Sensor Networks: Directed Diffusion and Other Proposals
Sensor Networks: Directed Diffusion and Other Proposals
ECE 256
Sensor Networking Why ??
Data Collection A basic need
Will the volcano erupt? Need temperature/gas signatures
Are poles melting due to GW? Need ocean current data
How many enemy tanks crossed through the jungle?
Did anyone enter my house while I was away?
3
San Fransiscos Moscone Center equipped with sensor network
4
Sensor Hardware
(Glimpse)
5
Sensor Nodes
Motivating factors for emergence
Applications
Moores Law in chips, MEMS
Advances in wireless technology
Challenges
Battery technology lagging
Solar cell
CMOS or III-V
1-2 mm 7
Example Hardware
Size
Golem Dust: 11.7 cu. mm
MICA motes: Few inches
Spec, 3/03
4 KB RAM
4 MHz clock
19.2 Kbps, 40 feet
Supposedly $0.30
Chemical
CO, CO2, radon
Biological
pathogen detectors
10
Berkeley Family of Motes
11
Sensor Software
(TinyOS Glimpse)
12
Programming TinyOS
Use a variant of C called NesC
NesC defines components
A component is either
A module
A module can be a Clock or LED
Or an user-defined software module
Or a configuration
set of other components wired together
Specifying the unimplemented methods invocation mappings
On your PC
Write NesC program
Compile to an executable for the mote
Plug the mote into the parallel port through a connector board
Install the program
On the mote
Turn the mote on, and its already running your application
14
TinyOS component model
Component specifies:
Internal Tasks Internal State
Commands Events
17
Sensor-node Operating System
Size of code and run-time memory footprint
Embedded System OSs inapplicable
Need hundreds of KB ROM
Workload characteristics
Continuous ? Bursty ?
18
TinyOS: Summary
Matches both
Hardware requirements
power conservation, size
Application requirements
diversity (through modularity), event-driven, real time
19
AdHoc and Sensors
Ad Hoc network lacking killer applications
Difficult to force co-operation among HUMAN users
Mobility/connectivity unreliable for a business model
Difficult to bootstrap critical mass required
21
This Talk
Directed Diffusion
Focusing on the shift from the ad hoc paradigm
The attention to energy conservation
SMAC
Energy-Aware Medium Access Control
22
Directed Diffusion
23
The Problem
A region requires event-
A sensor field
monitoring (harmful gas,
vehicle motion, seismic Sensor sources
vibration, temperature, etc.)
Event
Deploy sensors forming a
distributed network
Directed
Diffusion
On event, sensed and/or
processed information
delivered to the inquiring Sensor sink
destination
24
The Proposal
Proposes an application-aware paradigm to
facilitate efficient aggregation, and delivery of
sensed data to inquiring destination
Challenges:
Scalability
Energy efficiency
Robustness / Fault tolerance in outdoor areas
Efficient routing (multiple source destination pairs)
25
IP or not to IP
IP is the pivot of wired/wireless networks
All networking protocol over and below IP
Comments ?
26
Directed Diffusion
Typical IP based networks
Requires unique host ID addressing
Application is end-to-end, routers unaware
27
Data Naming
Expressing an Interest
Using attribute-value pairs
E.g., Type = Wheeled vehicle // detect vehicle location
Interval = 20 ms // send events every 20ms
Duration = 10 s // Send for next 10 s
Field = [x1, y1, x2, y2] // from sensors in this area
28
Gradient Set Up
Inquirer (sink) broadcasts exploratory interest, i1
Intended to discover routes between source and sink
29
Exploratory Gradient
Exploratory Request
Gradient
Event
Low Low
Low
30
Event-data propagation
Event e1 occurs, matches i1 in sensor cache
e1 identified based on waveform pattern matching
31
Reinforcement
Reinforced gradient
Reinforced gradient
Event
A sensor field Sink
32
Path Failure / Recovery
Link failure detected by reduced rate, data loss
Choose next best link (i.e., compare links based on
infrequent exploratory downloads)
Negatively reinforce lossy link
Either send i1 with base (exploratory) data rate
Or, allow neighbors cache to expire over time
Link A-M lossy
Event D A reinforces B
Src
M B reinforces C
A D need not
C A () reinforces M
B Sink
M () reinforces D
33
Loop Elimination
P Q
D M A
35
Average Dissipated Energy
flooding
Multicast Diffusion
flooding
Diffusion
Multicast
0%
10%
20%
39
Questions?
40
An Energy-Efficient MAC Protocol for
Wireless Sensor Networks (S-MAC)
Overhearing
Control Overhead
Idle Listening
Listening to possible traffic that is not sent
50%-100% energy drain compared with receiving
42
Avenues to Reduce Energy Consumption
43
(1) Periodic Listen and Sleep
44
Listen/Sleep Schedule Assignment
Follower
Listen Listen for a mount of time
B td Go to sleep after time t- td Sleep Hear SYNC from A, follow
As SYNC
Rebroadcasts SYNC after
Broadcasts
random delay td
45
Listen/Sleep Schedule Assignment
B td Sleep
Broadcasts
46
Keeping Clocks in SYNC
SYNC packets must not collide
Reserve separate time window for SYNC transmission
47
(2) Collision Avoidance
Identical to 802.11
RTS/CTS
Virtual carrier sense (NAV)
Physical carrier sense
48
(3) Overhearing Avoidance
Neighbors go to sleepon overhearing RTS/CTS
A is talking to B
D receives CTS from B -> sleep
Ds transmission will collide Bs
C receives RTS from A -> sleep
C cannot receive CTS/DATA from E
50
Acknowledgment to Pro. Jun Yang
51
Message Passing
Advantages:
Energy saving:
Neighbors go to sleep when sense transmissions
Reduces control overhead by sending multiple ACK
Disadvantage:
Node-to-node fairness reduces
However, message-level latency reduces
52
Listen time: 300ms
Experiment Sleeping time: 1s
SYNC: every 13s (10 listen/sleep period)
A, B, C use the same schedule
53
Energy save due
to periodic sleep 802.11
SMAC
55
SYNC overhead
56
Questions?
57
Energy Efficient Routing in Ad Hoc Disaster
Recovery Networks:
An Application Perspective
58
Motivation
Disasterrecovery emerging application for
adhoc/sensor networks
During Sep 11 attacks survivors were detected
through mobile phone signals
People often buried below earthquake disaster
59
Problem
Given some pkt generation rate at each badge
Design routing strategy that maximizes network
lifetime
60
Approach
Existing simplex techniques can be used to
solve the problem
Computation intensive due to several iterations for
convergence
61
Summary
Complexity of O(n3logT)
n3 for finding a feasible assignment of flows
Log T for the binary search
62
Other Research Challenges in Sensors
Coverage
Union of all sensing ranges need to cover entire region
Time synchronization
Data Aggregation
Calculating functions over a spatial distribution of sensors
Data Dissemination
Rumour routing, Ant colonies, swarm intelligence
Motion tracking, object guiding
Sensors + Actuators mobile robots !!!
63
Questions?
64
Message Complexity
Grid topology
N = 25
n = 5 Sources
m = 3 sinks
Nodes talk with
Adj. or diagonal
nodes
TinyOS:
Event-driven execution (reactive mote)
Size
Accommodate diverse set of applications (plug n play)
67
TinyOS Facts
Software Footprint 3.4 KB
Concurrency support:
At peak load CPU is asleep 50% of time
69
TinyOS: Size
70
Contribution
Network addressing is data centric
Probably correct approach for sensor type applications
Application-awareness a beneficial tradeoff
Data aggregation can improve energy efficiency
Better bandwidth utilization
Notion of gradient (exploratory and reinforced)
Fault tolerance
Implementation on Berkley motes
Network API, Filter API
71
Critique
Choice of path does not maximize aggregation
Least delay path does not max aggregation
Exploratory paths improve fault tolerance
But at the cost of additional msg./energy overhead
Overhead analysis omits the exploratory paths
Data overlap can be suppressed
2 sources, reporting overlapping data can be combined
Idle energy = 10% of receive, 5% of transmit
Explains the poor energy performance of flooding
Not realistic numbers optimistic assumption
72
Rumor Routing
LEACH
SPIN
73
Rumor Routing
74
LEACH
Proposes clustering of sensors + cluster leaders
Can aggregate data in single (local) cluster
Rotating cluster head balances energy consumption
Cluster formation distributed and energy efficient
Cluster-head
always awake
75
LEACH The Protocol
Time is divided into rounds
A node self-elects itself as the cluster head
Higher residual energy, higher probability to be head
Close-by sensors join this cluster-head
Cluster head does TDMA scheduling and gathers data
Gathered data compressed based on spatial correlation
Transmits data to Base Station (@ higher power)
In the next round, another cluster head elected
Probabilistic load balancing
Network lifetime can increase manifolds
76
SPIN: Information Via Negotiation
Flooding many sensors transmit same data
Redundant
Make sensors disseminate spatially/temporally
disjoint data sets
Name data with meta-data to define space/time property
Sensors compare overheard data with self-sensed data
Combine data to minimize overlap
Make sensors resource-adaptive
When low battery perform minimum activities
77
The SPIN 3-Step Protocol
78
The SPIN 3-Step Protocol