9/24/20011Mani SrivastavaUCLA - EE DepartmentRoom: 7702-B Boelter HallEmail: [email protected]: 310-267-2098WWW: http://www.ee.ucla.edu/~mbsCopyright 2001  Mani SrivastavaModeling Embedded SystemsEE202A (Fall 2001): Lecture #22Copyright 2001  Mani SrivastavaReading List for This Lecture MANDATORY READING E.A. Lee, “Embedded Software,” UCB ERL Memorandum M01/26.http://ptolemy.eecs.berkeley.edu/publications/papers/01/embsystems/ Harel, D.; Lachover, H.; Naamad, A.; Pnueli, A.; Politi, M.; Sherman, R.; Shtull-Trauring, A.; Trakhtenbrot, M. STATEMATE: a working environment for the development of complex reactive systems. IEEE Transactions on Software Engineering, vol.16, (no.4), April 1990. p.403-14.http://ielimg.ihs.com/iel1/32/1950/00054292.pdf OTHER READING None9/24/200123Copyright 2001  Mani SrivastavaHow to Design Embedded Systems? (Wescon 1975) “...deliberately avoid data processing aides such as assemblers, high-level languages, simulated systems, and control panels. These computer-aided design tools generally get in the way of cost-effective design and are more a result of the cultural influence of data processing, rather than a practical need.” “It’ s my observation that the bulk of real-world control problems require less than 2,000 instructions to implement. For this size program computer aided design does little to improve the design approach and does a lot to separate the design engineer from intimate knowledge of his hardware.”4Copyright 2001  Mani SrivastavaMethodical Design of Embedded Systems Ad hoc approach to design does not work beyond a certain level of complexity, that is exceeded by vast majority of embedded systems Methodical, engineering-oriented, tool-based approach is essential specification, synthesis, optimization, verification etc. prevalent for hardware, still rare for software One key aspect is the creation of models concrete representation of knowledge and ideas about a system being developed - specification model deliberately modifies or omits details (abstraction) but concretely represents certain properties to be analyzed, understood and verified one of the few tools for dealing with complexity9/24/200135Copyright 2001  Mani SrivastavaAbstractions and Models Foundations of science and engineering Activities usually start with informal specification However, soon a need for Models and Abstractions is established e.g.: Chess and Poker - No Rules, No Games Connections to Implementation (h/w, s/w) and Application Two types of modeling: system structure & system behavior the behavior and interaction of atomic components coordinate computation of & communication between components Models from classical CS FSM, RAM (von Neumann) CSP (Hoare), CCS (Milner) Turing machine, Post Machine, Universal Register Machine6Copyright 2001  Mani SrivastavaGood Models Simple Ptolemy vs. Galileo Amenable for development of theory to reason should not be too general Has High Expressive Power a game is interesting only if it has some level of difficulty! Provides Ability for Critical Reasoning Science vs. Religion Practice is currently THE only serious test of model quality Executable (for Simulation) Synthesizable Unbiased towards any specific implementation (h/w or s/w)9/24/200147Copyright 2001  Mani SrivastavaSeparate Behavior from Architecture8Copyright 2001  Mani SrivastavaElements of a Model of a Computation System: Language Set of symbols with superimposed syntax & semantics textual (e.g. matlab), visual (e.g. labview) etc. Syntax: rules for combining symbols well structured, intuitive Semantics: rules for assigning meaning to symbols and combinations of symbols without rigorous semantics, precise model behavior over time is not well defined full executability and automatic h/w or s/w synthesis is impossible E.g. operational semantics (in terms of actions of an abstract machine), denotational semantics (in terms of relations)9/24/200159Copyright 2001  Mani SrivastavaSimulation and Synthesis Two sides of the same coin Simulation: scheduling then execution on desktop computer(s) Synthesis: scheduling then code generation in C++, C, assembly, VHDL, etc. Validation by simulation important throughout design flow Models of computation enable Global optimization of computation and communication Scheduling and communication that is correct by construction10Copyright 2001  Mani SrivastavaModels Useful In Validating Designs By construction property is inherent. By verification property is provable. By simulation check behavior for all inputs. By intuition property is true. I just know it is. By assertion property is true. Wanna make something of it? By intimidation Don’t even try to doubt whether it is trueIt is generally better to be higher in this list9/24/2001611Copyright 2001  Mani SrivastavaHeterogeneous Systems Hierarchical composition of models Need to understand how models relate when combined in a single system[Evans]12Copyright 2001  Mani SrivastavaModeling Embedded Systems Functional behavior: what does the system do in non-embedded systems, this is sufficient Contract with the physical world Time: meet temporal contract with the environment temporal behavior important in real-time systems, as most embedded systems are simple metric such as throughput, latency, jitter more sophisticated quality-of-service metrics Power: meet constraint on power consumption peak power, average power, system lifetime Others: size, weight, heat, temperature, reliability etc.System model must support description of bothfunctional behavior and physical interaction9/24/2001713Copyright 2001  Mani SrivastavaImportance of Time in Embedded Systems: Reactive OperationComputation is in response to external events periodic events can be statically scheduled aperiodic events harder worst case is an over-design statistically predict and dynamically schedule approximate computation algorithms As opposed to Transformation Operation in Interactive SystemsTransformationalPhysicalProcessesREACTIVE14Copyright 2001  Mani SrivastavaReactive Operation (contd.) Interaction with environment causes problems indeterminacy in execution e.g. waiting for events from multiple sources physical