Measurement of V2X Prototypes Performance. Comparison with Simulation Results Iulia Ivan 1,2 , Philippe Besnier 2 , Xavier Bunlon 1 , Philippe Boutier 1 1 Renault, FR TCR RUC T93, 1 av. du Golf, 78288 Guyancourt, France 2 IETR, UEB, France, INSA, IETR, UMR CNRS 6164, F-35708 Rennes, France Abstract—This paper presents experimental results obtained from a test-bed with IEEE 802.11p modem prototypes and an RF channel emulator, and compares them with simulation results. Measurement results confirm some of the simulation results, and highlight the importance of using channel tracking algorithms at the receiver to support very high mobility. Index Terms—IEEE 802.11p standard, RF channel emulator, V2X communications I. INTRODUCTION Vehicle to Vehicle and Vehicle to Infrastructure (V2X) communication system is an important integrant part of the future ITS (Intelligent Transport Systems) global architecture. It operates at 5.9 GHz, and its applications mainly aim at improving road safety and traffic efficiency. IEEE 802.11p [1], published in July 2010, and its European profile, EN 202 663 [2], currently under development by ETSI, are the two standards dedicated to V2X communications. Physical (PHY) layer (i.e. layer 1) of the V2X system is derived from IEEE 802.11a, standard amendment for WLAN – Wireless Local Area Networks, designed for almost static, indoor and outdoor environments. Destined to completely different applications, V2X system must however prove its reliability at very important vehicle speeds. PHY layer performance in highly mobile environments was investigated in [3]-[5] by computer simulations with dedicated stochastic channel models derived from intensive measurement campaigns [6]-[7]. In [4] was highlighted the impact of the receiver structure on the system performance in high mobility, and the necessity of implementing channel tracking algorithms at the receiver. [5] stressed that the conventional V2X receiver, i.e. with channel estimation based on the long training symbols only, might not fulfill the performance requirements imposed by the 802.11p draft standard [1]. In this paper, we present experimental results obtained with prototypes from two different V2X modem suppliers, and an RF (radiofrequency) channel emulator. The chosen channel emulator allowed the reproduction of the channel models proposed in the V2X draft standard [1], and used in our simulations presented in [5]. The measurements with channel emulator were preceded by receiver sensitivity evaluation, and Packet Error Rate (PER) performance estimations in AWGN (Additive White Gaussian Noise) channel. Obtained experimental results evidence the performance difference between the two modems. While one of them has a better sensitivity, the other modem, due to channel tracking algorithms implemented at the receiver, shows much better robustness in time-variant fading conditions. Then, comparison with simulation results is made, and similarities with obtained experimental results are highlighted. The paper is organized as follows: Section II describes the receiver sensitivity measurement procedure and depicts the test-bed employed. In Section III, experimental results with RF channel emulator are compared with simulation results. Finally, in Section V, we draw some conclusions. II. TEST-BED AND EXPERIMENTAL RESULTS FOR SENSITIVITY ANALYSIS A preliminary step to modem performance evaluation with channel emulator consists in determining the minimum input receiver sensitivity. It represents the input power level for which 10 % of PER is reached in AWGN channel [1]. Additional receiver performance plots can be obtained by progressively decreasing the input power level, and then can be compared to simulation results. For comparison purpose, Fig. 1 shows PER results obtained by computer simulation of the baseband V2X PHY layer in AWGN channel vs. the SNR (dB). Two data rates are here considered, 6 Mb/s and 12 Mb/s (i.e. Mode 3 and Mode 6), and packet lengths of 100 bytes. The description of the simulated receiver is outlined in [4]. It should be noted however that the slopes of the obtained curves depend on the specific receiver implementation, and perfect match with experimental results is highly improbable. In the following paragraphs, we describe the modem prototypes employed in our study, the test-bed used for receiver performance evaluation in AWGN channel, and finally we present the associated experimental results. A. Modem Description V2X modem prototypes from two different suppliers, called here M_X and M_Y, were employed for the measurements. The modems tag the transmitted packets, allowing the receiver to count the number of packets correctly decoded. Measurement campaigns presented in this document were performed in the framework of Score@f project [9].