1. Chapter 6 Time Synchronization

2. Outline  6.1. The Problems of Time Synchronization  6.2. Protocols Based on Sender/Receiver Synchronization  Network Time Protocol (NTP)  Timing-sync Protocol for Sensor Networks (TPSN)  Flooding Time Synchronization Protocol (FTSP)  6.2.4. Ratio-based time Synchronization Protocol (RSP)  6.3. Protocols Based on Receiver/Receiver Synchronization  Reference Broadcast Synchronization (RBS)  Hierarchy Referencing Time Synchronization (HRTS)  6.4. Summary 2015/11/262

3. 6.1. The Problems of Time Synchronization 2015/11/263

4. The Problems of Time Synchronization  Why Need for Time Synchronization?  Many of the applications of WSN needs the event with time stamp  Ordering of the samples for reporting  Events are reported by multiple nodes  When WSN is energy save enabled, it need all nodes to be in sync in order to be in Idle or Active mode  Medium Access Layer (MAC) Scheduling (ex: TDMA)  Order of messages may change while transmission 2015/11/264

5. Sources of Inaccuracies  A local software clock of node i at time t L i (t) =  i H i (t) +  i  H i (t): hardware clock of node i at time t   i :clock drift rate of node i   i :phase shift of node i  Actual oscillators have random deviations from nominal frequency (drift, skew)  additional pulses or lost pulses over the time of one million pulses at nominal rate  Oscillator frequency is time variable  Long-term variation: oscillator aging  Short-term variation: environment (temperature, pressure, supply voltage,...) 52015/11/26

6. General Properties of Time Synchronization Algorithms  Physical time vs. logical time  External vs. internal synchronization  Global vs. local algorithms  Keep all nodes of a WSN synchronized or only a local neighborhood?  Absolute vs. relative time  Only accurate time difference  Sufficient to estimate the drift instead of phase offset 62015/11/26

7. General Properties of Time Synchronization Algorithms  Hardware vs. software-based mechanisms  A GPS receiver would be a hardware solution, but often too heavyweight/costly/energy-consuming in WSN nodes, and in addition a line-of-sight to at least four satellites is required  A-priori vs. a-posteriori synchronization  Is time synchronization achieved before or after an interesting event?  Post-facto synchronization: is triggered by an external event  Deterministic vs. stochastic precision bounds  Local clock update discipline  No backward jumps of local clocks  No sudden jumps 72015/11/26

8. Performance Metrics and Fundamental Structure  Metrics:  Precision: maximum synchronization error for deterministic algorithms, mean error /stddev /quantiles for stochastic ones  Energy costs, e.g. # of exchanged packets, computational costs  Memory requirements  Fault tolerance: what happens when nodes die? 82015/11/26

9. Fundamental Building Blocks of Time Synchronization Algorithms  Resynchronization event detection block:  when to trigger a time synchronization round?  Remote clock estimation block:  figuring out the other nodes clocks with the help of exchanging packets  Clock correction block:  compute adjustments for own local clock based on remote clock estimation  Synchronization mesh setup block:  figure out which node synchronizes with which other nodes 92015/11/26

10. Constraints for Time Synchronization in WSNs  Scale to large networks of unreliable nodes  Quite diverse precision requirements,  from ms to tens of seconds  Use of extra hardware is mostly not an option  Low mobility  Often there are no fixed upper bounds on packet delivery delay  Negligible propagation delay between neighboring nodes is negligible  Manual node configuration is not an option 102015/11/26

11. 6.2. Protocols Based on Sender/Receiver Synchronization 2015/11/2611

12. Protocols Based on Sender/Receiver Synchronization  In this kind of protocols, a receiver synchronizes to the clock of a sender  The classical Network Time Protocol (NTP) belongs to this class  We have to consider two steps: Pair-wise synchronization  How does a single receiver synchronize to a single sender?  Network wide synchronization  How to figure out who synchronizes with whom to keep the whole network / parts of it synchronized? 2015/11/2612

13. Network Time Protocol (NTP)  Synchronizing Physical Clocks  Computer Clocks in distributed system not in consistent  Need to synchronize clocks  External synchronization (ES)  Synchronized with an external reliable time source S  |S - C| < D, where C is computer’s clock  Internal synchronization (IS)  Synchronized with other computer in the distributed system  | C i - C j | < D  IS does not imply ES  Clock C i and C j may drift together  ES implies IS  Within bound 2D 2015/11/2613

14. Network Time Protocol (NTP)  Distributed System Type  Synchronous distributed system  Known upper bound on transmission delay  Simplifies synchronization  One process p 1 sends its local time t to process p 2 in a message m  p 2 could set its clock to t + T trans, where T trans is transmission delay from p 1 to p 2  T trans is unknown but min ≤ T trans ≤ max  Set clock to t + (max - min)/2 then skew ≤ (max - min)/2  Asynchronous distributed system  Internet is asynchronous system  T trans = min + x where x ≥ 0 2015/11/2614

15. Network Time Protocol (NTP)  Cristian’s method (1989) for an asynchronous system  A time server S receives signals from a UTC source  Process p requests time in m r and receives t in m t from S  p sets its clock to t - T round /2  Accuracy ± (T round /2 - min) :  because the earliest time S puts t in message m t is min after p sent m r.  the latest time was min before m t arrived at p  the time by S’s clock when m t arrives is in the range [t + min, t + T round - min]  T round is observed round-trip time  min is minimum delay between p and S mrmr mtmt p Time server S 2015/11/2615

16. Network Time Protocol (NTP)  Issues with Christian’s Algorithms  A single time server might fail, so they suggest the use of a group of synchronized servers  It does not deal with faulty servers  No authentication mechanism  Inaccuracy increases if the delay between messages is non- negligible 2015/11/2616

17. Network Time Protocol (NTP)  A time service for the Internet - synchronizes clients to UTC (Coordinated Universal Time) 1 2 3 2 33 Reliability from redundant paths, scalable, authenticates time sources Primary servers are connected to UTC sources Secondary servers are synchronized to primary servers Synchronization subnet - lowest level servers in users’ computers 2015/11/2617

18. Network Time Protocol (NTP)  Synchronisation of servers  The synchronization subnet can reconfigure if failures occur, e.g.  a primary that loses its UTC source can become a secondary  a secondary that loses its primary can use another primary  Modes of synchronization:  Multicast  A server within a high speed LAN multicasts time to others which set clocks assuming some delay (not very accurate)  Procedure call  A server accepts requests from other computers (like Cristiain’s algorithm). Higher accuracy. Useful if no hardware multicast.  Symmetric  Pairs of servers exchange messages containing time information  Used where very high accuracies are needed (e.g. for higher levels) 2015/11/26 18

19. Network Time Protocol (NTP)  Messages exchanged between a pair of NTP peers  All modes use UDP  Each message bears timestamps of recent events:  Local times of Send and Receive of previous message  Local times of Send of current message  Recipient notes the time of receipt ( we have T i-3, T i-2, T i-1, T i )  In symmetric mode there can be a non-negligible delay between messages TiTi T i-1 T i-2 T i-3 Server B Server A Time mm' Time 2015/11/2619

20. Network Time Protocol (NTP)  Accuracy of NTP  For each pair of messages between two servers,  NTP estimates an offset o i between the two clocks and a delay d i (total time for the two messages, which take t and t’)  T i-2 = T i-3 + t + o and T i = T i-1 + t’ - o  This gives us (by adding the equations) :  d i = t + t’ = T i-2 - T i-3 + T i - T i-1  Also (by subtracting the equations)  o = o i + (t’ - t )/2 where o i = (T i-2 - T i-3 + T i-1 - T i )/2  Using the fact that t, t’ >0 it can be shown that  o i - d i /2 ≤ o ≤ o i + d i /2.  Thus o i is an estimate of the offset and d i is a measure of the delay 2015/11/2620

21. Network Time Protocol (NTP)  Techniques to Improve Accuracy  NTP servers filter pairs, estimating reliability from variation, allowing them to select peers  High variability in successive pairs implies unreliable data  Accuracy depends on the delay between the NTP servers  Accuracy of 10s of millisecs over Internet paths (1 on LANs)  Peer selection  Lower layer peer favoured over higher layer server  Peer with lower synchronization dispersion is preferred  Synchronization dispersion is the sum of variability in data from the server to the root 2015/11/2621

22. LTS – Lightweight Time Synchronization  Overall goal:  Synchronize the clocks of all sensor nodes of a subset of nodes to one reference clock (e.g. equipped with GPS receivers)  Considers only phase shifts  Does not try to correct different drift rates 222015/11/26

23. LTS – Lightweight Time Synchronization  Two components:  Pair-wise synchronization:  based on sender/receiver technique  Network wide synchronization:  Minimum-height spanning tree construction with reference node as root 232015/11/26

24. 24 LTS – Pairwise Synchronization 2015/11/26

25. LTS – Pair-wise Synchronization  Assumptions:  Node i’s original aim is to estimate the true offset O = D(t 1 ) = L i (t 1 ) – L j (t 1 ), where L i (t j ) is the local software clock of node i at time t j  During the whole process the drift is negligible  the algorithm in fact estimates D(t 5 ) and assumes D(t 5 ) = D(t 1 )  Propagation delay τ and packet transmission delay t p are known to nodes i and j 252015/11/26

26. 26 Li(t5)Li(t5) 2015/11/26

27. 27 t 5 >= t 1 + τ +t p where τ :propagation delay t p :packet transmission time Li(t5)Li(t5) 2015/11/26

28. 28 t 5 <= t 8 - τ- t p where τ :propagation delay t p :packet transmission time L i (t 5 ) 2015/11/26

29. 29 The uncertainty is in the interval [L i (t 1 ) + τ + t p, L i (t 8 ) - τ – t p – (L j (t 6 ) – L j (t 5 )] Li(t5)Li(t5) 2015/11/26

30. LTS – Pair-wise Synchronization  Under the assumption that the remaining uncertainty is allocated equally to both i and j, node i can estimate L i (t 5 ) as 302015/11/26 This exchange takes two packets. If node j should also learn about the offset, a third packet is needed from i to j carrying O

31. LTS – Pair-wise Synchronization  Sources of inaccuracies:  Medium access delay  Interrupt latencies upon receiving packets  Delays between packet interrupts and time-stamping operation  Delay in operating system and protocol stack 312015/11/26

32. 322015/11/26

33. LTS – Pair-wise Synchronization  Improvements:  Let node i timestamp its packet after the MAC delay, immediately before transmitting the first bit  This would remove the uncertainty due to i’s operating system, protocol stack and the MAC delay!!  Let node j timestamp receive packets as early as possible, e.g. in the interrupt routine  This removes the delay between packet interrupts and time-stamping from the uncertainty, and leaves only interrupt latencies 332015/11/26

34. LTS – Pair-wise Synchronization – Error Analysis  Pairwise differences in time-stamping times at a set of receivers when time-stamping happens in the interrupt routine (Berkeley motes)  The motes raise an I/O pin at the same time they timestamp the packet  I/O signals were picked by an external logic analyzer  Another node send 160 pulse packets at random time  For each pulse packets, the difference of each of 10 possible receiver pairs are captured. 342015/11/26

35. LTS – Pairwise Synchronization – Error Analysis 35 Pair-wise difference in packet reception time (μsec) Number of trials 2015/11/26

36. LTS – Networkwide Synchronization  This way a spanning tree is created  But one should not allow arbitrary spanning trees  Consider a node i with hop distance h i to the root node  Assume that:  all synchronization errors are independent  Hence, a minimum spanning tree minimizes synchronization errors 362015/11/26

37. Timing-sync Protocol for Sensor Networks (TPSN)  Introduction  We present a Timing-sync Protocol for Sensor Networks (TPSN) that works on the conventional approach of sender-receiver synchronization  Pair-wise-protocol: time-stamping at node i happens immediately before first bit appears on the medium, and time-stamping at node j happens in interrupt routine 2015/11/2637

38. Timing-sync Protocol for Sensor Networks (TPSN)  Network Model  The network is “always-on”  Every node maintains 16-bit register as clock  Sensor has unique ID  Build hierarchical topology for the network  Node at level i can connect with at least one node at level i-1 2015/11/2638

39. Timing-sync Protocol for Sensor Networks (TPSN)  Level discovery Phase  Trivial  Synchronization Phase  Pair-wise sync is performed along the edge of hierarchical structure 2015/11/2639

40. Timing-sync Protocol for Sensor Networks (TPSN)  Level discovery Phase  The root node is assigned a level 0 and it initiates this phase by broadcasting a level_discovery packet  level_discovery packet contains the identity and the level of the sender  The immediate neighbors of the root node receive this packet and assign themselves a level (level = level +1)  This process is continued and eventually every node in the network is assigned a level. On being assigned a level, a node neglects any such future packets. This makes sure that no flooding congestion takes place in this phase 2015/11/2640

41. Timing-sync Protocol for Sensor Networks (TPSN)  Synchronization Phase  T1: A is sender, starting sync by sending synchronization_pulse packet to B  T2 = T1 + Δ + d, where  Δ is the clock offset  d is propagation delay  T3: B replies acknowledgement containing T1, T2, T3  T4: A receive ack. and T4 = T3 – Δ + d. So:  Δ = [(T2 － T1) － (T4 － T3)] / 2  d = [(T2 － T1) ＋ (T4 － T3)] / 2 2015/11/2641

42. Timing-sync Protocol for Sensor Networks (TPSN)  Synchronization Phase A B T1T1 T2T2 T1,T2,T3 T4T4 T1: A is sender, starting sync by sending synchronization_pulse packet to B with timestamp T1 T B receive the synchronization _pulse packet and ti2:mestamping immediately B replies acknowledgement containing T1,T2,T3 A receive an Ack and get timestamp T4 At time t1At time t4At time t2At time t3 2015/11/26 42

43. Timing-sync Protocol for Sensor Networks (TPSN)  Simulation and Comparison 2015/11/2643

44. Timing-sync Protocol for Sensor Networks (TPSN)  Simulation and Comparison 2015/11/2644

45. Flooding Time Synchronization Protocol (FTSP) 2015/11/2645

46. Flooding Time Synchronization Protocol (FTSP)  Introduction  The FTSP synchronizes the time to possibly multiple receivers utilizing a single radio message  Linear regression is used in FTSP to compensate for clock drift 2015/11/2646

47. Flooding Time Synchronization Protocol (FTSP)  Network Model  Every node in the network has a unique ID  Each synchronization message contains three fields:  TimeStamp  RootID  SeqNum  The node with the smallest ID will be only one root in the whole network 2015/11/2647

48. Flooding Time Synchronization Protocol (FTSP)  The root election phase  FTSP utilizes a simple election process based on unique node IDs  Synchronization phase 2015/11/2648

49. Flooding Time Synchronization Protocol (FTSP)  The root election phase  When a node does not receive new time synchronization messages for a number of message broadcast periods  The node declares itself to be the root  Whenever a node receives a message, the node with higher IDs give up being root  Eventually there will be only one root 2015/11/2649

50. Flooding Time Synchronization Protocol (FTSP)  Synchronization phase  Root and synchronized node broadcast synchronization message  Nodes receive synchronization message from root or synchronized node  When a node collects enough synchronization message, it estimates the offset and becomes synchronized node 2015/11/2650

51. B C Flooding Time Synchronization Protocol (FTSP) Root A Synchronized Node Unsynchronized node TimestamprootIDseqNumTimestamprootIDseqNumTimestamprootIDseqNum 2015/11/2651

52. Flooding Time Synchronization Protocol (FTSP)  Simulation and Conclusion 2015/11/2652

53. Ratio-based Time Synchronization Protocol (RSP) 2015/11/2653

54. Ratio-based Time Synchronization Protocol (RSP)  The RSP use two synchronization messages to synchronize the clock of the receiver with that of sender  The RSP also can extend to multi-hop synchronization  The nodes in the wireless sensor network construct a tree structure and the root of this tree is the synchronization root  The global time of the root is flooding out to the nodes through the tree structure 2015/11/2654

55. Ratio-based Time Synchronization Protocol (RSP)  The local clock time of a sensor device is provided by the quartz oscillator inside itself  Transformation formula between t and C i (t):   By (1), the local clock times of two sensor nodes i and j have the following relationship: : the local clock time of a sensor node i. t : the Coordinated Universal Time (UTC). : the drift ratio : the offset of node i’s clock at time t. (2) : relative drift ratio between nodes i and j : offset between the clocks of nodes i and j (1) 2015/11/2655

56. Ratio-based Time Synchronization Protocol (RSP) calculate the clock drift ratio θ = (T 3 − T 1 )/(T 4 −T 2 ). Reference node Sensor node 2015/11/2656

57. Each node can estimate the local time of reference node in the following way: Reference node Sensor node : the local time of sensor node :the corresponding local time of the reference node. : the initial offset between reference node and sensor node. (3) Ratio-based Time Synchronization Protocol (RSP) 2015/11/2657

58. It can be calculated using linear interpolation with the four time- stamps the can be derived as follows (4) Ratio-based Time Synchronization Protocol (RSP) 2015/11/2658

59.  Therefore, we can derive (5) from (3) and (4):  Each sensor node can estimate the local time of reference node, that is, the global time of the network (5) Ratio-based Time Synchronization Protocol (RSP) 2015/11/2659

60. RS (T1)(T1) Reference node Sensor node (T3)(T3) Ratio-based Time Synchronization Protocol (RSP) 2015/11/2660

61. 6.3. Protocols Based on Receiver/Receiver Synchronization 2015/11/2661

62. Protocols Based on Receiver/Receiver Synchronization  In this class of schemes  The receivers of packets synchronize among each other, not with the transmitter of the packet  Reference Broadcast Synchronization (RBS)  Synchronize receivers within a single broadcast domain  RBS does not modify the local clocks of nodes, but computes a table of conversion parameters for each peer in a broadcast domain 2015/11/2662

63. Reference Broadcast Synchronization (RBS)  Introduction  Reference broadcasts do not have an explicit timestamp  Receivers use reference broadcast’s arrival time as a point of reference for comparing nodes’ clocks  Receivers synchronizes with one another using the message’s timestamp (which is different from one receiver to another) 2015/11/2663

64. Reference Broadcast Synchronization (RBS)  Types of errors in traditional synchronization protocol  Send time latency  time spent at the sender to construct the message  Access time latency  time spent at the sender to wait for access to transmit the message  Prorogation time latency  time spent by the message in traveling from the sender to the receiver  Receive time latency  time spent at the receiver to receive the message from the channel and to notify the host 2015/11/2664

65. Reference Broadcast Synchronization (RBS)  Types of errors in RBS  Phase error  due to nodes’ clock that contains different times  Clock skew  due to nodes’ clock that run at different rates 2015/11/2665

66. Reference Broadcast Synchronization (RBS)  Difference between RBS & Traditional synchronization protocol  RBS  Synchronizes a set of receivers with one another  Supports both single hop and multi-hop networks  Traditional  Senders synchronizes with receivers  mostly supports only single hop networks 2015/11/2666

67. Reference Broadcast Synchronization (RBS)  The phase offset with the clock skew is estimated by:  Least-squares linear regression graph  From the best-fit line of the graph, following can be inferred:  Slope of the line : Clock skew of the nodes’ clock  Intercept of the line : Phase of the nodes’ clock 2015/11/2667

68. Reference Broadcast Synchronization (RBS)  Basic idea to estimate phase offset and clock skew for non-deterministic receivers:  Transmitter broadcasts m reference packets  Each of the n receivers records the time that the reference was received, according to its local clock  The receivers exchange their observation  Each receiver i can compute its phase offset to any other receiver j  Drift can be neglected when observations are exchanged quickly after reference packets  Drift can be estimated jointly with offset O when a number of periodic observations of O i,j have been collected 2015/11/2668

69. Reference Broadcast Synchronization (RBS) 69

70. Reference Broadcast Synchronization (RBS)  Formula for calculating the phase offset and clock skew of receiver r 1 with another receiver r 2 :  Let t r,b be r’s clock when it received broadcast b  for each pulse k that was received by receivers r 1 and r 2, we plot a graph : x = t r1, k y = t r2,k – t r1,k  Diagonal line drawn through the points represents the best linear fit to the data 2015/11/2670

71. Reference Broadcast Synchronization (RBS)  Diagonal line minimizes the residual error (RMS)  Therefore, we go for calculating the slope and intercept of the diagonal line  Time value of r 1 is converted to time value of r 2 by combining the slope and intercept data obtained 2015/11/2671

72. Reference Broadcast Synchronization (RBS) Step1: Transmitter broadcasts Step2: Receiver records its local clock, and exchange observation Step3:Use Least-squares linear regression to estimate phase offset B A Transmitter Receiver Reference Packet A:Local time B:Local time Finish RBS 2015/11/2672

73. Reference Broadcast Synchronization (RBS)  Communication costs:  Be m the number of nodes in the broadcast domain  First scheme: reference node collects the observations of the nodes, computes the offsets and sends them back  2 m packets  Second scheme: reference node collects the observations of the nodes, computes the offsets and keeps them, but has responsibility for timestamp conversions and forwarder selection  m packets 732015/11/26

74. Reference Broadcast Synchronization (RBS)  Communication costs:  Be m the number of nodes in the broadcast domain  Third scheme: each node transmits its observation individually to the other members of the broadcast domain  m (m-1) packets  Fourth scheme: each node broadcasts its observation  m packets, but unreliable delivery 742015/11/26

75. Reference Broadcast Synchronization (RBS)  Conclusion  Collisions are a problem: the reference packets trigger all nodes simultaneously to tell the world about their observations  Computational costs: least-squares approximation is not cheap!  Can be used without external timescales  Does not require tight coupling between sender and its network interface 2015/11/2675

76. Hierarchy Referencing Time Synchronization (HRTS) 2015/11/2676

77. Hierarchy Referencing Time Synchronization (HRTS)  Goal :  Synchronize the vast majority of a WSN in a lightweight manner  Idea  Combine the benefits of LTS and RBS 2015/11/2677

78. Hierarchy Referencing Time Synchronization (HRTS)  LTS : Lightweight Time Synchronization  Goal  Synchronize the clocks of all sensor nodes of a subset of nodes to one reference clock  It considers only phase shifts and does not try to correct different drift rates 2015/11/2678

79. Hierarchy Referencing Time Synchronization (HRTS)  LTS : Pairwise Synchronization n1n1 At time t 1 Sync packet n2n2 At time t 2 Record t 2 Reply packet Record t 1 At time t 3 At time t 4 Record t 4 Record t 3 In this packet contains t 2 and t 3 2015/11/2679

80. Hierarchy Referencing Time Synchronization (HRTS) 2015/11/2680

81. Hierarchy Referencing Time Synchronization (HRTS)  LTS : Pairwise Synchronization  Offset : O = Δ(t 5 )=L i (t 5 ) - L j (t 5 )= [L i (t 8 )+ L i (t 1 )- L j (t 6 )- L j (t 5 )] / 2  Benefit : only two packet transmissions with each pair  Benefit of RBS  Idea : ignore transmission delay  By this idea, one packet can synchronize every node in one hop  Combining the two protocol’s benefit, the HRTS finds good solution to synchronize nodes in hierarchical way 2015/11/2681

82. Hierarchy Referencing Time Synchronization (HRTS) 2015/11/2682

83. Hierarchy Referencing Time Synchronization (HRTS)  Timeline:  Root node triggers time synchronization at t 1 with timestamp L R (t 1 )  Node i timestamps packet at time t 2 with L i (t 2 ) and node j timestamps it at t 2 ’ with L j (t 2 ’ )  Node i formats a packet and timestamps it at time t 3 with L i (t 3 ) – the packet includes the values L i (t 2 ) and L i (t 3 )  Root node R timestamps the answer packet at time t 4 with L R (t 4 ) and computes its offset O R,i with node i ’ s clock  O R,i =L i (t 2 ) - L R (t 2 ) = L i (t 2 ) – (L R (t 1 ) + L R (t 4 ) – (L i (t 3 )- L i (t 2 ))/2 =[ (L i (t 2 ) - L R (t 1 )) – (L R (t 4 )- L i (t 3 ))]/2  Root node R broadcasts the values O R,i and L i (t 2 ) 83 2015/11/26

84. Hierarchy Referencing Time Synchronization (HRTS)  The root node R can estimate the offset O R,i between its own clock and the local clock i in a similar fashion as the protocol LTS  Root R broadcast the values O and L i (t 2 ) to all nodes  Node i simply subtract the offset O R,i from its local clock  Node j can compute O j,i directly as O j,i = L i (t 2 ) – L j (t’ 2 ) and O R,j = O R,i – O j,i 842015/11/26

85. Hierarchy Referencing Time Synchronization (HRTS)- Discussion  Node j is not involved in any packet exchange  by this scheme is possible to synchronize an arbitrary number of nodes to R ’ s clock with only three packets!!  The synchronization uncertainty comes from:  The error introduced by R when estimating O R,i  The error introduced by setting t 2 = t 2 ’  This makes HRTS only feasible for geographically small broadcast domains 852015/11/26

86. Hierarchy Referencing Time Synchronization (HRTS)- Discussion  Both kinds of uncertainty can again be reduced by:  timestamping outgoing packets as lately as possible (relevant for t 1 and t 3 )  timestamping incoming packets as early as possible (relevant for t 2, t 2 ’, t 4 )  The authors propose to use extra channels for synchronization traffic when late timestamping of outgoing packets is not an option  Rationale: keep MAC delay small 862015/11/26

87. Hierarchy Referencing Time Synchronization (HRTS) 2015/11/2687

88. Summary  Time synchronization is important for both WSN applications and protocols  Using hardware like GPS receivers is typically not an option, so extra protocols are needed  Post-facto synchronization allows for time synchronization on demand, otherwise clock drifts would require frequent resynchronization and thus a constant energy drain 2015/11/2688

89. Summary  Some of the presented protocols take significant advantage of WSN peculiarity like:  small propagation delays  the ability to influence the node firmware to timestamp outgoing packets late, incoming packets early 2015/11/2689

90. References [1] Ed. Ivan Stojmenovic, Handbook of Sensor Networks Algorithms and Architectures, 2005. [2] F. Sivrikaya,and B.Yener, Time Synchronization in Sensor Networks: A Survey,2004. (www.cs.rpi.edu/~yener/PAPERS/WINET/timesync04.pdf) [2] J. Elson, L. Girod, and D. Estrin,Fine-Grained Network Time Synchronization using Reference Broadcasts. (In Proceedings of the Fifth Symposium on OSDI 2002) [3] S. Ganeriwal, R. Kumar, and M. Srivastava, Timing-Sync Protocol for Sensor Networks. (SenSys ’03) [5] D. L. Mills. Network Time Protocol (Version 3) Specification, Implementation and Analysis. RFC 1305, 1992. [6] D. L. Mills. Improved Algorithms for Synchronizing Computer Network Clocks. IEEE/ACM Transactions on Networking, 3(3): 245–254, 1995. [7] D. L. Mills. Adaptive Hybrid Clock Discipline Algorithm for the Network Time Protocol. IEEE/ACM Transactions on Networking, 6(5): 505–514, 1998. 2015/11/2690

91. References [8] S. Ganeriwal, R. Kumar, S. Adlakha, and M. Srivastava. Network-Wide Time Synchronization in Sensor Networks. Technical Report NESL 01-01-2003, Networked and Embedded Systems Lab (NESL), University of California, Los Angeles (UCLA), 2003. [9] S. Ganeriwal, R. Kumar, and M. B. Srivastava. Timing-Sync Protocol for Sensor Networks. In Proceedings of the 1st ACM International Conference on Embedded Networked Sensor Systems (SenSys), pages 138–149, Los Angeles, CA, November 2003. [10] Miklós Maróti, Branislav Kusy, Gyula Simon, Ákos Lédeczi, The Flooding Time Synchronization Protocol, In Proceedings of the 2ed ACM International Conference on Embedded Networked Sensor Systems (SenSys), pages 39 – 49, Baltimore, MD, USA, 2004. [11] J.-P. Sheu, W.-K. Hu, and J.-C. Lin, Ratio-Based Time Synchronization Protocol in Wireless Sensor Networks, Telecommunication Systems, Vol. 39, No. 1, pp. 25-35, Sep. 2008. [12] J. Elson, L. Girod, and D. Estrin. Fine-Grained Network Time Synchronization using Reference Broadcasts. In Proceedings of the Fifth Symposium on Operating Systems Design and Implementation (OSDI 2002), Boston, MA, December 2002. [13] H. Dai and R. Han. TSync: A Lightweight Bidirectional Time Synchronization Service for Wireless Sensor Networks. ACM SIGMOBILE Mobile Computing and Communications Review, 8(1): 125–139, 2004. 2015/11/2691

92. Recommend Reading  Particular Challenges and Constraints for Time Synchronization Algorithms in WSN  J. Elson and K. R¨omer. Wireless Sensor Networks: A New Regime for Time Synchronization. In Proceedings of the First Workshop on Hot Topics In Networks (HotNets-I), Princeton, NJ, October 2002.  J. E. Elson. Time Synchronization in Wireless Sensor Networks. PhD dissertation, University of California, Los Angeles, CA, Department of Computer Science, 2003.  Other Time Synchronization Protocol  Lightweight time synchronization protocol (LTS)  J. V. Greunen and J. Rabaey. Lightweight Time Synchronization for Sensor Networks. In Proceedings of the 2nd ACM International Workshop on Wireless Sensor Networks and Applications (WSNA), San Diego, CA, September 2003. 2015/11/2692