prolonging network lifetime prolonging network lifetime
play

Prolonging Network Lifetime Prolonging Network Lifetime by Cross- - PDF document

Eyes - energy efficient sensor networks Prolonging Network Lifetime Prolonging Network Lifetime by Cross- by Cross -layer Optimization layer Optimization Energy efficient sensor networks Tim Nieberg, Jian Wu, Lodewijk van Hoesel Overview


  1. Eyes - energy efficient sensor networks Prolonging Network Lifetime Prolonging Network Lifetime by Cross- by Cross -layer Optimization layer Optimization Energy efficient sensor networks Tim Nieberg, Jian Wu, Lodewijk van Hoesel Overview Cross-layer Approach EMACs TDMA-based, self-organizing MAC-scheme Connected Active Set Identify nodes that are needed for connectivity (“ACTIVE” nodes) Other nodes can follow sleeping pattern (“PASSIVE” nodes) ESR (Eyes Source Routing) On-demand, dynamic routing protocol Limited flooding to reduce routing overhead during dynamic changes in topology energy efficient sensor networks 2 1

  2. Eyes - energy efficient sensor networks ������� ������������ Self-organizing, TDMA-based MAC protocol Nodes can autonomously chose time slot No base stations needed Collision-free communication Supports efficient transmission of short multicast messages Used in clustering , routing etc. Scalable, adaptive for network topology Allows sleep patterns Acknowledgements of messages is decided on higher protocol layers energy efficient sensor networks 3 ��������������������� Timeslot CR TC Data CR TC energy efficient sensor networks 4 2

  3. Eyes - energy efficient sensor networks ��������������������� Timeslot CR TC Data CR TC Traffic Control Traffic Control • ��������������������������� � ������������������������������������ ��������� � ������������������������������� � ����������������������������������� ���������������� �������� � ��������������������� energy efficient sensor networks 5 ��������������������� Timeslot CR TC Data CR TC Data Data • ������������������������������ � ���������������������������������� �������������� � ����������������������� energy efficient sensor networks 6 3

  4. Eyes - energy efficient sensor networks ��������������������� Timeslot CR TC Data CR TC Communication Request Communication Request • ������������������������ ��������������������������� � ������������������� energy efficient sensor networks 7 ���������������������������� Nodes choose time slots locally TC-Section contains list of neighboring slots Controlled time slot ...00101 1 0... Result of occupied time slots for ? 6 ...1111110... ...01 1 0110... ...1001 1 10... 3 ...0111 1 00... 5 Free time slot 5 ? ? 7 4 4 2 ...010 1 100... ...100 1 100... 1 ...1 1 10100... ... 1 10110... 2 = Active node, that claimed time slot 2 ? = New active node in the network energy efficient sensor networks 8 4

  5. Eyes - energy efficient sensor networks ���������� 50 nodes in 800x500 m area One sink, 5 sensing nodes Other nodes are “intermediate relays” Random Waypoint model 2-10 m/s Transmission range: 150m Metric: Network Lifetime Physical model of RFM Time until 30% of relay nodes run out of energy Comparison to S-MAC energy efficient sensor networks 9 ������� �������������������� Static network topology DSR routing 1,6 EMACs Relative network lifetime 1,4 SMAC 1,2 EMACs gives a 20-50% 1 increase of network 0,8 lifetime 0,6 0,4 0 5 9 13 19 25 38 75 150 300 EMACs performs msg/min comparable for static and dynamic topology Dynamic network topology 1,6 EMACs Relative network lifetime 1,4 SMAC SMAC suffers from: 1,2 Increased number of 1 routing messages 0,8 Increased listen interval 0,6 In mobile scenario 0,4 0 5 9 13 19 25 38 75 150 300 msg/min energy efficient sensor networks 10 5

  6. Eyes - energy efficient sensor networks ������������������� EMACs supports three energy modes: Active: Node controls a time slot Node can communicate collision-free Passive: Node does NOT control a time slot Node uses CR of other (active) nodes No collision-free communication Dormant: Node shuts down for agreed interval Not considered in simulations… Passive mode saves energy… … but how to decide? energy efficient sensor networks 11 �������������������� Local decision algorithm to create connected subset of nodes which will remain active Passive nodes can follow sleeping pattern Each node is neighbor to non-passive node Idea: 1. Identify Anchor Nodes Nodes form Independent, Dominating Set (IDS) 2. Introduce Bridges for connectivity May have to use distributed bridge consisting of 2 intermediate nodes energy efficient sensor networks 12 6

  7. Eyes - energy efficient sensor networks ���������������������� ��������� Distributed Bridge Anchor Bridge Passive energy efficient sensor networks 13 ���������������������� ��������� “Mesh”-like backbone of active nodes energy efficient sensor networks 14 7

  8. Eyes - energy efficient sensor networks ���������������������� (Re-) Decision on status is invoked at wake-up when topology changes Active nodes only Node will listen to one more frame before becoming passive Detect current changes of neighbors energy efficient sensor networks 15 ������������������� No extra transmissions needed (!) TC-Section of EMACs contains all necessary information Extra Field: AID (active id) to encode all information Description Encoding Anchor AID = ID Bridge AID = (Anchor1 XOR Anchor2) 1 Undecided Active AID = 0 Nonmember AID = ID (lowest Anchor) 1) 1 st bit of ID is always 0 -- if 1 st bit of AID = 1 then node is bridge All control-messages can be inferred (=> not needed) energy efficient sensor networks 16 8

  9. Eyes - energy efficient sensor networks ��������������������� 2’500 nodes, different network densities energy efficient sensor networks 17 ������� �������������������������� EMACs w passive Static network topology Static scenario SMAC EMACs 1,8 Roles are not rotated 1,6 Relative network lifetime No dramatic increase of 1,4 lifetime 1,2 1 0,8 Mobile scenario 0,6 0,4 Energy consumption is 0 5 9 13 19 25 38 75 150 300 more evenly distributed msg/min in the network EMACs w passive Dynamic network topology SMAC Roles are changed EMACs 2,9 Relative network lifetime 2,4 >120% of network lifetime 1,9 increase!! 1,4 0,9 0,4 0 5 9 13 19 25 38 75 150 300 msg/min energy efficient sensor networks 18 9

  10. Eyes - energy efficient sensor networks ���������������������������������� On Demand driven routing algorithm three phases Setup : routes created only needed by a source node Maintenance : existing routes maintained by route maintenance procedures Reestablishment : reconstruct route when maintenance fails fast route recovery relying on MAC route optimization during maintenance locally restricted flooding with high efficiency only two neighbor ’s ID stored in the node all routing messages with short fixed size characteristics: low overhead, can handle mobility and failures energy efficient sensor networks 19 Route Setup Request Reply destination only replies first initial flooding with very short request request greatly reduced energy consumption only the best route is confirmed fitted into the TC section of EMAC nodes on the best route record nodes only forward the first received the best neighbor to the destination request node store the best neighbor to the source nodes Not on the best route node release the route info Source Source Destination Destination energy efficient sensor networks 20 10

Recommend


More recommend