International Journal of Digital Content Technology and its Applications Volume 4, Number 4, July 2010 15 Signal Framing Methods of Automotive Embedded Networked Control System for Bandwidth Consumption Optimization 1 Jian HU, 2 Gangyan LI, 3 Yeqiong SONG *1, 2 School of Mechanical and Electronic Engineering, Wuhan University of Technology, 122, Luoshi Road, 430070, Wuhan, Hubei Province, P. R. China, hujian@whut.edu.cn, gangyanli@whut.edu.cn 3 LORIA - Equipe TRIO,615, Rue Du Jardin Botanique, 54600 Villers Les Nancy, France, song@loria.fr doi: 10.4156/jdcta.vol4.issue4.2 Abstract Parameter (or variable) of automotive embedded networked control system is named as signal. In order to improve system performance and provide more remaining bandwidth for system extension, it is necessary to minimize the bandwidth consumption of frames transmitting the datum of signals. In this paper, the signal framing problem and methods of automotive embedded networked control system for bandwidth consumption optimization are studied. Besides OSpF (One Signal per Frame), other different signal framing methods, BFD-FNO (Best Fit Decreasing for Frame Number Optimization), BFBCD-BCO (Best Fit with Bandwidth Consumption Decreasing for Bandwidth Consumption Optimization) and BFPI-BCO (Best Fit with Period Increasing for Bandwidth Consumption Optimization), are presented based on bin-packing heuristic algorithm. Applying NETCARBENCH and algorithm implementation tool, the signal framing heuristic algorithm performance of automotive embedded networked control system for bandwidth consumption optimization is analyzed and evaluated. By comparing the performances of different heuristic algorithms, it is concluded that no matter the ECU number and signal load, BFD-FNO heuristic algorithm retains the signal framing efficiency for bandwidth consumption optimization. Keywords: Automotive Electronic, Automotive Embedded Networked Control, Signal Framing, Bandwidth Consumption, Heuristic Algorithm 1. Introduction In this paper, the parameter (or variable) of automotive embedded networked control system is named as signal. All kinds of status signal and control signal are basic element for well-working of automotive embedded networked control system. The instance of signal, termed as the data of this signal, identifies different parameter value in automotive embedded networked control system. For example, “engine speed” is a signal. 1000rpm is one data of engine speed and a parameter value of EEC (Engine Electronic Control) subsystem and other subsystems. In a word, the basic function of automotive embedded networked control system is to share datum of signals between different ECUs (Electronic Control Units) based on in-vehicle network technology. Signal transmission in automotive embedded networked control system is implemented by frame (in this paper CAN (Controller Area Network) [1] [2] is used as the in-vehicle network protocol) which is the data unit exchanged in data link layer of ISO/OSI (International Standardization Organization/Open System Interconnect) Reference Model. Before sending the data of signal, it is packaged into CAN data frame according to CAN standard/extended frame format in data link layer. Every CAN data frame includes not only the data of signal but also some additional control information. After receiving CAN data frame, it is unpacked in data link layer and transferred the data of signal to upper layers. Because the signal transmission is implemented by CAN data frame, the temporal characteristics of frame depend on the temporal characteristic of signal [3]. After being provided ECU set, the signal set sent over the network of each ECU and the characteristics, such as period, deadline and size etc., of each signal, one can build the frame set on each ECU, the signals packaged into each frame and furthermore obtain the characteristics, such as size,