Power Aware Routing in Mobile Ad-Hoc Networks

1
Power Aware Routing in Mobile Ad-Hoc Networks
31st Oct, 2002

• Sumit I Eapen
• - Joy Ghosh

2
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

3
Introduction Power Concerns

• The lifetime of a network is defined as the time
  it takes for a fixed percentage of the nodes in a
  network to die out.
• Portability of wireless nodes being critical its
  almost mandatory to keep the battery sizes to a
  bare necessary
• Since battery capacity is thus fixed, a wireless
  mobile node is extremely energy constrained
• Hence all network related transactions should be
  power aware to be able to make efficient use of
  the overall energy resources of the network

4
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

5
Traditional routing metrics

• Aims to minimize hop counts and propagation delay
• Fails to take into account the power usage of
  nodes
• Results in poor lifetime of networks

6
Power Aware Metrics

• Intuition
• conserve power and share cost of routing packets
  to ensure increase in life of node and network
• Metrics
• 1. Minimize energy consumed / packet
• 2. Maximize time to Network Partition
• 3. Minimize variance in node power levels
• 4. Minimize cost / packet
• 5. Minimize maximum node cost

7
1. Minimize energy consumed / packet

• Definitions
• T(a,b) energy consumed in transmitting and
  receiving one
• packet over one hop from a to b
• ej Sk-1i1 T(ni, ni1) total energy spent
  for packet j
• Goal
• Minimize ej for all packets j
• Note
• In lightly loaded networks this automatically
  finds shortest hop path
• In heavily loaded networks due to contention it
  might not be shortest

8
2. Maximize time to network partition

• Definition
• Cut Set set of nodes that divide the network
  into two partitions
• As soon as one node in the set dies the delay
  experienced increases
• Goal
• - To balance load of the nodes in the Cut Set to
  maximize network life
• Problems
• The problem is similar to scheduling tasks to
  multiple servers so that
• the response time is minimized, which is
  known to be NP-complete

9
3. Minimize variance in node power levels

• Goal
• To keep all nodes up and running together for as
  long as possible
• Concept
• Build a route that takes into account the amount
  of data waiting to be transmitted in all the
  intermediate nodes
• Merit
• Achieve some kind of load balancing to ensure
  similar rates of dissipation of energy throughout
  the network

10
4. Minimize cost / packet

• Definition
• Total cost of sending packet j
• cj Sk-1i1 fi (xi)
• Where,
• xi is the energy dissipated in node i till now
• fi(xi ) is the cost of node i
• fi(xi) 1 / (1 g(xi))
• Where
• g(xi) is the normalized battery capacity
• Goal
• - Minimize cj for all packets j

11
4. Minimize cost / packet (contd.)

• Advantage
• - The remaining batter power level is
  incorporated into the routing decision
• This also balances load by avoiding usage of weak
  nodes in presence of stronger ones
• Network congestion can be taken care of by
  increasing node cost in presence of contention.

12
5. Minimize maximum node cost

• Definition
• Ci(t) cost of routing a packet through node i
  at time t
• C(t) maximum of the Ci(t)s
• Goal
• - Minimize C(t), for all t gt 0
• Side effects
• Delays node failure
• Reduces variance in node power levels

13
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

14
MANET Routing Protocols

• Broad Classifications
• Proactive Protocols
• Table Driven
• Frequent topology updates
• Each node knows about all destinations
• Distance Vector, Link State Routing, etc.
• Reactive Protocols
• On Demand
• A node learns of other nodes through actual
  communications
• DSR, AODV, etc

15
Low Power Routing - I

• Transmission Power
• P (i, j) is the Link Cost defined as the power
  expended for transmitting and receiving a packet
  between two consecutive nodes i and j
• Minimize Si,j?path P (i, j)
• Fixed transmit power
• P(i,j) b x packet_size c
• Where b packet size dependent energy
  consumption
• And c fixed cost for MAC layer control
  negotiation
• Varying transmit power
• P(i,j) k x daij
• Where dij distance between i and j
• And a parameter depending on physical
  environment

16
Low Power Routing - II

• Remaining Battery Power
• Ri(t) is the remaining power of node i at time t
• Simple Approach
• Minimize Si?path 1/Ri(t)
• Min-Max Approach
• Avoid routes with nodes having minimum battery
  capacity among all nodes in all possible routes
• Conditional Min-Max Approach
• Till all nodes in route have energy above a
  threshold, choose route with minimum total
  transmission power
• As energy falls below threshold, use the min-max
  algorithm suggested above

17
Power-Aware Source Routing (PSR)

• This is a Reactive (On demand) protocol based on
  DSR
• Cost Function
• The cost of route p at time t is C (p,t)
• C (p,t) Si?p Ci(t)
• where Ci(t) is the cost of node i at time t
• Ci(t) ?i . Fi/ Ri(t)a
• ?i transmit power of node i
• Fi full-charge battery capacity of node i
• Ri(t) remaining battery power of node i at time
  time t
• a a positive weighting factor
• This Cost function takes into account both
  transmission power and remaining battery power

18
PSR Route Discovery

• RREQ broadcast initiated by source
• Intermediate nodes can reply to RREQ from cache
  as in DSR
• If there is no cache entry, receiving a new RREQ
  an intermediate node does the following
• Starts a timer
• Keeps the path cost in the header as Min-cost
• Adds its own cost to the path cost in the header
  and broadcast
• On receiving duplicate RREQ an intermediate node
  re-broadcasts it only if the following is true
• The timer for that RREQ has not expired
• The new path cost in the header is less than
  Min-cost
• Destination also waits for a specific time after
  the first RREQ arrives
• It then replies to the best seen path in that
  period and ignores others that come later
• The path cost is added to the reply and is cached
  by all nodes that hear the reply

19
PSR Route Discovery Illustration
5
12
5
7
t2
11 _at_ t1
3
t1
13, 11
2
8
9 _at_ t3
8
S
D
t3
6
reply to 11
4
2
2
20
PSR Route Maintenance

• Node mobility
• Connections between some nodes on the
  path are lost due to their movement. In this case
  a new RREQ is issued and the corresponding entry
  in the cache is purged.
• Energy Depletion
• Energy of some intermediate node maybe
  depleting very quickly. This can be addressed in
  two ways
• Semi-global approach
• Here the source monitors the remaining battery
  level of the path by periodically polling the
  intermediate nodes
• Local approach
• Each intermediate node is allowed to
  send back a route error at time t if the
  following condition is met

21
PSR Route Cache Invalidation

• Once the cost of a node has increased beyond the
  threshold for a particular route, all cache
  entries to various destinations are invalidated
• However if a path was newly added to the cache,
  the node makes some allowance by lowering the
  threshold by some normalized amount for
  forwarding packets only in that path
• Invalidated routes are purged from cache after
  some time
• A node can use an invalidated route for its own
  message initiations but not for relaying other
  nodes packets

22
PSR vs DSR Simulation on NS(2)

• Test bed of 20 nodes confined in 1000 x 1000 m2
  area
• Range of each node is 250 m
• 100 reliable and random ftp connections
• Average duration of connection is 20 sec
• Total simulation time 10000 sec
• Speed of movement is 10 m/s
• Random mobility with pause time of 4 sec

23
PSR vs DSR network lifetime
24
PSR vs DSR varying threshold
25
PSR Points to Ponder

• Threshold timers increase latency
• Destination has to wait gt blocking nature
• The choice of the time-out period is critical
• Route invalidation based on the cost increase
  threshold is also a sensitive decision
• Too low can force frequent route discoveries
• Too high can over use a node in a path

26
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

27
Local Energy-Aware Routing (LEAR)

• Aims to balance energy consumption with shortest
  routing delays
• Takes into account a nodes willingness to
  participate in the routing path which is based on
  its remaining battery power
• Destination does not wait to reply gt
  non-blocking
• Efficient use of route cache

28
The basic LEAR Algorithm

• Source uses a sequence number for new request
• If it gets no reply back it increases the
  sequence number and re-broadcasts

29
LEAR Basic Algorithm

• Problems
• Cannot utilize route cache in the basic form
  since upstream nodes cannot freely decide on
  behalf of downstream nodes
• May incur repeated route request messages due to
  dropping of requests by intermediate nodes in
  cascade
• Solutions four additional routing control
  messages
• DROP_ROUTE_REQ
• ROUTE_CACHE
• DROP_ROUTE_CACHE
• CANCEL_ROUTE_CACHE

30
LEAR DROP_ROUTE_REQ

• The Cascading effect
• Say the path is A -gt B -gt C1 -gt C2 -gt D
• Each of the intermediate nodes say have low
  energy
• On 1st request from A to D, B will drop request
  and adjust threshold
• On 2nd request from A to D, C1 will drop and
  adjust, and so on
• D will finally get the request on 4th attempt
• DROP_ROUTE_REQ
• On 1st attempt from A to D, B drops and adjusts
  itself and also forwards DROP_ROUTE_REQ along the
  path to D
• This causes C1 and C2 to adjust their threshold
• D will receive the request on the 2nd attempt

31
LEAR ROUTE_CACHE

• Destination may receive multiple ROUTE_REQ and
  ROUTE_CACHE
• It replies to only the first one

32
LEAR DROP_ROUTE_CACHE CANCEL_ROUTE_CACHE

• On receiving CANCEL_ROUTE_CACHE from C1, B
  invalidates that entry

33
LEAR Complete Algorithm
34
LEAR Simulation on GloMoSim

• Test bed of 40 nodes confined in 1000 x 1000 m2
  area
• Range of each node is 250 m
• 5 Constant Bit Rate source and destination pair
  chosen
• 1024 byte packets sent every sec for a specified
  duration
• Total simulation time 500 sec
• Random waypoint mobility
• Speed of movement is 5 m/s
• Pause time is varied from 50 to 400 sec
• Simulation results shown next are average of 100
  runs
• Initial Threshold value set to 90 of nodes
  initial power
• The value of adjustment d is taken as 0.1 or 0.4

35
LEAR Standard Deviation of energy distribution

• Energy Consumption measured at radio layer
• 35 improved energy balance with high mobility
  (50 sec pause time)
• 10 improvement with moderate mobility (400 sec
  pause time)
• The d value does not affect much

36
LEAR Ratio of accepted ROUTE_REQ

• Ratio total route_reqs accepted / total
  route_reqs received
• Even DSR does not have 100 ratio due to TTL
• d 0.1 drops requests more frequently due to
  lower adjustment

37
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

38
Geographical Energy Aware Routing (GEAR)

• Mostly appropriate for static data-centric sensor
  networks
• The basic concept comprises of two main parts
• Route packets towards a Target region through
  geographical and energy aware neighbor selection
• Disseminate the packet within the region
• The concept of the 1st part can also be applied
  to mobile ad-hoc networks

39
GEAR Energy aware neighbor computation

• Each node N maintains state h(N,R) which is
  called learned cost to region R
• Each node infrequently updates neighbor of its
  cost
• When a node wants to send a packet, it checks the
  learned cost to that region of all its neighbors
• If the learned cost of a neighbor to a region is
  not available, the estimated cost is computed as
  follows
• c(Ni, R) xd(Ni, R) (1-x)e(Ni)
• Where,
• x tunable weight,
• d(Ni, R) normalized distance of neighbor to
  region
• e(Ni) normalized consumed energy at node i

40
GEAR Packet forwarding

• When a node wants to forward a packet to a
  destination, it checks to see if it has any
  neighbor closer to destination than itself
• In case of multiple choices it aims to minimize
  the learned cost h(Ni, R)
• It then sets its own cost to
• h(N, R) h(Ni, R) C(N, Ni)
• Where,
• C(N, Ni) combination of remaining energy of N
  and Ni and the distance between them

41
GEAR Forwarding around holes

• Incase there are no neighbors closer to
  destination than itself, the node forwards to the
  neighbor with the least learned cost
• It updates its own cost accordingly
• So next time it wont lie in the route to that
  region

42
GEAR Discussions on hole avoidance

• If the length of the path from S to T is n, the
  learned cost will converge after S delivers n
  packets to same target T
• Convergence of learned cost only affects
  efficiency of hole avoidance not its correctness
• Propagating learned cost further upstream through
  the update procedure will enable earlier chances
  to avoid holes

43
GEAR Dissemination

• Once the target region is reached the packets are
  disseminated within the region by recursive
  geographic forwarding
• Forwarding stops when a node is the only one in a
  sub-region

44
GEAR Drawback I

• Inefficient Transmission
• Recursive geographic forwarding vs. Restricted
  flooding

45
GEAR Drawback II

• Non-Termination
• When network density is low compared to (sub)
  target region size

46
GEAR proposed solution

• Node degree is used as a criteria to
  differentiate low density networks from high
  density ones
• Choice of restricted flooding over recursive
  geographic forwarding is made accordingly

47
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

48
Minimum Energy Wireless Network

• What is Minimum Energy Network?
• -- It is a network where there is a path
  from node i to j
• that consumes the least transmission
  power
• . Minimum Energy Network Design
• --given a set of wireless nodes, for each
  node find a selected set of nodes called
  neighbors, set a directed link from the node to
  its neighbor (enclosure graph)
• --design an algorithm that will do the above
  function
• --protocol is distributed
• Design first for a stationary wireless network
  and then extend it to a mobile scenario

49
Minimum Energy Network Power Losses

• 1. Transmission loss which is proportional to dn
  where d is the distance between transmitter and
  receiver. n gt 2
• 2. Receiver power loss constant C.
• 3. CPU computation loss negligible.
• Due to 1 above, it can be seen that relaying
  packets through intermediate nodes might save
  energy instead of directly transmitting packets.

50
Relaying Concept

• Relay through b if tdnab tdnbc C lt tdnac

• Relay Region
• R i-gtr of the transmit-relay node pair
  (i,r) is
• R i-gtr (x,y) P i-gtr-gt(x,y) lt P
  i-gt(x,y)
• e.g, Ra-gtb c in the above example

51
Relay Region
52
Neighbors

• Neighbors N(i) of a node i are those nodes that
  do not fall in the relay region of any other node
  with respect to i
• Ei nkeN(i) Rc i-gtk n DN
• N(i) n e N(xn,yn) e Ei, n ? i
• Enclosed Node
• A node i is said to be enclosed if it has
  communication links to each of its neighbors and
  to no other node.

53
Algorithm to find the Enclosure Graph

• The distributed protocol to find the enclosure
  graph consists of two steps
• for each node i, find its neighbors
• set up directional links from each node to all
  its neighbors
• This graph is strongly connected
• Search for Neighbors (Phase 1)
• A search algorithm is used to determine the above
• Each node sends a signal to its search region.
  This signal contains the position of the node.
  
• The node also listens to signals. When it
  receives the signals it can find the relay region
  of the corresponding node.

54
Algorithm (contd.)

• Nodes found in the search fall into two
  categories.
• Alive nodes
• Dead nodes
• When the search algorithm terminates for node i
  then the set of alive nodes is the set of
  neighbors for node i.
• The only outgoing communication links from i will
  be to these set of alive nodes.

55
Determining Paths (Phase II)

• Apply an algorithm similar to bell ford to
  enclosure graph
• Lets assume that all nodes wish to find the
  minimum power path to a particular node called
  the Master node
• Path Determination
• Each node broadcasts its cost to its neighbors
• The cost of a node i is defined as the minimum
  power necessary for it to reach the master node
• Each node finds minimum cost it can attain given
  costs of its neighbors.
• If n e N(i), when i receives the information
  cost(n), it computes
• Ci,n Cost(n) Ptrans(i,n) Preceiver(n)
• Cost(i) min neN(i) Ci,n
• Picks the link corresponding to this minimum cost
  neighbor

56
Distributed Mobile Network

• Protocol developed so far was for a stationery
  network
• Localized nature of the search algorithm makes it
  applicable to mobile scenarios too
• Here each node periodically executes phase 1 and
  phase 2.
• This time interval should not be too large or too
  small
• Thus the protocol can be made self
  reconfigurable.
• Demerit of Minimum Energy Networks
• The remaining battery power is not taken
  into consideration.

57
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

58
Low Energy Adaptive Clustering Hierarchy (LEACH)

• In this we consider a micro-sensor network where
• 1. The base station is fixed and located
  far from sensors
• 2. All nodes are homogeneous and energy
  constrained
• Key features of LEACH
• Localized coordination and control for cluster
  setup and operation
• Randomized rotation of the cluster heads and the
  corresponding clusters.
• Local compression to reduce global compression

59
LEACH - Algorithm Details

• Operation of Leach broken into rounds
• Round
• Set-up phase
• Advertisement phase
• Cluster Set-up Phase
• Schedule Creation
• Data transmission
• Steady-state phase

60
Advertisement Phase

• Each node decides whether or not to become a
  cluster head for a round based on a threshold.
• Each node say node n generates a random number
  between 0 and 1. If the random number is less
  than a threshold T(n) then the node elects itself
  to be a cluster head.
• T(n) P / ( 1 P(r mod 1/p)) if n e G
• 0
  otherwise
• P desired percentage of cluster heads (P
  0.05)
• r current round
• G is the set of nodes that have not been
  cluster head in last 1/P rounds

61
Advertisement Phase (contd.)

• Each node that elects itself cluster-head for
  current round broadcasts a message to the rest of
  the nodes
• All cluster-heads transmit their advertisement
  with the same transmit energy
• Non cluster heads keep their receivers on
• Based by the received signal strength, each
  non-cluster node decides to which cluster head to
  join( assuming symmetric propagation channels)

62
Cluster Set up Phase

• Each non-cluster-head node informs the
  cluster-head to whom it wants to join.
• During this phase all heads should keep their
  receivers on
• Schedule Creation
• Each cluster head based on the number of nodes
  in its cluster creates a TDMA schedule which is
  broadcasted to its cluster

63
Data Transmission

• Radios of non-heads are off when its not
  transmitting, to preserve energy.
• When all data has been received from all the
  nodes the head performs signal processing to
  compress the data into a single signal
• This is then send directly to the base station by
  a high energy transmission.

64
Direct Transmission vs- LEACH
65
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

66
Sensor Protocols For Information via Negotiation

• A family of adaptive protocols that efficiently
  disseminate information among sensors in a energy
  constrained wireless sensor network.
• Uses Meta-data high level data descriptor
• Meta-data negotiations to eliminate redundant
  information
• Why data dissemination? classic flooding can be
  used but has 3 demerits
• Implosion
• Overlap
• Resource Blindness

67
Implosion Example
68
Overlap Example
69
SPIN Negotiation Resource Management

• To overcome the problem of implosion and overlap,
  SPIN nodes negotiate before they transmit data.
• To negotiate in an energy efficient manner
  meta-data is used
• Nodes use a resource manager to find out their
  battery reserves
• If low then they cut back on certain activities
  like forwarding third party information.

70
SPIN MESSAGES

• ADV new data advertisement. When a node has new
  data to send it sends an ADV that contains the
  meta-data
• REQ this is in response to a ADV. This contains
  the meta-data that it wants
• DATA data message. This contains the actual
  sensor data that the REQ asked for. It has a meta
  data header.

71
SPIN1 3 way handshake
72
Energy Dissipation Comparison
73
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

74
Hierarchical Power Aware Routing

• Discusses about an online power aware routing
  algorithm in large sensor networks
• Path selection takes into consideration both the
  transmission power and the minimum battery power
  of the node in the path. It tries to compromise
• Makes use of zones to take care of the large
  number of sensor nodes

75
HPAR - Definitions

• Pmin power of the path with minimal power
  consumption
• P(Vi) initial power of node Vi
• Pt(Vi) power of node Vi at time t
• eij energy to transmit message between node i
  and j.
• Utij residual power fraction
• Utij (Pt(Vi) - eij) / P(Vi)

76
HPAR max-min zPmin Algorithm

• Find the path with the least power consumption,
  Pmin by using the Dijkstra algorithm
• Find the path with least power consumption in the
  graph.
• If the power consumption is greater than
  zPmin or no path is found, then the previous
  shortest path is the solution.
• Find the minimal utij on that path, let it be
  umin.
• Find all the edges whose residual power fraction
  utij is no greater than umin, remove them from
  the graph.
• Goto 1.

77
HPAR Empirical Experimental Analysis
78
HPAR - Zone Based Routing

• Max-min zPmin algorithm requires accurate power
  level information for all nodes in the network
• This is not feasible for a large network with
  lots of nodes
• So the whole network is divided into a small
  number of zones
• Each message is routed across zones using the
  information of the power estimate for the zones

79
HPAR - Zone Power Estimation

• Each zone has a controller node that polls each
  node in the zone for their power level
• Power estimation measures the number of messages
  that can flow through the zone
• Estimation is done relative to direction of
  message transmission
• Once the controller node determines the power
  estimate in each direction it broadcasts these to
  the other zones
• This is feasible because the number of zones is
  small

80
Zone Power Estimation
D
C
B
A
E
81
HPAR Power Graph
82
HPAR Zone Power Estimation Algorithm
83
HPAR - Global Path Selection
84
Local Path Selection

• The max-min zPmin algorithm is used directly to
  route a message within a zone.
• There could be multiple entry points into the
  zone and multiple exit points. So how are 2 paths
  in adjacent zones which are supposed to be part
  of a common global path connected.
• For this we associate a count with each node
  which tells how many times did a path start from
  the node when the power estimation in each
  direction was done.
• Then whenever we find paths we take the start and
  end node in each zone to be the ones the highest
  count.

85
HPAR Path Connection amongst Zones
86
Contents

• Introduction
• Metrics for power awareness
• Routing Protocols
• gt Power Source Routing (PSR)
• gt Local Energy Aware Routing (LEAR)
• gt Geographical and Energy Aware Routing
  (GEAR)
• gt Minimum Energy Mobile Wireless
  Networks
• gt Low Energy Adaptive Clustering
  Hierarchy (LEACH)
• gt Sensor Protocols for Information via
  Negotiation
• gt Hierarchical Power Aware Routing in
  Sensor Networks
• References

87
References - I
1 Power-Aware Routing in Mobile Ad Hoc Networks
Suresh Singh, Mike Woo, C.S. Raghavendra 1 Pow
er-aware Source Routing Protocol for Mobile Ad
Hoc Networks Morteza Maleki, Karthik Dantu, and
Massoud Pedram 2 Non-Blocking Localized Routing
Algorithm for Balanced Energy Consumption in
Mobile Ad Hoc Networks Kyungtae Woo, Chansu Yu,
Hee Yong Youn, Ben Lee 3 Hierarchical
Power-aware Routing in Sensor Networks Qun Li,
Javed Aslam, Daniela Rus 4 Minimum Energy
Mobile Wireless Networks Volkan Rodoplu, Teresa
H. Meng 5 A Location-aided Power-aware Routing
Protocol in Mobile Ad Hoc Networks Yuan Xue,
Baochun Li
88
References - II
6 Geographical and Energy Aware Routing a
recursive data dissemination protocol for
wireless sensor networks Yan Yu, Ramesh
Govindan, Deborah Estrin 7 Energy-Efficient
Communication Protocol for Wireless Microsensor
Networks - Wendi Rabiner Heinzelman, Anantha
Chandrakasan, Hari Balakrishnan 8 Adaptive
Protocols for Information Dissemination in
Wireless Sensor Networks - Wendi Rabiner
Heinzelman, Joanna Kulik, Hari Balakrishnan 9 GP
SR Greedy Perimeter Stateless Routing for
Wireless Networks Brad Karp, H.T.
Kung 10 Dynamic Source Routing in Ad Hoc
Wireless Networks David B. Johnson, David A.
Maltz
89
Thank You!!!

• 31st Oct, 2002