Dedicated short-range communication (DSRC) and 4G-LTE are two widely used candidate schemes for Connected Vehicle (CV) applications. It is thus of great necessity to compare these two most viable communication standards and clarify which one can meet the requirements of most V2X scenarios with respect to road safety, traffic efficiency, and infotainment. To the best of our knowledge, almost all the existing studies on comparing the feasibility of DRSC or LTE in V2X applications use software-based simulations, which may not represent realistic constraints. In this paper, a Connected Vehicle test-bed is established, which integrates the DSRC roadside units, 4G-LTE cellular communication stations, and vehicular on-board terminals. Three Connected Vehicle application scenarios are set as Collision Avoidance, Traffic Text Message Broadcast, and Multimedia File Download, respectively. A software tool is developed to record GPS positions/velocities of the test vehicles and record certain wireless communication performance indicators. The experiments have been carried out under different conditions. According to our results, 4G-LTE is more preferred for the nonsafety applications, such as traffic information transmission, file download, or Internet accessing, which does not necessarily require the high-speed real-time communication, while for the safety applications, such as Collision Avoidance or electronic traffic sign, DSRC outperforms the 4G-LTE.
In 2011, the US Department of Transportation (US DOT) announced plans to support the introduction of vehicle-to-vehicle (V2V) communication among light-duty vehicles in the USA, commonly known as “Connected Vehicles” [
Many vehicle manufacturers also pay due attention to the research and implementation of Connected Vehicle. The Crash Avoidance Metrics Partnership (CAMP) Vehicle Safety Consortium Communications (VSCC) comprising BMW, Daimler Chrysler, Ford, GM, Kia, Nissan, Toyota, and Volkswagen, in partnership with USDOT, proposed more than 57 application scenarios about Connected Vehicle, like safety applications, nonsafety applications, high potential benefit safety applications, and other applications [
Active safety latency requirements (units: seconds).
Traffic signal violation warning | 0.1 |
Curve speed warning | 1.0 |
Emergency electronic brake lights | 0.1 |
Precrash sensing | 0.02 |
Cooperative forward collision warning | 0.1 |
Left turn assistant | 0.1 |
Lane change warning | 0.1 |
Stop sign movement assistance | 0.1 |
US DOT has developed a Connected Vehicle Reference Architecture (CVRA) to help guide deployment of components by road operators and automotive, highway, and aftermarket equipment manufacturer and service providers [
Between DSRC and LTE, which one is an appropriate technology for Connected Vehicle applications? Will the combination or hybrid solution be more promising? This is an urgent open question which has been discussed recently.
Considering the harsh vehicular environment and related communication concerns, such as high level of the mobility of the nodes, multipath, and environmental dynamics caused by vehicles and pedestrians, IEEE proposed a modified version of the Wireless Local Area Network (WLAN) protocol, IEEE802.11p (commonly called “DSRC”), for vehicle-vehicle and vehicle-infrastructure communication. A dedicated bandwidth of 75 MHz in the 5.850 to 5.925 GHz band has been allocated by the US Federal Communications Commission (FCC) [
The existing cellular wireless infrastructure, particularly the 4G-LTE, has potential to be redesigned as a communication basis for vehicular cooperative safety systems, which can offer low latencies and high throughputs simultaneously, thus enabling more bandwidth-demanding and real-time critical services for end-users [
In China, there still exists a debate on whether DSRC should be utilized as the communication standard of the physical layer in the Connected Vehicle architecture of China, since the LTE networks are widely deployed all over the country and the cellular capabilities have already been on the roadmap for many vehicle manufacturers and for telematics applications [
Many researchers have expressed considerable research interests on the vehicular networking and proposed various solutions. However, the majority of these studies are based on computer simulation to avoid the high costs in the real field experiments. Vinel [
Compared with field testing, the software-based protocol simulation testing has the advantages of low cost, short deployment cycle, and flexible parameters setting. The shortcomings are also obvious: (1) the mathematical models used in a software-based simulation is formulated in an idealistic environment. The signal attenuations caused by vehicles, surrounding buildings, or vegetation are neglected; (2) the differences among the communication equipment and user terminals such as the performance diversity and installation conditions are hardly considered; and (3) the results of a software-based simulation are generally better than the actual field tests. However, in the practical application environment, these results will degrade to some extent.
On the basis of the above literature survey, it can be confirmed that it is greatly necessary to test the function and performance of DSRC and LTE in a real road environment, so as to verify the simulation results of various theoretical models and provide important data for the development of various vehicular networking applications.
This study tests the communication quality and reliability of the DSRC and LTE in a real road environment under 3 classic Connected Vehicle scenarios: Collision Avoidance, Traffic Message Broadcast, and Multimedia File Download between vehicle and infrastructure. This can be a critical baseline for many other Connected Vehicle applications. This paper made 3 key contributions. First, we build an independent and private wireless network platform made up of DSRC and LTE in Chang’an University Cooperative Vehicle and Infrastructure System (CU-CVIS) test-bed. This platform integrates the 4G-LTE cellular stations and the core network, DSRC roadside units, and vehicular on-board terminals. With the platform, lots of vehicular communication experiments under different mobility conditions can be conducted. Second, three Connected Vehicle application scenarios are set as Collision Avoidance, Traffic Text Message Broadcast, and Multimedia File Download, respectively. A software tool is developed to record the GPS-based position and velocity of the vehicles that are moving on the test-bed. With the tool, some crucial wireless communication performance indicators of DSRC and 4G-LTE such as the throughput, data loss rate, and the latency time under the above scenarios are acquired. Third, the results of the on-field vehicular communication performance are analyzed. The suitability of DSRC and 4G-LTE on the most popular Connected Vehicle scenarios is discussed. Moreover, some suggestions are given on how to integrate these two kinds of technologies in the future Connected Vehicle applications.
This paper is organized as follows. Section
As shown in Figure
Chang’an University Cooperative Vehicle-Infrastructure System test-bed.
The bird-view satellite image of CU-CVIS test-bed
The logic architecture of CU-CVIS test-bed
On the test-bed, we install a short-range wireless system with 4 DSRC roadside units (RSUs) and a 4G-LTE cellular communication system, which comprise a specialized vehicular communication platform. Both DSRC and LTE systems are isolated with the public telecom network. The locations of the RSUs and LTE antennas are denoted in Figure
The DSRC RSUs are mounted on the gantries established on the test-bed (see Figure
The DSRC RSU on the test-bed.
The LTE network deployed on the test-bed.
The 4G-LTE system is a private cellular wireless network operating on the frequency of 1.88–1.9 GHz with TD-LTE protocol, which is developed by China Datang Mobile Co., Ltd. As shown in Figure OBU connects LTE network through CPE, which realizes LTE protocol stack and TCP/IP protocol stack by wireless or wired mode and transmits data to the destination through 4G network. eNode-B is made up of RRU and BBU, which is a terminal of the air interface protocol and the first node to contact with user equipment (UE). eNode-B is responsible for the wireless bearer, downlink dynamic wireless resources, data packet scheduling, and mobility management. EPC is made up of Serving Gateway (SGW), PDN Gateway (PGW), Mobility Management Entity (MME), and Policy and Charging Rules Function (PCRF), which is responsible for data exchange and processing.
The application servers play very important roles, which are used for data collection, storage, and processing, and are connected to the base station through Gigabit Ethernet. The LTE platform has 4 directional antennas with the maximal transmission power of 60 w, which realize the full wireless signal coverage of the whole test-bed. The LTE platform is divided into 4 cells. Each cell can accommodate up to 200 items of UE with Uplink rate of 20 Mb/s and downlink rate of 80 Mb/s.
The constructed vehicular communication platform can conduct 4 kinds of wireless communication experiments: (1) V2V via DSRC; (2) V2V via LTE; (3) V2I via DSRC; (4) V2I via LTE. There are many famous test-beds around the world for testing connected and automated vehicles, such as the MCity of the University of Michigan and the GoMento test site located in Contra Costa County, California. These two test-beds focus on demonstration testing for applications of future Intelligent Transportation Systems, which mainly aims to verify the integrity performance and data-flow logic of the applications developed for the connected and automated vehicles (CAVs). CU_CVIS test-bed of Chang’an University focuses on the metafunction testing for each part of CAVs and the comprehensive performance testing under the limit conditions. For instance, CU_CVIS can test the performance of different communication modes and verify whether these communication methods can meet the requirements of some intelligent transportation applications on the real time and reliability of data transmission. It can test some fundamental functions of CAVs, such as positioning accuracy, target recognition accuracy based on vision, vehicle lateral or longitudinal control ability in high-speed condition, complex environment, and bad weathers.
We have two midsize vehicles to carry out the experiment. The setup of the on-board devices is shown in Figure
The setup of the on-board devices.
The connection of on-board devices
The mounting positions of the antennas
We develop a testing software kit to test the dynamic performance of DSRC and LTE. The software kit encrypts the API functions from the device manufacturer, which can achieve the transmission of packages or files between the vehicles and roadside devices. The software kit enables 3 functions. The first is to synchronize the on-board units with the GPS clock time. The second function is to calculate the communication parameters using the encrypted APIs. For instance, the Round-Trip Time (RTT) is acquired based on the Ping function supplied by Linux OS, the Packet Loss Rate is calculated through the success rate of the transmission of the WSMP packages, and the throughput is computed on the counts of the received UDP packages during a specified time segment. The last function is to record the GPS data of the vehicles, including longitude, latitude, and velocity, which is used to analyze the effect of the distance and movement on the wireless communication performance. The software interfaces are shown in Figure
The developed testing software.
The client interface
The server interface
We distill 3 general scenarios from the 75 safety applications proposed by CAMP [
In this scenario (see Figure
Collision Avoidance scenario.
In this scenario (see Figure
Traffic Text Message Broadcast scenario.
In this scenario, when the vehicles are passing by the DSRC RSU or entering the LTE cell, the OBUs will request file download. This scenario can also be extended to the applications of map download, Video on Demand (VOD), and so forth. In this scenario, we only employ one car to conduct the testing in order to avoid the two cars’ competition for the wireless bandwidth. Once the car enters the effective communication area, it will send a download request to the server. If the request is accepted, the downloading starts and the OBU begins to record the throughput of the link. The running model of the car is the same as the above two scenarios. In order to ensure the reliability of the testing data, the testing cars will run on the circular test road for 5 times at different speeds to execute the application of Multimedia File Download.
Figure
The performance degradation of DSRC and LTE under Scenario I.
PLR versus the vehicle running velocity
Average RTT delay versus the vehicle running velocity
As can be seen in Figure
Figure
The DSRC performance versus different communication distance when the vehicle velocity is 120 Km/h.
PLR versus the distance between OBU and DSRC RSU
Average RTT delay versus the distance between OBU and DSRC RSU
Figure
The performance degradation of DSRC and LTE under Scenario II.
PLR of both DSRC and 4G-LTE versus the vehicle running velocity
Average RTT delay of both DSRC and 4G-LTE versus the vehicle running velocity
From Figure
Figure
The throughput performance of DSRC and LTE in Scenario III.
In this paper, three Connected Vehicle application scenarios are set as Collision Avoidance, Traffic Text Message Broadcast, and Multimedia File Download, respectively. The communication performance of DSRC and LTE is investigated and analyzed with the developed hardware and software platform. Compared with the traditional computer simulation method for vehicle network performance, there are several innovations as follows: The performance of LTE is worse than that of DSRC under Collision Avoidance scenario, which is mainly caused by Doppler Effect and cellular handoff of LTE network. It means that LTE cannot meet the lowest requirement of 100 milliseconds for the safety applications. DSRC is more suitable for Collision Avoidance and other safety-related V2V traffic applications. Under the scenario of Traffic Text Message Broadcast, LTE has a long coverage range while DSRC has not. But, within the effective communication range, DSRC has better communication performance than LTE. Both LTE and DSRC are acceptable for the broadcast of the nonsafety text message. For the safety application such as the message broadcast from an electronic traffic sign, DSRC outperforms LTE. Because of the high costs of the dense deployment of DSRC RSUs, it is suggested that DSRC RSU can be installed on the spots where it is strongly related to safety. Under the scenario of Multimedia File Download, the throughput performance of LTE is significantly higher than that of DSRC at different vehicle running velocities. Furthermore, LTE has a long coverage range, which is more suitable for Multimedia File Download than DSRC. The combination of DSRC and LTE should be a good solution for Connected Vehicles. It not only enables the safe driving but also can supply high-quality telematics service to the drivers. In the future, the comparison experiments with more vehicles need to be conducted, because it is closer to the real-world scenarios and can test communication performance of DRSC and LTE in the extreme conditions.
The authors declare that there are no conflicts of interest regarding the publication of this paper.
The research is supported by the National Natural Science Foundation of China (no. 51278058), the Fundamental Application Research Program of China Ministry of Transport (no. S2013JC9397), the 111 project (no. B14043) and the Joint Laboratory of Internet of Vehicles sponsored by Ministry of Education and China Mobile. The authors also thank the anonymous reviewers and the editor for their valuable comments.