BASEMENT - AN ARCHITECTURE AND METHODOLOGY FOR DISTRIBUTED AUTOMOTIVEREAL-TIME SYSTEMS

Citation
H. Hansson et al., BASEMENT - AN ARCHITECTURE AND METHODOLOGY FOR DISTRIBUTED AUTOMOTIVEREAL-TIME SYSTEMS, I.E.E.E. transactions on computers, 46(9), 1997, pp. 1016-1027
Citations number
41
Categorie Soggetti
Computer Sciences","Engineering, Eletrical & Electronic","Computer Science Hardware & Architecture
ISSN journal
00189340
Volume
46
Issue
9
Year of publication
1997
Pages
1016 - 1027
Database
ISI
SICI code
0018-9340(1997)46:9<1016:B-AAAM>2.0.ZU;2-L
Abstract
BASEMENT(TM) is a distributed real-time architecture developed for veh icle internal use in the automotive industry. BASEMENT covers applicat ion development, as well as the hardware and software that provide exe cution and communication support. This paper gives an overview of the BASEMENT concept, as well as presenting two system realizations. The f irst realization is based on the commercial real-time kernel Rubus, wh ile the second is an ultra-dependable architecture (DACAPO) with provi sions for fault tolerance at various system levels. BASEMENT is design ed for the automotive systems of the future. These systems will be req uired to simultaneously handle multiple safety critical functions and a large number of less critical functions. All of these features are t o be provided at a production cost substantially lower than that of cu rrent systems, and, at the same time, with a reliability allowing vehi cles to be built without mechanical backup systems, even for safety cr itical subsystems such as braking and steering. The key constituents o f the concept are: 1) resource sharing (multiplexing) of processing an d communication resources, 2) a guaranteed real-time service for safet y critical applications, 3) a best-effort service for nonsafety critic al applications, 4) a communication infrastructure providing efficient communication between distributed devices, 5) a program development m ethodology allowing resource independent and application oriented deve lopment of application software, and 6) a straightforward and well-def ined operation principle enabling efficient fault tolerance mechanisms to be employed.