optimal strategies for maintaining a chain of relays
play

Optimal strategies for maintaining a chain of relays between an - PowerPoint PPT Presentation

Optimal strategies for maintaining a chain of relays between an explorer and a base camp Lukas Humbel 2. Mai 2012 1/55 2/55 3/55 4/55 5/55 6/55 7/55 8/55 9/55 Outline Model Definition 1 Problem Statement Time/Relay Model What to


  1. Optimal strategies for maintaining a chain of relays between an explorer and a base camp Lukas Humbel 2. Mai 2012 1/55

  2. 2/55

  3. 3/55

  4. 4/55

  5. 5/55

  6. 6/55

  7. 7/55

  8. 8/55

  9. 9/55

  10. Outline Model Definition 1 Problem Statement Time/Relay Model What to measure Manhattan Hopper Strategy 2 Strategy Description Static Scenario Performance Dynamic Scenario Performance Conclusion 3 10/55

  11. Problem Statement 11/55

  12. Problem Statement v n + 1 v 0 v 1 . . . Grid size: 0 . 5 Transmission distance: 1 11/55

  13. Problem Statement Grid size: 0 . 5 Transmission distance: 1 11/55

  14. Problem Statement Grid size: 0 . 5 Transmission distance: 1 11/55

  15. Time/Relay Model 12/55

  16. Time Model Synchronized Look – Compute – Move 12/55

  17. Relay Model - Sensory Input α d 1 d 2 Sees its chain neighbors Memoryless No communication 13/55

  18. Relay Model - Sensory Input α d 1 d 2 Sees its chain neighbors Memoryless No communication 13/55

  19. Relay Model - Sensory Input α d 1 d 2 Sees its chain neighbors Memoryless No communication . . . must sense when predecessor has stepped 13/55

  20. Relay Model - Movement Moves with constant speed 14/55

  21. Relay Model - Movement Moves with constant speed Can be removed everywhere Inserted only at home 14/55

  22. Chain Attributes Valid condition Optimal condition 15/55

  23. What to measure 16/55

  24. Static Scenario Explorer fixed Quality measurement: Time to optimal chain 16/55

  25. Dynamic Scenario Chain in optimal condition Explorer moving Quality measurement: Possible speed of explorer Maximal chain length 17/55

  26. What can we expect? Dynamic Scenario Explorer can move as fast as a relay constant 18/55

  27. What can we expect? Dynamic Scenario Explorer can move as fast as a relay constant Chain length? O ( minimal length ) 18/55

  28. What can we expect? Dynamic Scenario Explorer can move as fast as a relay constant Chain length? O ( minimal length ) Static Scenario There are cases where a (constant speed moving) relay needs n timesteps to get close to the direct line. 18/55

  29. Strategy Description 19/55

  30. Manhattan Hopper All stations move on a grid Chain remains valid Relays move at most constant distance Uses Manhattan distance 19/55

  31. Manhattan Hopper All stations move on a grid Chain remains valid Relays move at most constant distance Uses Manhattan distance d = ∆ x + ∆ y 19/55

  32. Manhattan Hopper Description v n + 1 v 0 v 1 . . . Executed sequentally. v i + 1 moves after v i One sequence is called a run 20/55

  33. Manhattan Hopper Description Neighbors not in line → move Neighbors in line → stay 21/55

  34. Manhattan Hopper Description v i + 2 v i + 1 v i If v i moves to v i + 2 . v i + 1 and v i + 2 are removed. 22/55

  35. Manhattan Hopper Description v i + 2 v i + 1 v i If v i moves to v i + 2 . v i + 1 and v i + 2 are removed. v i + 1 and v i + 2 are removed. 22/55

  36. Manhattan Hopper Description v i If v i moves to v i + 2 . v i + 1 and v i + 2 are removed. v i + 1 and v i + 2 are removed. A remove operation ends the run. 22/55

  37. A little example 23/55

  38. Static Scenario Performance 24/55

  39. Static Scenario Theorem 1 After n runs, the chain has optimal length 24/55

  40. Configuration � u i = position ( v i + 1 ) − position ( v i ) 25/55

  41. Configuration � u i = position ( v i + 1 ) − position ( v i ) C = ( ⇒ , ⇑ , ⇑ , ⇒ , ⇒ , . . . , ⇒ , ⇑ , ⇑ ) = ( � u 0 , � u 1 , . . . , � u k ) 25/55

  42. Configuration � u i = position ( v i + 1 ) − position ( v i ) C = ( ⇒ , ⇑ , ⇑ , ⇒ , ⇒ , . . . , ⇒ , ⇑ , ⇑ ) = ( � u 0 , � u 1 , . . . , � u k ) u i and � � u j are oppositional ↔ � u i = − � u j 25/55

  43. Configuration � u i = position ( v i + 1 ) − position ( v i ) C = ( ⇒ , ⇑ , ⇑ , ⇒ , ⇒ , . . . , ⇒ , ⇑ , ⇑ ) = ( � u 0 , � u 1 , . . . , � u k ) u i and � � u j are oppositional ↔ � u i = − � u j Optimal (Manhattan) length configuration? 25/55

  44. Static Scenario - Strategy Effects On Configuration Lemma 2 Let C = ( � u 0 , � u 1 , � u 2 . . . , � u k ) . Assume a run finishes without removing any relay. C ′ = ( � u 1 , � u 2 , . . . , � u k , � u 0 ) is the configuration after the run. Also afterwards � u 0 is not oppositional to any other. 26/55

  45. Static Scenario - Strategy Effects On Configuration C = ( ⇒ , ⇑ � �� � , ⇑ , ⇒ , ⇒ , . . . , ⇒ , ⇑ , ⇑ ) 27/55

  46. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇒ , ⇑ � �� � , ⇒ , ⇒ , . . . , ⇒ , ⇑ , ⇑ ) 28/55

  47. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇑ , ⇒ , ⇒ � �� � , ⇒ , . . . , ⇒ , ⇑ , ⇑ ) 29/55

  48. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇑ , ⇒ , ⇒ , ⇒ � �� � , . . . , ⇒ , ⇑ , ⇑ ) 30/55

  49. Static Scenario - Strategy Effects On Configuration If � u 0 is oppositional to any other � u i , � u 0 will meet it at some point C = ( . . . , ⇒ , ⇐ � �� � , . . . ) triggers a removal 31/55

  50. Static Scenario - Strategy Effects On Configuration Lemma 3 Let C = ( � u 0 , � u 1 , � u 2 . . . , � u k ) . The run finishes with removing v i and v i + 1 if and only if u i + 1 is the first vector oppositional to � � u 0 . C ′ = ( � u 1 , � u 2 , . . . , � u i , � u i + 2 , . . . � u k ) is the configuration after the run. 32/55

  51. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇑ , ⇒ , ⇒ , ⇒ , ⇓ , . . . , ⇑ , ⇑ , ⇒ ) � �� � 33/55

  52. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇑ , ⇒ , ⇒ , ⇒ , ⇓ , . . . , ⇑ , ⇑ , ⇒ ) � �� � 34/55

  53. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇒ , ⇑ , ⇒ , ⇒ , ⇓ , . . . , ⇑ , ⇑ , ⇒ ) � �� � 35/55

  54. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇒ , ⇒ , ⇑ , ⇒ , ⇓ , . . . , ⇑ , ⇑ , ⇒ ) � �� � 36/55

  55. Static Scenario - Strategy Effects On Configuration C = ( ⇑ , ⇒ , ⇒ , ⇒ , ⇑ , ⇒ , ⇓ , . . . , ⇑ , ⇑ , ⇒ ) � �� � C ′ = ( ⇑ , ⇒ , ⇒ , ⇒ , ⇒ , ⇑ , ⇓ , . . . , ⇑ , ⇑ , ⇒ ) � �� � C ′′ = ( ⇑ , ⇒ , ⇒ , ⇒ , ⇒ , . . . , ⇑ , ⇑ , ⇒ ) 37/55

  56. Static Scenario - Strategy Effects On Configuration Lemma 3 Let C = ( � u 0 , � u 1 , � u 2 . . . , � u k ) . The run finishes with removing v i and v i + 1 if and only if u i + 1 is the first vector oppositional to � � u 0 . C ′ = ( � u 1 , � u 2 , . . . , � u i , � u i + 2 , . . . � u k ) is the configuration after the run. 38/55

  57. Static Scenario - Some Observations Vectors are never created, label them uniquely C 1 = ( � a 0 , � a 1 , . . . , � a k ) 39/55

  58. Static Scenario - Some Observations Vectors are never created, label them uniquely C 1 = ( � a 0 , � a 1 , . . . , � a k ) In every run � u i ( i � = 0) reduces its position at least by one Case 1: No removal Case 2: Removal happens and � u i is before the removal Case 3: Removal happens and � u i is after the removal 39/55

  59. Static Scenario Assume after n runs, there is an oppositional pair � u p and � u q with p < q . C = ( . . . , u p , . . . , u n ) � �� � Distance: n − p At most n − p + 1 runs earlier, � u p was at position 0 and hence would have been removed. 40/55

  60. Static Scenario Assume after n runs, there is an oppositional pair � u p and � u q with p < q . C = ( . . . , u p , . . . , u n ) � �� � Distance: n − p At most n − p + 1 runs earlier, � u p was at position 0 and hence would have been removed. After n rounds, there are no more oppositional pairs. 40/55

  61. Static Scenario It takes n rounds to reach minimal length. Timesteps? 41/55

  62. Static Scenario It takes n rounds to reach minimal length. Timesteps? Pipeline! Start new run every 3 time steps. 41/55

  63. Static Scenario It takes n rounds to reach minimal length. Timesteps? Pipeline! Start new run every 3 time steps. After 3 n + n = 4 n time steps the chain is optimal 41/55

  64. Dynamic Scenario Performance 42/55

  65. Dynamic Scenario Must handle explorer moves 42/55

  66. Dynamic Scenario Must handle explorer moves Perform Follow run Then perform Hopper run The Hopper run is what we have seen before 42/55

  67. Follow Run Explorer moves 43/55

  68. Follow Run Relays follow 43/55

  69. Follow Run Base inserts new relay 43/55

  70. Follow Run 43/55

  71. Dynamic Scenario Performance Lemma 4 Let the chain have optimal length prior to the explorer’s movement. Then after the explorer’s movement, the Hopper and Follow run bring the chain to an optimal length. 44/55

Recommend


More recommend