1 / 39

Q uantitative E valuation of E mbedded S ystems

Q uantitative E valuation of E mbedded S ystems. QUESTION DURING CLASS? Email : qees3TU@gmail.com. FAIL!. Thank you, Dimitrios Chronopoulos !. Q uantitative E valuation of E mbedded S ystems. Matrix equations - revisited Monotonicity Eigenvalues and throughput

azra
Download Presentation

Q uantitative E valuation of E mbedded S ystems

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Quantitative Evaluation of Embedded Systems QUESTION DURING CLASS? Email : qees3TU@gmail.com

  2. FAIL! Thank you, DimitriosChronopoulos!

  3. Quantitative Evaluation of Embedded Systems Matrix equations - revisited Monotonicity Eigenvalues and throughput Eigenvalues and the MCM

  4. Quantitative Evaluation of Embedded Systems Matrix equations - revisited Monotonicity Eigenvalues and throughput Eigenvalues and the MCM

  5. Recall! Cycles with a 0 execution time cause livelocks But when logging events, this is mathematically okay... y 0 ms

  6. Recall! A B Theorem: The number of tokens on any cycle is constant! Therefore, every cycle must contain at least one token, otherwise a deadlock occurs. y u C D 2ms 1ms 3ms 4ms

  7. Recall!

  8. Calculation rules

  9. Solving the (max,+) matrix equations Using induction

  10. Quantitative Evaluation of Embedded Systems Matrix equations - revisited Monotonicity Eigenvalues and throughput Eigenvalues and the MCM

  11. Monotonicity Theorem: (max,+) matrix addition is a monotone operator

  12. Monotonicity Given that x(1) = 0 and for all n : u(n) ≥ 0

  13. Best-case approximation… A x1 B C x3 y u 2ms 1ms x2 3ms

  14. Best-case approximation Tokens Theorem: (max,+) matrix addition is a monotone operator and as a consequence, removing the input gives a best-case approximation of behavior. Time (s)

  15. Quantitative Evaluation of Embedded Systems Matrix equations - revisited Monotonicity Eigenvalues and throughput Eigenvalues and the MCM

  16. Eigenvalues In linear algebra... a pair (x,λ) is an eigenpair for a matrix A if: with denoting elementwise multiplication and assuming that

  17. Eigenvalues In (max,+) algebra... a pair (x,λ) is an eigenpair for a matrix A if: with denoting elementwise addition and assuming that

  18. Eigenvalues In linear algebra, we find for any eigenpair (x, λ) that If (x, λ) is an eigenpair then so is (αx, λ), for any scalar α

  19. Eigenvalues And so, in (max,+) algebra we find for any eigenpair (x, λ) If (x, λ) is an eigenpair then so is (α+x, λ), for any shift α

  20. Find an eigenpair… A x1 B C x3 y 2ms 1ms x2 3ms

  21. Eigenvalues for a given eigenpair (x,λ) with x≤0

  22. Eigenvalues Tokens Time (s)

  23. Eigenvalues Tokens Theorem: The best-case throughput is always smaller than 1/λ, with λthe biggest eigenvalue of the associated matrix. Time (s)

  24. Quantitative Evaluation of Embedded Systems Matrix equations - revisited Monotonicity Eigenvalues and throughput Eigenvalues and the MCM

  25. Eigenvalues and the MCM A x1 B C x3 y 2ms 1ms x2 3ms

  26. Eigenvalues are cycle means A x1 B C Theorem: Every eigenvalue is the cycle mean of some cycle... x3 y 2ms 1ms x2 where Al(k,k) is the duration of some cycle with l tokens on it hence λ = Al(k,k)/l is a cycle mean. 3ms

  27. The MCM is an eigenvalue Let λ=MCM and xi represent a critical token... A x1 B Now given matrix A of a dataflow graph,let us construct the following two matrices: C x3 y 2ms 1ms x2 3ms In an entry (i,j) representsthe max duration of any path from i to j minus λ for each token on that path.

  28. The MCM is an eigenvalue Since xi represents a critical token: A x1 B So for the ithcolumn we find: Theorem: The maximal cycle mean of a graphis the maximum eigenvalue of its (max,+) matrix. C x3 y And consequently: 2ms 1ms x2 3ms In an entry (i,j) representsthe max duration of any path from i to j minus λ for each token on that path. So is an eigenpair of A

  29. Summary: (max,+) matrix addition is a monotone operator Thus removing the input gives a best-case approximation of behavior. Every eigenvalue is the cycle mean of some cycle... But not every cycle mean is an eigenvalue! The maximal cycle mean of a graphis the maximum eigenvalue of its (max,+) matrix and 1/MCM is the maximal throughput. See the book by [Baccelli, Cohen, Olsder and Quadrat] for more! Like for the question “can the 1/MCM throughput actually be achieved?”

  30. Simulate 6 firings

  31. Give the (max,+) matrix equations

  32. Calculate the MCM

  33. Determine a periodic schedule for arbitrary µ

  34. Plot the latency for a period µ

  35. Optimize the periodic schedule for µ = 15

  36. Optimize the periodic schedule for arbitrary µ

  37. Plot the delayed latency for a period µ

  38. Plot the minimal delay for a period µ

  39. Optimize the periodic schedule for arbitrary µ

More Related