEVs) in a microscopic traffic simulation environment for three different demand levels: no congestion, mild congestion, and heavy congestion. The Eco-CAC system includes an Eco-router, a speed harmonization (SPD-HARM) controller, an Eco-CACC-I controller (also known as green light optimized speed advisory [GLOSA]), and an Eco-CACC-U controller (also known as vehicle string control or platooning). The study found that the Eco-CAC system reduces fuel consumption by up to 16.8%, energy consumption by up to 36 .9%, travel time by up to 21.7%, total delay by up to 43.1%, stopped delay by up to 68.7%, and CO2 emissions by up to 16.8%. The study also found that different vehicle types generate different benefits. For example, the Eco-CAC system reduces fuel consumption, travel time, total delay, stopped delay, and CO2 emissions for ICEVs in heavily congested conditions. The benefits of the controllers mostly increase as the CAV MPR increases. However, the controllers negatively affect ICEV fuel consumption, travel time, total delay, stopped delay, and CO2 emissions for no or mild congestion. For BEVs, the study found that the Eco-CAC system improves the energy consumption but negatively affects travel time, total delay, and stopped delay for all congestion levels. The study also tested the Eco-CAC system considering both the current and a future vehicle composition on the LA network and found that the Eco-CAC system effectively reduces fuel and energy consumption, travel time, total delay, and stopped delay in heavily congested conditions for both vehicle compositions. However, the study found that different vehicle compositions produced different results. Specifically, the maximum energy consumption savings of BEVs (36.9% saving) for the current vehicle composition were observed at a 10% CAV MPR in mild congestion, while the maximum savings for the future vehicle composition (35.5%) were observed at a 50% CAV MPR with no congestion. The study found that the developed Eco-CAC system effectively reduces the fuel and energy consumption, travel time, total delay, stopped delay, and CO2 emissions for ICEVs, BEVs, and HEVs for specific scenarios. Additionally, the effectiveness of the Eco-CAC system and the controllers depends on the traffic conditions, including the level of traffic congestion, the network configuration, the CAV MPR, and the vehicle composition [4] . The Netherlands Organization for Applied Science (TNO) investigated various CV use cases. The study reviewed available literature on the environmental impact of CV use cases and the study performed simulation studies using a microscopic emission calculation tool (EnViVer) to quantify the potential environmental impacts of CV applications. EnViVer was developed using the VISSIM traffic microsimulation software and VERSIT+, a vehicle emission simulation tool. The identified mechanisms that were taken into consideration include reduction of trips; reduction and departure time shift; mode shift, reduction of vehicle dynamics, and the powertrain operation. The results showed that an ecodriving technique that reduces vehicle stops is very beneficial, generating a CO2 reduction in the range of 13-45%. An ecodriving technique that reduces deceleration and acceleration shows benefits as well (3-7% improvement) . The study found that cooperative adaptive cruise control (CACC) reduces vehicle emissions by 6% compared to ACC [5] . Olia et al. studied the potential impacts of CV applications using PARAMICS. The simulation testbed was a network in the north of Toronto. The simulation results showed that CVs were able to navigate through multiple routing options and therefore could save up to 58% in travel time at an MPR of 50% on certain corridors. As for improving safety indicators, increasing the MPR of CVs can improve the safety index (the probability of incidents) by up to 45%. The Comprehensive Modal Emission Model (CMEM) was used to estimate emission factors. A reduction of 30% in CO2 emissions was identified [6] . In a study by Vahidi et al. , the authors reviewed the energy saving potential of CAVs based on first principles of motion and optimal control theory. The study found that connectivity to other vehicles and infrastructure allows better anticipation of upcoming events, such as hills, curves, slow traffic, state of traffic signals, and movement of neighboring vehicles. The review paper concluded that automation allows vehicles to adjust their motion more precisely in anticipation of upcoming events to save energy, and that cooperative driving can further increase a group of vehicles' energy efficiency by allowing them to move in a coordinated manner. The study concluded that CAVs' energy efficient movement could have a harmonizing effect on mixed traffic, leading to additional energy savings for neighboring vehicles [7] . A study by Rahman et al. investigated the safety impact of CVs and lower-level automation (CVLLA) using simulation. Two features were tested as CVLLA: automated braking and lane keeping assistance. Segment and intersection crash risks were estimated using surrogate safety assessment modeling techniques. The results showed that both CV and CVLLA technology significantly reduced conflict frequency. Higher MPRs of CVs generate higher benefits, where the maximum improvement was found to be at 100% MPR. However, at least 40% MPR is needed to achieve the safety benefits of reduced intersection crash risks and 30% is needed for reduced segment crash risks [8] . > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < A study by Ahn et al. developed a multi-objective ecorouting algorithm (eco-and travel time-optimum routing) for BEVs and ICEVs in a CV environment and investigated the network-wide impacts of the multi-objective Nash optimum (user equilibrium) traffic assignment on a large-scale network. Unlike ICEVs, BEVs are more energy efficient on low-speed arterial trips compared to highway trips. The authors demonstrated that different energy consumption patterns require different eco-routing strategies for ICEVs and BEVs. This study found that single objective eco-routing could significantly reduce BEVs' energy consumption but also significantly increased their average travel time and thus the authors developed a multi-objective routing model (eco-and travel time-routing) to improve both energy and travel time measures. The simulation study found that multi-objective routing could reduce BEVs' energy consumption by 13.5%, 14.2%, 12.9%, and 10.7%, as well as ICEVs' fuel consumption by 0.1%, 4.3%, 3.4%, and 10.6% for "not congested, "slightly congested," "moderately congested," and "highly congested" conditions, respectively. The study also found that multiobjective user equilibrium routing reduced the average vehicle travel time by up to 10.1% compared to the standard user equilibrium traffic assignment for highly congested conditions, producing a solution closer to the system optimum traffic assignment. The study concluded that the multi-objective ecorouting strategy can reduce vehicle fuel/energy consumption effectively with minimum impacts on travel times for both BEVs and ICEVs [9] . In a study conducted by Abdelghaffar et al., a dynamic freeway speed controller based on sliding mode theory was developed and tested. The advantages of the sliding mode control are its simple design, global stability, and robustness that can address the discontinuity of the fundamental diagram due to the capacity drop. A SPD-HARM controller was developed in their study to dynamically identify bottlenecks and regulate the speeds of CVs. A decentralized phase split/cycle length controller was used to optimize all traffic signals in the network. The results showed that average travel time was reduced by 12%, total delay was reduced by 21%, and CO2 emissions were reduced by 3.3%. In addition, the authors' results showed that freeways benefit from the controller more than arterials [10] . In summary, we found that the benefits of CV applications were highly dependent on the application type, network settings, MPRs, and test algorithms. There were a limited number of studies focusing on the benefits at a network/system level. More studies concentrated on either freeways only or signalized intersection corridors only. The following two sections will be used to illustrate these two categories individually. Numerous studies investigated and quantified the impacts of CV-enabled applications on freeway sections, as summarized in Table II . Some of these studies concentrated on car-following and lane-changing behavior using a basic multi-lane freeway segment without any other complex roadway configurations. Some concentrated on studying the disturbance of merging or diverging traffic on a mainline freeway by analyzing freeway sections with on-and off-ramps. For example, in a study by Jang et al., the authors analyzed crash risks and estimated the safety benefits of the forward hazardous situation warning information presented by a Connected Intelligence Transport System (C-ITS) pre-deployment project for Korean freeways. C-ITS is composed of on-board units, roadside units (RSUs), and a traffic management center. The crash potential index (CPI) was adopted to quantify the crash benefits of CVs. A total of 700 CVs, including 400 buses, 264 trucks, and 36 sports utility vehicles, were instrumented with data acquisition systems and V2X communication devices. The study found that the average speed decreased by 10.2% and the time-to-collision (TTC) increased by 5.3% when warning information was provided in a CV environment. In addition, the achievable reduction in the CPI was approximately 20.7% due to the provision of warning information [11] . To study the effectiveness and network communication efficiency of connected (V2V and V2I) vehicular technologies in alerting motorists when they are approaching a hazardous zone, such as using CV technology to recommend a proper speed for travelers going through a low visibility area, Outay et al. compared the performance of V2V and V2I communications in an extensive computer simulation experiment. The authors adapted the iTetris platform for various scenarios-a baseline scenario, a deactivated alert scenario, a V2V cooperative alert scenario, and a V2I alert scenario. The authors also explored, via simulations, whether Cooperative Hazard Awareness and Avoidance systems, based on V2V and V2I communications can potentially contribute towards eco-driving by reducing CO2 emissions. The results revealed that the alerting system based on V2I communication yields better message reception rates and better safety efficiency. The results also showed that the Cooperative Hazard Awareness and Avoidance system can reduce CO2 emissions using SPD-HARM [12] . To explore the impacts of CACC on vehicle fuel efficiency in mixed traffic, Liu et al. conducted analyses at a freeway merge bottleneck. Their results showed that CACC string operation resulted in a maximum of 20% reduction in energy consumption compared to the human driver only case. At a 100% MPR, CACC-equipped vehicles consumed 50% less fuel than ACC vehicles without V2V communication and cooperation. At lower CACC MPRs, the vehicle fuel efficiency could be improved via use of a dedicated CACC lane or by implementing wireless connectivity on the manually driven vehicles. In the case of a CACC MPR of 40%, those strategies brought about a 15% to 19% capacity increase without decreasing vehicle fuel efficiency. The authors' results imply the importance of incorporating the V2V cooperation component into an automated speed control system and highlight the necessity of deploying CACC-specific operation strategies at lower CACC MPRs [13] . A study by Li et al. used simulation coding in MATLAB to test the effectiveness of reducing rear-end collisions in an I2V > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < system integrating variable speed limit (VSL) and ACC. The results showed that both the surrogate crash risk measurestime exposed time to collision (TET) and time integrated time to collision TIT-were reduced. VSL-only and ACC-only methods had a positive impact in reducing the TET and TIT values (reduced by 53.0 and 58.6% and 59.0 and 65.3%, respectively). The I2V system combined the advantages of both ACC and VSL to achieve the most safety benefits (reduced by 71.5 and 77.3%, respectively) [14] . A study by Rahman and Abdel-Aty attempted to evaluate longitudinal safety of CV platoons by comparing the implementation of managed-lane and all-lane CV platoons (for the same MPR) over a non-CV scenario. A high-level control algorithm of CVs in a managed-lane was proposed to form platoons with three joining strategies: rear join, front join, and cut-in join. Five surrogate safety measures-standard deviation of speed, TET, TIT, time exposed rear-end crash risk index, and sideswipe crash risk-were utilized as indicators for safety evaluation. The results showed that with CVs, the safety in the studied expressway significantly improved. Managed lane control produced better improvements compared to all-lane control for the same MPR [15] . Jin et al. used micro simulation to evaluate the mobility and environmental benefits of a CV application that involved a realtime optimal lane selection algorithm. On average, the travel time was reduced up to 3.8% and the fuel consumption was reduced by 2.2%. In addition, the reduction in emissions of criteria pollutants, such as CO, HC, NOx and PM2.5, ranged from 1% to 19%, depending on the different congestion levels of the roadway segment [16] . A study conducted by Guériau integrated multi-agent cooperative traffic modeling into the MovSim, a traffic simulator, to model complex interactions between cooperative vehicles and potentially among vehicles and infrastructure. The authors discussed some potentialities of C-ITS for traffic management with the methodological issues following the expansion of such systems. The operational goal of this work was to develop a decision-making tool validated in simulation conditions tailored to cooperative strategies. The model involved three layers that couple different dynamics to take into account information reliability while limiting traffic disturbances, and hence homogenize traffic flow. The results showed that at a 40% to 50% MPR, the connected environment had benefits in terms of homogenization and was able to avoid congestion for the merging traffic from on ramps and a scenario where a lane was closed [17]. Monteil et al. constructed a multi-agent framework that inherits knowledge from traffic theory. Their goal was to set up the modeling bricks of a cooperative auto-adaptive system with the assumption that modeling and technology uncertainties should affect the traffic physics and communications. A threelayer framework with physical, communication, and trust layers was used to achieve traffic flow homogenization. By modeling the trust as a function of distances, along with a communication layer and a physical layer, the authors conducted simulations using two lanes of freeway traffic with an entrance flow distribution from US-101 sample data. The results showed that the operation could decrease the speed variance and therefore the likeliness that traffic would fall into local congestion phenomena. The cooperation can also limit the impact of aggressive drivers (who oppose a global gain in acceleration when making a lane change). An increasing MPR reduces the impact of aggressive lane-changing behaviors, as it increases the stability domains and homogenizes traffic and reduces propagation of perturbations [18] . A study conducted by Liu et al. examined the impact of CACC string operations on vehicle speed and fuel economy on a 13-mile section of the SR-99 corridor near Sacramento, CA. The authors used simulation to evaluate the performance of the corridor under various CACC MPR scenarios and traffic demand levels. The CACC string operation was also analyzed when a vehicle awareness device (VAD) and CACC managed lane strategies were implemented. The results revealed that the average vehicle speeds increased by 70% when the CACC MPR increased from 0% to 100%. The highest average fuel economy, expressed in miles per gallon (MPG), was achieved under the 50% CACC scenario with an MPG at 27. This was 10% higher than the baseline scenario. However, when the CACC MPR was 50% or higher, the vehicle fuel efficiency only had minor increases. When the CACC MPR reached 100%, the corridor allowed 30% more traffic to enter the network without experiencing reduced average speeds. Results also indicated that the VAD strategy increased the speed by 8% when the CACC MPR was 20% or 40%, while there was a minor decrease in the fuel economy. The managed lane strategy decreased the corridor performance when implemented alone [19] . A study conducted by Ard et al. demonstrated the effectiveness of an anticipative car-following algorithm in reducing energy use of gasoline engine and electric CAVs. Without sacrificing safety and traffic flow, the authors implemented a vehicle-in-the-loop testing environment. Experimental CAVs were driven on a track, interacting with surrounding virtual traffic in real-time. The authors explored the energy savings in microsimulations. Model predictive control handled high level velocity planning and benefited from communicated intentions of a preceding CAV or estimated probable motion of a preceding human driven vehicle. A combination of classical feedback control and data-driven nonlinear feedforward control of pedals achieved acceleration tracking at the low level. The controllers were implemented in a robot operating system and energy was measured via calibrated OBD-II readings. The authors reported energy savings of 30% [20] . > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < A model-free approach was proposed by Weng et al. to incorporate the differences of varied vehicle types and combine the goal of fuel economy and mobility platooning. A utility function was formulated using the Nelder-Mead approach. This approach relies on the communication of instantaneous speed and fuel consumption between vehicles in the platoon with mixed vehicle types. The simulation and experimental results showed that this method was effective in increasing the objective function, with less fuel consumption and higher mobility. However, due to the data-driven nature, convergence required some time [21] . A study conducted by Kamal et al. constructed a model predictive control (MPC) framework to efficiently drive a vehicle on multi-lane roads by enhancing the vehicle's capability in lane change and speed adjustment. MPC uses present state information to predict the future behavior through the explicit use of a process model. The results showed that the MPC generated optimal acceleration of the vehicle and the optimal timing to move to the next lane if long term performance gain was anticipated by predicting surrounding traffic [22] . Wang et al. proposed a strategy called Altruistic Cooperative Driving (ACD), where vehicles that are causing congestion should yield the right of way to other vehicles by slowing down or changing lanes. By defining the vehicle driving conditions into maximum, deadlock, and free-run, the strategy changes the deadlock situation into a free-run or maximum to improve overall efficiency. A simulator was generated using Java for a section of multi-lane road. The results showed that the ACD strategy achieved higher speed efficiency with up to 15% improvement and can perform cooperative driving to resolve deadlock conditions in a timely fashion [23] . A study by Ghiasi et al. designed a CAV-based trajectorysmoothing method to control CAVs upstream of a bottleneck to harmonize traffic, improve fuel-efficiency, and reduce environmental impacts. Four steps were adopted: information update, trajectory prediction, shooting heuristic, and damping control. The results showed that the proposed methodology smoothed CAV movements and harmonized the following human-driven vehicles. Improvements were achieved in throughput, speed variations, fuel consumption and surrogate safety measures [24] . Liu et al. conducted a study to model the impacts of CACC on mixed traffic flow on multi-lane freeway facilities. A modeling framework that adopts a new vehicle dispatching model to generate the high-volume traffic flow was proposed. The framework also ensures realistic CACC vehicle behaviors. By incorporating new lane changing rules and automated speed control algorithms, the proposed modeling framework could further reproduce traffic flow dynamics under the influence of CACC operation strategies. Case studies on four-lane freeway segments with on-and off-ramps were illustrated and the results indicated that their proposed modeling framework could improve system mobility for different MPRs. Specifically, at a 60% or lower MPR, managed lane and vehicle awareness devices were helpful and led to a capacity improvement ranging from 8% to 23%. At a 100% MPR, the freeway capacity was 90% higher than the base case [25] . Bichiou et al. proposed an input minimal platooning controller. This controller incorporates various dynamic and kinematic constraints, such as acceleration, velocity, and collision avoidance constraints. The authors tested the controller using a calibrated real network-freeways in downtown Los Angeles-in a simulation environment utilizing the INTEGRATION microscopic simulation software. The results showed that a significant reduction in system-wide travel time, delays, and fuel consumption was achievable. It was also observed that, although the controller only controlled vehicles in the platoon, all the vehicles in the network benefited from the controller [26] . In summary, CV-enabled applications in uninterrupted traffic flow typically regulate car following and lane changing behavior to minimize fuel consumption, maximize throughput, and improve surrogate crash measures, such as the standard deviation of speed, time to collision, etc. The results varied case by case depending on the different applications, congestion levels, and/or MPRs of CVs. In most cases, the percentage of CVs needed to be above a certain level to achieve significant results. The majority of the previous studies used simulation tools to evaluate the effectiveness of the applications, while field tests were rare due to the scarcity of equipped facilities as well as safety concerns. Dresner and Stone were the first to introduce the concept of signal-head-free intersection control [27] . In their approach they considered a first-in first-out intersection control mechanism. Mirheli et al. also designed a signal-head-free intersection control logic using a dynamic programming model. The authors used a stochastic look-ahead technique on a Monte Carlo tree search algorithm to determine the near-optimal actions over time to prevent conflicts. By doing so, the model developed by the authors maximized the intersection throughput. The simulation results showed that travel time was significantly reduced at intersections for different demand patterns. Depending on various demand patterns, the improvement ranged between 59% and 84%, compared to intersections controlled by fixed-time and fully-actuated traffic signals [28] . Similarly, Zohdy et al. developed the intersection CACC system to regulate the flow of traffic proceeding through signalhead free intersections [29] [30] [31] . This work was then extended to optimize the flow of traffic proceeding through a roundabout [32] . Work by Elhenawy et al. investigated the use of game theory to optimize vehicle movements through the intersection [33] . Bichiou and Rakha developed a signal-head free offline controller in [34] and extended it to real-time vehicle trajectory optimization considering dynamic constraints to enhance the mobility of intersections. The results showed that the proposed algorithm outperformed other intersection control strategies by not only producing lower delay but also decreasing vehicle fuel > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < consumption and CO2 emissions [35] . In summary, CACC and i-CACC are efficient in decreasing delays and fuel consumptions comparing to conventional signal controls, abandoning the concepts of intersection control by optimizing the scheduling and movement of vehicles traversing an intersection. Table III lists the studies for signal-free intersections. The existing studies were mainly conducted in a simulation environment. Numerous studies have attempted to quantify the benefits of V2X enabled applications along signalized roadways, as summarized in Table IV . Some of the previous studies tested V2X-enabled applications along arterials with one or more signalized intersections, where there are traffic controls that force vehicles to decelerate, stop, and/or accelerate at the onset, during, or at the conclusion of a red indication. By using the data transmitted from other vehicles and infrastructures, the V2X-enabled applications can guide vehicles through intersections with minimum delays, emissions, and without colliding into other vehicles by sharing Signal Phasing and Timing (SPaT) data. Table IV summaries these studies. Early work on GLOSA or eco-driving in the vicinity of signalized intersections started with the work of Kamalanathsharma and Rakha [36] [37] [38] [39] . The authors introduced a dynamic programming approach to deriving the fuel-optimum vehicle trajectory. Yang et al. extended this work by predicting the queue at the intersection approach [40] and considering multiple signalized intersections in deriving the optimum solution. By varying demand level, MPRs, phase splits and offsets, as well as the distances among the consecutive traffic signals, they concluded that a fuel consumption saving of 13.8% can be achieved given a 100% MPR. Combining higher MPRs with shorter phase lengths produced larger fuel savings. Demand levels, traffic signal offset, and traffic signal spacing all affect the results [41] . Later Almannaa et al. tested the controller in a controlled field environment. Three scenarios were tested: normal driving, driving with a speed advisory, and automated Eco-CACC. The field experiment tested four red indication offset values randomly delivered to drivers along an uphill and downhill, totaling 1,563 trips by 32 different participants. The results showed that the proposed Eco-CACC system can reduce fuel consumption significantly; a saving in fuel consumption of 31% and travel time of 9% was achieved. The results also demonstrated that automatic control yields more significant benefits than a human control scenario [42] . A GLOSA system was implemented in a study by Bradaï et al. This system coaches the driver to adapt their vehicle speeds such that the drive can safely proceed through upcoming traffic signals during a green indication. It allows reducing stop times and unnecessary accelerations in urban traffic situations, thereby saving fuel and reducing CO2 emissions. The results obtained show a significant reduction in CO2 emissions. However, the authors also stated that the results were obtained under specially designed circumstances where the trip length was only 1,500 meters and the experiment environment can be considered as a straight line with no other vehicles [43] . An eco-approach application was designed to provide drivers with recommendations to encourage "green" driving while approaching, passing through, and departing intersections in a study conducted by Xia et al. Both simulation experimentation and field operational testing were carried out to demonstrate the eco-approach application and to quantify its potential fuel and CO2 savings. The results showed that a communication platform based on a 4G/LTE C-V2X network link and a cloud-based server infrastructure was effective and sufficient for this kind of application. It was found in both the simulation experiment and the field operational testing that an average of 14% fuel and CO2 savings could be achieved [44] . A study by Wang et al. developed a cluster-wise cooperative eco-approach and departure (Coop-EAD) application for CAVs to reduce energy consumption and compared its performance to existing Ego-EAD applications. Instead of considering CAVs traveling through signalized intersections one at a time, the authors' approach strategically coordinates CAVs' maneuvers to form clusters using various operating modes: initial vehicle clustering, intra-cluster sequence optimization, and cluster formation control. The novel Coop-EAD algorithm is applied to the cluster leader, and CAVs in the cluster follow the cluster leader to conduct EAD maneuvers. A preliminary simulation study for a given scenario showed that, compared to an Ego-EAD (speed and location) application, the proposed Coop-EAD application achieves an 11% reduction in energy consumption, up to an 18% reduction in pollutant emissions, and a 50% increase in traffic throughput, respectively [45] . In a study conducted by Moser et al., a CACC approach using stochastic linear model predictive control strategies was proposed. Both V2V and V2I communication were assumed to be present. The objective of this approach was to minimize the fuel consumption in a vehicle-following scenario. By means of a conditional Gaussian model, the probability distribution of the upcoming velocity of the preceding vehicle was estimated based on current measurements and upcoming traffic signal timings. The evaluation of the controllers showed a significant reduction in vehicle fuel consumption compared to the predecessor while increasing safety and driving comfortably [46] . A study conducted by Bento et al. designed a novel intersection traffic management system for CAVs. The developed intelligent traffic management (ITM) techniques were proved to be successful in reducing the delays and emissions without colliding with each other. The data needed for the ITM system is supported by V2X communication where the vehicle position and speed are exchanged among vehicles and infrastructure. In addition to the savings in CO2 emissions and delays, the authors also found that the improvement is more significant when the traffic is heavier [47] . The study conducted by Wan et al. proposed a speed advisory system (SAS) for pre-timed traffic signals to minimize fuel consumption. They showed that the minimal fuel driving strategy alternates between periods of maximum acceleration, > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < engine shut down, and sometimes constant speed. Instead of using this bang-singular-bang control, they employed a suboptimal solution without significantly sacrificing the drivability but with significant improvement in fuel economy. The SASequipped vehicles not only improved their own fuel economy, but also benefit other conventional vehicles. A higher MPR generated more fuel savings. The cost in traffic flow and travel time was minimum [48] . A simulation-based case study implemented on a hypothetical four-way single-lane approach intersection under varying congestion conditions showed that the Cooperative Vehicle Intersection Control (CVIC) algorithm significantly improved intersection performance compared to conventional actuated intersection control, with a 99% and 33% reduction in stopped delay and total travel time, respectively. In addition, the CVIC algorithm significantly improved air quality and energy savings: 44% reduction of CO2 and 44% fuel consumption savings [49] . The study by Ho Chaudhuri et al. developed a fuel-efficient control strategy for a group of connected HEVs considering congested traffic and traffic signals that are common in urban road conditions. A higher-level controller develops an optimal velocity profile focusing on minimizing the average tractive energy consumption and reducing red light idling. The lowerlevel controller tracks the velocity profile obtained from the higher-level controller by optimally splitting power between the vehicle engine and the battery using an adaptive equivalent consumption minimization strategy (ECMS). The simulation results showed that using the proposed method, no vehicle had to stop for red lights. ECMS can track the velocity supplied by the higher level controller perfectly and keeps the state of charge of the battery within 1.5% of the initial battery state of charge [50] . Liu et al. designed a cooperative signal control algorithm utilizing CACC datasets and data collected by traditional fixed traffic sensors. The control strategy proved to be effective with mixed traffic as well as 100% CACC traffic. The average vehicle speed and the average vehicle MPG increased by more than 10% and speed increased by 13% when the CACC MPR was 100%, while the speed increased by 36% and MPG improved by more than 34% when the CACC MPR was 40%. Even with a 0% CACC, the proposed method can generate benefits with speed and MPG improvements of 12.5% and 12.2%, respectively [51] . In a study conducted by Lebre et al., a GLOSA system was developed. The results showed that GLOSA could reduce vehicle emissions, waiting time, and travel times. One feature of this study is that it involved real scenario testing. The authors tested the traffic signal with communication devices that transmitted identity, timestamp, location, and phase information. The data were sent to the V2X equipment, which was composed of an IEEE802.11p complaint WiFi router and an antenna mounted to the traffic light, through an Ethernet connection. In the case when there was only one vehicle, both simulation and the in-field test achieved similar savings in CO2 emissions (11% and 13%). When there was more than one vehicle included (only in simulation) a 5% saving in CO2 emissions and 30% of waiting time saving were observed when the MPR was around 50%. The simulations were conducted with two other networks and the savings in time and CO2 emissions were also significant-for the grid network, a reduction of 10.5% in CO2 emissions and an 80% decrease in waiting time was observed with an MPR of 100%;, while for the Valeo shuttle transit route, the reduction of the travel time was about 2.05% with 50% vehicle MPR and 50% MPR for infrastructure. The saving increased quickly to 16% when the CAV MPR increased to 50% and 60% for infrastructure [52] . Chen and Rakha developed a connected eco-driving controller for BEVs that assists BEVs negotiating signalized intersections by minimizing energy consumption. This controller features multiple realistic constraints, including vehicle acceleration and deceleration behavior, BEV energy consumption behavior, and the intercorrelation of speed, location, signal timing in a CV environment. The results showed that the authors' controller can effectively reduce stopand-go behavior while at the same time generating a savings of 9.3% in energy consumption and 3.9% in delays [53] . In another study, the same authors developed a HEV Eco-CACC system. The system computes real-time energy-optimized trajectories for HEVs. Within the system, two models are developed: one HEV energy model to compute the instantaneous fuel consumption and one vehicle dynamics model to capture the relationship between speed, acceleration as well as tractive/resistance forces. The results revealed that on an arterial with three signalized intersections, the proposed system produced a 7.4% reduction in energy consumption, a 5.8% reduction in delay, and a 23% reduction in vehicle stops. Elouni et al. designed an adaptive traffic signal control system using two methods-game-theoretic decentralized and centralized perimeter control-and compared the performance of both systems. The results showed that the Nash Bargaining (NB) traffic signal controller can prevent congestion from building and improve the performance of the entire network. The NB controller outperforms both gating and non-gating controllers with significant reductions in vehicle travel time, emissions, fuel consumption, and delays. Specifically, the decentralized NB controller led to significant reductions of, with or without gating, 21% to 41% in total delay, 40% to 55% in CO2 emission levels, and 12% to 20% fuel consumption [54] . A study by Maile tested a cooperative intersection collision avoidance system using vehicles from five OEMs: Daimler, Ford, GM, Honda and Toyota, among which the GM vehicle contained the full prototype. The intersection at 5th Ave. and El Camino Real in Atherton, CA was used for the testing. Results showed that the system can almost 100% reliably send out warnings and therefore prevent crashes from happening [55] . > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < The system can reliably send out warning messages 100% of the time and therefore prevent crashes from happening. Field test > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < In summary, V2X-enabled applications in interrupted traffic flow typically involve a speed advisory system where vehicles are advised to go through a traffic signal-controlled intersection to avoid unnecessary deceleration, acceleration or stops. By adjusting the speeds of vehicles upstream of traffic signals, these applications can help vehicles go through the intersection with minimum interruption and fuel consumption. The existing studies were mainly conducted in a simulation environment. Most of them studied conventional ICEVs. Some tested other vehicle types, such as BEVs, or trucks and buses. Existing studies illustrated the promising results of using V2X-enabled applications in terms of improvements in environment, mobility, and safety. This section briefly describes the work on the modeling of the interdependencies of the transportation and communication systems as it relates to the evaluation of V2X applications, as summarized in Table V . Veins, one of the well-known frameworks in this category, integrates SUMO, a traffic simulator, and OMNET++, a communication simulator, using the TraCI interface. TraCI is a messaging standard that uses the Transmission Control Protocol (TCP) connections to share messages across the two simulators. One advantage of using TraCI is that it allows for bidirectional coupling of the two simulators. However, a key shortcoming of Veins is that it is unable to model large-scale networks. Veins by default does not support C-V2X mode 4; however, OpenCV2X [56] has extended Veins to support C-V2X modeling. Specifically, OpenCV2X implemented the C-V2X standard in SimuLTE, which is built on top of OMNET++ to support LTE communication [57] . Another integrated simulator was developed which entailed integrating the SUMO traffic and NS-3 communication simulators. The authors extended NS-3 to support C-V2X release 14. The authors did not mention how the integration between NS-3 and SUMO was done, but we assume that they used the TraCI interface as was done in Veins because TraCI is the external interface for SUMO developed by the SUMO authors [58] . In two studies conducted by Elbery et al., the authors developed an integrated simulator, VNetIntSim, which combined the OPNet communication simulator with the INTEGRATION traffic simulator. The integrated framework was used to demonstrate the impact of mobility parameters (traffic stream speed and density) on the communication performance through different applications including the File Transfer Protocol using TCP and Voice over Internet Protocol based on the User Datagram Protocol [59, 60] . Hoque et al. attempted to address the scalability problem by parallelizing one or both simulators to speed up the execution time and thus support simulation of large-scale road networks with hundreds of thousands of vehicles. The authors developed an Integrated Distributed Connected Vehicle Simulator (IDCVS) by incorporating hardware-in-the-loop simulation with the integration of SUMO and OMNET++. The authors provided a partitioning heuristic algorithm that partitions the complex traffic network into two sets of partitions: one for SUMO and one for OMNET++. The tool was then used to model DSRC for connected vehicles considering different levels of market penetration [61] . The Vehicular Network Simulator (VNS); which integrates the traffic simulator DIVERT 2.0 and the communication system simulator, NS-3; falls into the third category of methods. VNS supports the 802.11p communication standard as it was developed before the release of the C-V2X LTE standard. VNS differs from the previous work we discussed in the way that NS-3 and DIVERT are integrated. Since both simulators are developed in the same programming language (C++), they were put into one executable environment instead of having the two executing programs communicating with each other. Although the two simulators share the same execution environment, they still communicate using TCP connections through the network integration module. In VNS, at each simulation time step, the traffic simulator is run first followed by the communication simulator. The communication simulator has node entities that are mapped to the vehicle entities in the traffic simulator. Each node entity has access to its corresponding vehicle entity. One last difference in VNS is the adaptation of the NS-3 network simulator. The authors adapted the implementation of NS-3 to support large-scale simulations. They applied the concept of nearest neighbors and the locality of vehicle position updates by using QuadTrees to accelerate the performance of the NS-3 communication simulator. However, the integrated simulator is still constrained by the computational speed of NS-3 [62] . Elbery and Rakha developed a network simulator that simultaneously models the traffic and communication systems. Unlike the previous work, the communication system abstraction is not a network simulator, but instead is an analytical model, which allows for large-scale modeling of CV applications. The authors developed an analytical communication model for the DSRC Media Access Control layer protocol that estimates the packet drop probability and delay using a Markov chain and a queuing model. They implemented the analytical model in the INTEGRATION microscopic traffic simulator and tested the integrated framework using a dynamic eco-routing application [63, 64] . > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < [65] . In Table VI , we summarize the various parameters associated with the different simulator studies. This includes the different road networks, simulation times, number of vehicles simulated, and execution time for the various CV application simulators. In summary, developing a fully integrated traffic and communication simulator is essential for the modeling CV applications. Most of the work integrates the two systems using dedicated simulators in the respective fields, which models low level details of both systems but lacks the capability of largescale simulations. Recent work uses analytical modeling techniques to substitute one of the dedicated simulators for the benefits of scaling to large scale simulations while not sacrificing significantly on the accuracy of the modeling tool. This section describes how the tools developed in the previous section are used to assess the performance of various CV applications while capturing the mutual interdependencies of the transportation and communication systems. Schiegg et al. studied the environmental awareness problem of vehicles on the road, which they called incomplete vehicle perception due to the limited range of the vehicle's on-board sensors. They proposed extending the vehicle's perception by sharing the vehicle's information, collected by the vehicle's sensors, with other vehicles on the road using C-V2X communication technology. The service of sharing information within vehicles on the road is called collective perception. The authors proposed an analytical model to evaluate the performance of the service using the C-V2X Mode 4 standard and used an analytical model to model the C-V2X standard. The authors found that the collective perception service enhanced the information when using C-V2X to share sensors' information. They also found that although C-V2X was useful, more enhancement was needed regarding the latency requirements of vehicle safety applications [66] . In the study by Segata et al., the authors studied the performance of the C-V2X communication standard in the context of platooning applications. The authors investigated the impact of the scheduling algorithm for Mode 4 on the platoon formation using several controllers. Findings suggested that although C-V2X performed very well in terms of Packet Delivery Ratio, it did not perform well in terms of packet loss bursts. The authors showed that the packet loss bursts can hinder C-V2X usefulness in safety applications, and claimed that the packet loss bursts are due to the scheduling algorithm and the half-duplex nature of the C-V2X channel (vehicles cannot receive and transmit at the same time) [67] . A study by Rajab and Miucic investigated the performance of the C-V2X communication standard on the performance of two safety applications: Emergency Vehicle Alert and High Beam Assist. The authors modeled the C-V2X communication standard using empirical data from the Crash Avoidance Metrics Partners performance assessment project to create several Packet Error Ratio (PER) curves for different conditions (ideal, medium, and severe). They smoothed the PER curves using splines applied to the empirical data [68] . A truck platoon application was tested in [69] using C-V2X communication. The authors found that the best scenario is to use C-V2X mode 3 in areas covered with LTE-infrastructure and C-V2X mode 4 in areas not covered to get the best benefits from the platoon (i.e., smaller inter-truck gaps). The authors, however, noted that one critical configuration of the C-V2X mode 4 is the re-selection counter, which must be tuned carefully to achieve good performance [69] . Marco et al. developed an open-source framework for testing V2X applications. They evaluated two applications: emergency vehicle (V2V) and area speed advisory (V2I/V2N) using DSRC and C-V2X communication technologies. They showed that the two applications were valuable and did help the emergency vehicle to maintain high speed while crossing intersections (in the V2V case) and decreased the number of collisions in the area of the speed advisory applications (in the V2I/V2N case) [70] . Mouawad et al. tested and evaluated a cooperative collision avoidance V2I application at urban intersections. The vehicles shared their Lidar sensor measurements (local occupancy maps) with RSUs. The RSUs fused all the messages into a global occupancy map and sent it back to all vehicles in range. The authors found that the configuration for the best performance was a message size of 1,685 bytes for global occupancy maps, which led to the lowest obstacle mis-detection rate and a packet generation rate of 10 Hz when the number of vehicles was less than 70 and 5 Hz otherwise [71] . A C-V2I-based system for collision avoidance was introduced and evaluated in a study by Malinverno et al. The authors evaluated the effectiveness of their collision detection algorithm using Cooperative Awareness Messages that were sent by vehicles and pedestrians using C-V2I. It was not clear what the configuration of the communication protocol was, but the authors mentioned that they used SimuLTE-veins for simulating the C-V2I. The authors assumed an MPR of 100% and tested the location of the application server in two settings: the cloud and at the RSU. The results of their algorithm was not affected by the location of the application server [72] . Video-assisted overtaking maneuver application using image processing and video streaming over C-V2X was introduced in a study by Magalhaes et al. The C-V2X enhanced the performance of the application by having low packet-loss and high video quality; however, it affected the latency. The > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < authors used a C-V2X onboard unit with a Cooperative Awareness Messages frequency of 10 Hz, HARQ enabled, and transmission power of 20 dBm with Line-of-sight assumption [73] . A study conducted by 5GAA analyzed the capability of 3GPP LTE-V2X PC5 (LTE side-link) and IEEE 802.11p (DSRC or ITS-G5) in reducing the number of fatalities and serious injuries. The model used the number of crashes as a baseline number and the fraction of signal delivery reliability, effectiveness of receiving alert/warning message, as well as some other ratios to estimate the number of crashes that could be avoided. The results showed that LET-V2X would avoid a greater number of crashes when compared to 802.11p due to a combination of the superior performance of LTE-V2X and the market led conditions that better favor the deployment of LTE-V2X [74] . Rebbeck et al. concluded that C-V2X enhances road safety and traffic efficiency. The base and equitable 5.9 GHz user scenario appeared to the be the most beneficial way to deploy C-V2X considering the cost of upgrading the roadside infrastructure. Additional benefits can be achieved if LTE PC5 communication is integrated in smartphones. The study also concluded that the cost of upgrading the in-vehicle C-ITS system would be significant [75] . Beyrouty et al. evaluated seven bundles of C-ITS services that are mature and are expected to be deployed in the short or medium term, including safety-based V2V services, V2I services that deliver the most benefit on motorways, V2I services mostly appliable in urban areas, services intended to provide information regarding parking, services intended to provide traffic and smart routing information, and V2X vulnerable road user protection services. The authors also assessed policy options considering six key themes: privacy, security, interoperability, compliance assessment, continuity, and enabling conditions. They estimated the savings in avoiding crashes, fuel consumption savings, and emission reductions for different policy options [76] . Using the INTEGRATION simulation tool Elbery and Rakha evaluated the mutual impact of an eco-routing transportation application and the V2I communication system using the 802.11p protocol. They tested the eco-routing application performance using different measures of effectiveness for different CV MPRs and congestion levels. The results of their study showed that reasonable fuel savings were achieved at a low-to-medium MPR. However, at high MPRs and high vehicular traffic congestion levels the eco-routing application performance degraded and caused higher fuel consumption and even a gridlock in the road network. This was because of the sub-optimal routes produced by the eco-routing application caused by the high packet drop rate [77, 78] . Alternatively, ignoring the communication system constraints produced benefits that increased as the CV MPR increased. This study clearly demonstrated the need to model the communication system as part of the evaluation of CV applications and demonstrated that without such an integrated modeling framework, erroneous conclusions could be derived. In summary, V2X-enabled intelligent transportation applications like collective perception, platooning, emergency vehicle alert, area speed advisory, collision avoidance, and video-assisted overtaking maneuvers have been tested in the literature. Despite the promising benefits of V2X technology, there is still more room for enhancement, specifically from the latency perspective. This paper reviewed numerous research efforts that attempted to quantify the impacts of various CV applications on the transportation system's mobility, safety, and the environment. Based on the literature review, the following conclusions can be derived: 1. CV applications are highly efficient in improving the transportation system's mobility and safety and minimizing vehicle emissions and fuel consumption levels. Assuming perfect wireless communication among the different components of the CV environment, the mobility, safety, and environmental impacts typically increase as the MPR of CVs increases. 2. The MPR of CVs is a critical factor that affects the performance of CV applications. A minimum MPR was recommended in many studies to reach a significant level of savings in travel time, delays, and emissions. Depending on the types and features of the different applications, the minimum MPR varies from 10% to 40%. 3. The benefits of CV applications generally increase with the increase of MPR if communication constraints are not explicitly accounted for. The benefits, however, do not follow a linear relationship with the MPR. Some applications will start to show effects once a minimum number of CVs are in the traffic flow, but the benefits flatten out after the MPR reaches a certain level. Some applications do not generate benefits until the MPR reaches a certain level; however, the benefits increase accordingly with the increasing percentage of CVs in the traffic flow. 4. Most of the previous research efforts assumed that the communication among different parts of the system is latency and/or error-free, assuming the C-V2X data are accepted and applied by all users without an error. This assumption should be modified in future research. Due to the delays in signal transmission, bandwidth, and varied level of acceptance and cooperation of travelers in the system, the benefits will be lower in real-world situations. The accurate benefits in the real world need to be estimated accordingly accounting for limitations in the communication system. 5. A unique study of an eco-routing application showed that a minimum MPR of about 10% was needed to achieve benefits, with the benefits peaking at a CV MPR of 30-40%. Higher MPRs loaded the 802.11p communication system and at an MPR of 75%, the application produced gridlock because of the loss and latency in the packet transmission. This finding is unique because it > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < demonstrates the importance of including the communication system constraints in the modeling of CV applications. 6. Most previous research efforts were based on microscopic simulations. Very few studies tested CV technologies in the field given the high cost of such testing. Even in those cases with field testing, the testbed typically was a closed environment involving a small sample of CVs (no more than 10), thus not overloading the communication system. 7. The benefits of CV applications vary based on the application type, study design, vehicle type, test location or network type, roadway condition, and utilized energy/emission model. The differences might also be caused by different modeling algorithms, modeling assumptions or parameter settings in the model. Based on these findings, we recommend the following research directions: Inventory of U.S. Greenhouse Gas Emissions and Sinks United States Environmental Protection Agency (EPA) 7 Types of Vehicle Connectivity Evaluating an Eco-Cooperative Automated Control System (Eco-CAC)," presented at the The 101th Transportation Environmental Benefits of C-V2X Assessing the Potential Impacts of Connected Vehicles: Mobility, Environmental, and Safety Perspectives Energy saving potentials of connected and automated vehicles Cooperative Highway Traffic: Multiagent Modeling and Robustness Assessment of Local Perturbations Mobility and Energy Consumption Impacts of Cooperative Adaptive Cruise Control Vehicle Strings on Freeway Corridors Energy and flow effects of optimal automated driving in mixed traffic: Vehicle-in-theloop experimental results Model-free speed management for a heterogeneous platoon of connected ground vehicles Efficient vehicle driving on multi-lane roads using model predictive control under a connected vehicle environment Cooperative autonomous driving for traffic congestion avoidance through vehicle-to-vehicle communications A mixed traffic speed harmonization model with connected autonomous vehicles Modeling impacts of Cooperative Adaptive Cruise Control on mixed traffic flow in multi-lane freeway facilities A Cooperative Platooning Controller for Connected Vehicles Multiagent Traffic Management: An Improved Intersection Control Mechanism Development of a signal-head-free intersection control logic in a fully connected and autonomous vehicle environment Intersection management for autonomous vehicles using iCACC Game theory algorithm for intersection-based cooperative adaptive cruise control (CACC) systems Intersection management via vehicle connectivity: The intersection cooperative adaptive cruise control system concept Enhancing Roundabout Operations via Vehicle Connectivity An intersection game-theory-based traffic control algorithm in a connected vehicle environment Developing an Optimal Intersection Control System for Automated Connected Vehicles Real-time optimal intersection control system for automated/cooperative vehicles Agent-based modeling of Eco-Cooperative Adaptive Cruise Control systems in the vicinity of intersections Leveraging Connected Vehicle Technology and Telematics to Enhance Vehicle Fuel Efficiency in the Vicinity of Signalized Intersections Multistage dynamic programming algorithm for eco-speed control at traffic signalized intersections Fuel-Optimal Vehicle Throttle Control: Model Logic and Preliminary Testing Eco-Cooperative Adaptive Cruise Control at Signalized Intersections Considering Queue Effects Eco-Driving at Signalized Intersections: A Multiple Signal Optimization Approach Field implementation and testing of an automated eco-cooperative adaptive cruise control system in the vicinity of signalized intersections A Green Light Optimal Speed Advisor for Reduced CO2 Emissions Field operational testing of ecoapproach technology at a fixed-time signalized intersection Cluster-wise cooperative eco-approach and departure application for connected and automated vehicles along signalized arterials Cooperative adaptive cruise control applying stochastic linear model predictive control strategies A study of the environmental impacts of intelligent automated vehicle control at intersections via V2V and V2I communications Optimal speed advisory for connected vehicles in arterial roads and the impact on mixed traffic Development and Evaluation of a Cooperative Vehicle Intersection Control Algorithm Under the Connected Vehicles Environment Hierarchical control strategies for energy management of connected hybrid electric vehicles in urban roads Traffic signal control by leveraging Cooperative Adaptive Cruise Control (CACC) vehicle platooning capabilities Real scenario and simulations on GLOSA traffic light system for reduced CO2 emissions, waiting time and travel time Battery Electric Vehicle Eco-Cooperative Adaptive Cruise Control in the Vicinity of Signalized Intersections Adaptive Traffic Signal Control: Game-Theoretic Decentralized vs. Centralized Perimeter Control Cooperative Intersection Collision Avoidance System for Violations (CICAS-V) for Prevention of Violation-Based Intersection Crashes OpenCV2X Mode 4 A Simulation Extension for Cellular Vehicular Communication Networks Bidirectionally Coupled Network and Road Traffic Simulation for Improved IVC Analysis Performance Analysis of C-V2X Mode 4 Communication Introducing an Open-Source C-V2X Ieee Vts Veh Technol VNetIntSim -An Integrated Simulation Platform to Model Transportation and Communication Networks An Integrated Architecture for Simulation and Modeling of Small-and Medium-Sized Transportation and Communication Networks Smart Cities, Green Technologies, and Intelligent Transport Systems Parallel Closed-Loop Connected Vehicle Simulator for Large-Scale Transportation Network Management: Challenges, Issues, and Solution Approaches VNS: An Integrated Framework for Vehicular Networks Simulation Vehicular Communication and Mobility Sustainability: the Mutual Impacts in Large-scale Smart Cities Large-Scale Modeling of VANET and Transportation Systems INTEGRATION Large-Scale Modeling Framework of Direct Cellular Vehicle-to-All Analytical Performance Evaluation of the Collective Perception Service in C-V2X Mode 4 Networks A Critical Assessment of C-V2X Resource Allocation Scheme for Platooning Applications Assessment of Novel V2X Applications Using a Simulation Platform 3GPP C-V2X and IEEE 802.11p for Vehicle-to-Vehicle communications in highway platooning scenarios A Multi-stack Simulation Framework for Vehicular Applications Testing Impact of LTE-V2X Connectivity on Global > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE-CLICK HERE TO EDIT) < Occupancy Maps in a Cooperative Collision Avoidance (CoCA) System Performance Analysis of C-V2I-based Automotive Collision Avoidance Video-assisted Overtaking System Enabled by C-V2X Mode 4 Communications Ieee Conf Wirel Mob An Assessment of LTE-V2X(PC5) and 802.11p direct communications technologies for improved road safety in teh EU Sociao-Economic Benefits of Cellular V2X Support study for imparc assessment of cooperative intelligent transport systems City-Wide Eco-Routing Navigation Considering Vehicular Communication Impacts VANET communcation and mobility sustainability The authors would like to thank Georgios Efraimidis, Anne-Lise Thieblemont, and Jim Misner of Qualcomm for their input on the manuscript. She is the author/co-author of over 50 papers and reports. She is the principal investigator (PI) and faculty investigator on multiple projects in the area of driving behavior, connected vehicle research, traffic modeling and simulation, Intelligent Transportation Systems (ITS), advanced traffic signal control systems, Geographic Information Systems (GIS) in transportation, and transportation safety modeling. She has extensive experiences on network monitoring and controlling, optimization of traffic signals, traffic simulation, and spatial analysis. She is the registered professional engineer (PE) in the state Virginia. She is also a member of the ITS America and TRB Committee on Geographic Information Science (AED40).