reti wireless di sensori: problematiche generali ed esempi...
TRANSCRIPT
1
University of Bologna at CesenaDepartment of Electronics, Computer Science and Systems
Reti wireless di sensori:problematiche generali
ed esempi di applicazioni
Prof. Davide DardariIng. Matteo Lucchi
2
Prof. Davide Dardari, Ing. Matteo Lucchi
•Introduction to wireless sensor networks•Applications and main issues•Standards •Agricultural application example
Outline
2
3
Prof. Davide Dardari, Ing. Matteo Lucchi
Wireless Ages…
Past – Group CommunicationsOne wireless communication device per thousands of persons.
Present – Personal CommunicationsOne wireless communication device per person.
Future – Ambient CommunicationsEvery person using thousands wireless communication devices.
The surrounding environmentcommunicates…
Introduction to WSNs
4
Prof. Davide Dardari, Ing. Matteo Lucchi
The sensor might measure:temperaturepressure,light,acceleration,pollution,humidity,………..
SENSOR
POWERSUPPLY
CPU
COMMUNICATION
NODE
INFRARED
ACOUSTIC
SEISMIC
IMAGE
MAGNETIC…
ELECTRO-MAGNETICINTERFACE
8-bit, 10 MHzSlow Computations
128KB-1MBLimited Storage
1Kbps - 1Mbps, 3-100 Meters,
Lossy Transmissions
BatteryLimited Lifetime
What is a wireless sensor?
3
5
Prof. Davide Dardari, Ing. Matteo Lucchi
Target: low-size, low-cost, low-energy consumption
UC Berkeley: Smart Dust
JPL: Sensor Webs
Crossbow: Mote
What is a wireless sensor?
6
Prof. Davide Dardari, Ing. Matteo Lucchi
“A network of nodes that sense the environmentand may control it, enabling interaction betweenpeople and their personalised surroundings”
What is a wireless sensor network?
4
7
Prof. Davide Dardari, Ing. Matteo Lucchi
Wireless sensor networks are spatially distributed networks of small deviceswith sensing (also actuators) and communication (wireless) capabilities
Key technology to “map” the physical world into the internet space
A potentially huge number of possible applications
InternetLong-range link
short-range link
The vision: “Internet of Things”
8
Prof. Davide Dardari, Ing. Matteo Lucchi
• Environment & Habitat Monitoring
• Industrial Sensing• Traffic Control• Seismic Studies• Life Sciences• Infrastructure security• Health• Animal tracking• Planets exploration• Military
Circulatory Net
Application areas
5
9
Prof. Davide Dardari, Ing. Matteo Lucchi
THE QUAKE
It was in the early afternoon of an otherwise unremarkable Thursday thatthe Great Quake of 2053 hit Southern California.
…………………
From “Sensor Networks: A Bridge to the Physical World”By Jeremy Elson and Deborah Estrin,In “Wireless Sensor Networks”, Ed. Raghavendra, Sivalingam, Znati, Kluwer, 2004.
Why a Wireless Sensor Network?
10
Prof. Davide Dardari, Ing. Matteo Lucchi
The earth began to rupture several miles under the surface of an unihabitated part of the Mohave desert.
Decades of pent-up energy was violently released, sending huge shear waves speeding toward Los Angeles.
The quake was enormous, even by California standards, as its magnitude surpassed 8 on the Richter scale.
…………………………..
Mohave desert
The Quake (cont’d)
6
11
Prof. Davide Dardari, Ing. Matteo Lucchi
Residents had long ago understood such an event was possible.This area had been instrumented by scientists for more than a century.
By the turn of the century, the situation had improved considerably.Many seismometers were connected to the Internet …
If a sensor was close enough to the epicenter, and the epicenter was far enough from a population center,the alarm could be raised 20 or 30 seconds before the city started to shake.The idea was promising.
…Mohave desert
The Quake (cont’d)
12
Prof. Davide Dardari, Ing. Matteo Lucchi
Mohave desert
But, in the half century leading up to the Great Quake of 2053,technological advances changed everything.
By the mid 2040’s, the vast, desolate expanse of the desert floor was home to nearly a million tiny, self-contained sensors.
…
The Quake (cont’d)
7
13
Prof. Davide Dardari, Ing. Matteo Lucchi
It was just a few dozen of those seismometers – closest to the epicenter –that first sensed unusual acceleration in the ground.
As the number of confirmed observations grew, so did the likelihood that this event was not simply random noise. It was real!
In a few tenths of a second, the earth’s movement had the attentionof thousands of seismometers within a few miles of the epicenter.
The network soon reached consensus: this was an earthquake.It was a dangerous one.
………… Mohave desert
The Quake (cont’d)
14
Prof. Davide Dardari, Ing. Matteo Lucchi
The information hopped from one node to the next.After 41 miles, it finally reached the first sign of civilization: a wired communication access point.
Four seconds had passed since the quake began.
Once on the wired grid, the alarm spread almost instantly to every city.
The new generation of smart structures in Los Angeleslearned of the quake nearly thirty seconds before it arrived.
………. Mohave desert
The Quake (cont’d)
8
15
Prof. Davide Dardari, Ing. Matteo Lucchi
Tense moments passed.Sirens blared as every traffic light turned red,every elevator stopped and opened at the nearest floor …
Finally, the silence was broken, a low rumble, a deafening roar.The earth rolled and shook violently.
…….
Mohave desert
The Quake (cont’d)
16
Prof. Davide Dardari, Ing. Matteo Lucchi
The city could not completely escape damage. Many older homes collapsed.
Rescue crews arrived with Portable Emergency Survivor Locators.Each was a nylon package the size of a bottle containing thousands of tiny sensors that could disperse themselves as the package was thrown.
Back at the rescue truck, a map of the structure began to appear.People were visible as heat sources.
Chemical sensors began to detect abnormal traces of natural gas …
…Mohave desert
The Quake (cont’d)
9
17
Prof. Davide Dardari, Ing. Matteo Lucchi
By Monday, Southern California had returned to normal.The 2053 quake came and went, thanks largely to the pervasive sensors…
…
IS ALL THIS INVENTION, OR WILL IT BE REALITY?
Mohave desert
The Quake (cont’d)
18
Prof. Davide Dardari, Ing. Matteo Lucchi
To many people in 2003, using technology to prevent such a catastrophemust have seemed fanciful and improbable.
It seemed as improbable as 2003’s globally interconnected network of nearly a billion computers (the Internet!) must have seemed to those in 1953.
Indeed, perhaps even as improbable as 1953’s “electronic brain”must have seemed to those in 1903, who would soon experiencetheir own Great Earthquake in San Francisco.
WSNs: enablers of the Ambient Wireless Communications paradigm
The Quake (cont’d)
10
19
Prof. Davide Dardari, Ing. Matteo Lucchi
Circulatory Net Embed numerous distributed devices to monitor and interact with physical world: in work-spaces, hospitals, homes, vehicles, and “the environment” (water, soil, air…)
Sensing the world
20
Prof. Davide Dardari, Ing. Matteo Lucchi
Organize into Clusters Route
Wake/DiagnosisDeploy
WSN set-up example
11
21
Prof. Davide Dardari, Ing. Matteo Lucchi
random deployment no a priori topology knowledge - localization
long network lifetime (years) energy efficiency – distributed proc.
self-organization no centralized control - no human intervention
large number scalability - low cost
Information coverage & reliability connectivity - node density
WSN main issues and requirements
22
Prof. Davide Dardari, Ing. Matteo Lucchi
Environmental monitoring• Energy efficiency• Low duty cycle (network
synchronization)• Low data rate• One-way communication• Wireless backbone
Example of typical application requirements
Industrial• Reliability (e.g., interference)• Security• Interoperability (standards)• High data rate• Two-way communication (actuators)• Wired backbone
12
23
Prof. Davide Dardari, Ing. Matteo Lucchi
Where is energy consumed?
Energy efficiency
0.00
5.00
10.00
15.00
20.00
25.00
30.00M
CU
Act
ive
Mod
e, 1
6 M
Hz
MC
U A
ctiv
eM
ode,
32
MH
z
MC
U A
ctiv
e an
dR
X M
ode
MC
U A
ctiv
e an
dTX
Mod
e, 0
dBm
Pow
er m
ode
1
Pow
er m
ode
2
Pow
er m
ode
3
ADC
whe
nco
nver
ting
Energy Consumption (mA)
[Ref. ChipCon CC2430]
24
Prof. Davide Dardari, Ing. Matteo Lucchi
Physical layer • Low power circuit(CMOS, ASIC) design
• Optimum hardware/software function division
• Energy effective waveform/code design
• Adaptive RF power control
MAC sub-layer • Energy effective MAC protocol
• Collision free, reduce retransmission and transceiver on-times
• Intermittent (low duty-cycle), synchronized operation
• Rendezvous protocols
Link layerNetwork layer
Application layer
• FEC versus ARQ schemes; Link packet length adapt.
• Multi-hop route determination
• Energy aware route algorithm, network coding
• Route cache, directed diffusion, self-organizing algorithms• Video applications: compression and frame-dropping
• In-network data aggregation and fusion
See Jones, Sivalingam, Agrawal, and Chen survey article in ACM WINET, July 2001;See Lindsey, Sivalingam, and Raghavendra book chapter in Wiley Handbook of Mobile Computing,
Ivan Stojmenovic, Editor, 2002.
How to get energy efficiency
13
25
Prof. Davide Dardari, Ing. Matteo Lucchi
Supervisor
Node
Random Scenario
Direct communication with the supervisor is not energy efficient.
Far nodes tend to discharge prematurely since they must transmitwith higher power levels
Multi hop communication which routing protocol?
Multi-hop communication
26
Prof. Davide Dardari, Ing. Matteo Lucchi
Multi-Hop Routing issues:-Irregular Topologies-Self-organizing-Scalability-Fault tolerant-Energy efficient -Data transport aware-Connectivity-Coverage
Multi-hop communication
14
27
Prof. Davide Dardari, Ing. Matteo Lucchi
SVCH
CH
CH
At each round (supervisor interrogation):
•The SV triggers nodes•Each node has a probability x to self-elect cluster head (CH)•Each non-CH node selects the strongest CH •Each CH is responsible for data collection within its cluster area and for the transmission to the SV
LEACH-based clustered protocol
Self-organizing clustered network (example)
28
Prof. Davide Dardari, Ing. Matteo Lucchi
-Energy efficient self-organizing routing protocols (*)-Cooperative transmission for event detection (*)-Distributed spatial random process estimation through WSN in realistic environments (*)-Connectivity-Ultrawide bandwidth (UWB) techniques-Ranging and localization techniques-Time synchronization-Test and implementation on hardware (experimental activity)
Some research activities at the University of Bologna
15
29
Prof. Davide Dardari, Ing. Matteo Lucchi
Scenario A: supervisor (sink) located outside the sensor field
Scenario B: sink located inside the sensor field
L
0
dsink
= -50
dsink
= 50
Alan Campanini, Roberto Verdone and Davide Dardari “Performance of an Energy-efficient Multi Hop Protocol for Wireless Sensor Networks”
Example of routing protocols performance comparison
30
Prof. Davide Dardari, Ing. Matteo Lucchi
Network lifetime (scenario A)
16
31
Prof. Davide Dardari, Ing. Matteo Lucchi
Network lifetime (scenario B)
32
Prof. Davide Dardari, Ing. Matteo Lucchi
Estimation of a random spatial process(e.g., temperature, pressure, …)
Detection of a binary event (e.g., fire)
Environmental Monitoring application example
17
33
Prof. Davide Dardari, Ing. Matteo Lucchi
Idea: adoption of cooperative transmission techniques to achieve high energy efficiency (by exploiting diversity)
Fusion center (sink, collector)
Event
M. Lucchi, A. Giorgetti and M. Chiani, “Cooperative Diversity in Wireless Sensor Networks”, WPMC'05, Aalborg, Denmark, sept. 2005.
Detection of a binary event
34
Prof. Davide Dardari, Ing. Matteo Lucchi
Scenario A: each sensor node simplysends its own local decision independentlyto the fusion center.
Scenario B: nodes exchange and agree onthe common local information beforetransmitting to the fusion center
The “Hamlet dilemma”: to cooperate or not to cooperate?(… and how?)
Detection of a binary event
tN
10 / HH
tN
10 / HH
tN
tN
18
35
Prof. Davide Dardari, Ing. Matteo Lucchi
T. Q. S. Quek, D. Dardari, and M. Z. Win, “Energy efficiency of dense wireless sensornetworks: To cooperate or not to cooperate,” IEEE JSAC (Special Issue on Cooperative Communications and Networking), vol. 25, Feb 2007.
Depends on the event intensity, but not only on that ….... a lot of open problems are still waiting to be solved
Energy gain due to cooperation
Research activity in cooperation with MIT - USA
Event intensity
Detection of a binary event
36
Prof. Davide Dardari, Ing. Matteo Lucchi
Let be Z(s) the physical entity or signal under monitoring, such as temperature, pressure or electromagnetic field, at the position s in the 2/3-dimensional space.
Purpose: starting from the collected samples the supervisor tries to estimate the original function (process) by using a wireless sensor network in an efficient way.
Estimation of a random spatial process
19
37
Prof. Davide Dardari, Ing. Matteo Lucchi
Given a target process estimation accuracy
•How many nodes to deploy (node density)?
•Centralized vs distributed processing?
•How to route and aggregate the information?
Process estimation accuracy - Network lifetime (energy efficiency) tradeoff
Some questions:Estimation of a random spatial process
38
Prof. Davide Dardari, Ing. Matteo Lucchi
The analytical model developed
It takes into account•Random sampling theory (Point processes)•Clustered information routing•Distributed vs centralized processing•Energy consumption model•MAC protocol•Propagation effects (path-loss and shadowing)
D. Dardari, A. Conti, C. Buratti, and R. Verdone, “Mathematical evaluation of environmental monitoring estimationerror through energy-efficient wireless sensor networks,” IEEE Trans. Mobile Comput., vol. 6, jul 2007.
Estimation of a random spatial process
20
39
Prof. Davide Dardari, Ing. Matteo Lucchi
Nodes are randomly placed with a spatial density ρ in the monitored environment (e.g., nodes thrown by an airplane).
Supervisor
Node
RandomScenario
The sample space is finite owing to the limited WSN extension, and it defines a finite area A (sample space), inside which the process is observed. Thus, the actual (truncated) signal of interest is
( ) ( ) ( )= ⋅s s sAx z r
x(s) has finite energy, E0bandwidth per dimension B
1( )
0∈⎧
= ⎨⎩
ssA
Ar
otherwise
z(s) is a realization of the target process Z(s)
A
Building the model: spatial random sampling
40
Prof. Davide Dardari, Ing. Matteo Lucchi
Bit transmission is much more energy consuming than bit processing
It could be convenient to perform partial local processing at each node cluster (distributed processing) before data transmission to reduce the network throughput
•Node complexity constraints•Process estimation degradation (if lossy compr. algorithms adopted)
•Distributed compression techniques•Re-sampling techniques (at Nyquist rate)
Problems
Energy efficiency
21
41
Prof. Davide Dardari, Ing. Matteo Lucchi
10−4
10−3
10−2
10−1
100
101
ρ [m−2
]
10−4
10−3
10−2
10−1
100
ε
case 1case 2case 3Sim.: case 3
DDSP
no DDSP
The impact of DDSP on the estimation error (example)
Nodes density
Normalized MSE
Distributed Signal Processing (DDSP)
42
Prof. Davide Dardari, Ing. Matteo Lucchi
Mean node life duration without distributed signal processing (DDSP)
Nodes density
Number of roundsper Joule of charge
10−4
10−3
10−2
10−1
100
101
ρ [m−2
]
102
103
104
105
106
107
108
Nro
und
x=1 E−2, α=0.1x=1 E−3, α=0.1x=1 E−4, α=0.1x=1 E−2, α=0.001x=1 E−3, α=0.001x=1 E−4, α=0.001Sim.: x=1 E−3, α=0.1
Mean node life duration
22
43
Prof. Davide Dardari, Ing. Matteo Lucchi
Nodes density
10−4
10−3
10−2
10−1
100
101
ρ [m−2
]
102
103
104
105
106
107
108
Nro
und
x=1 E−2, α=0.1x=1 E−3, α=0.1x=1 E−4, α=0.1x=1 E−2, α=0.01x=1 E−3, α=0.01x=1 E−4, α=0.01
Number of roundsper Joule of charge
Mean node life durationMean node life duration with distributed signal processing (DDSP)
44
Prof. Davide Dardari, Ing. Matteo Lucchi
•a normalized estimation error not larger than 5%,•α = 0.1, x = 0.001,•c = 10, •β = 4 10E-6 m^(-2), •Nodes density ρ= 2 10E-3 m^(-2)
Requirements:
Without DDSP the mean number of rounds per Joule of charge is about 700.
With DDSP the mean number of rounds per Joule of charge is about6000!
Numerical example
23
45
Prof. Davide Dardari, Ing. Matteo Lucchi
… but a lot of problems are still open!
46
Prof. Davide Dardari, Ing. Matteo Lucchi
Other advanced topics for WSNs (1/2)
• Network coding– Distributed source coding (reduced network communications)– Distributed error correction to mitigate malicious errors injection
and node failures
Protocol domain
• Secure communication using RF channel reciprocity– Secret key generation and sharing using channel reciprocity– Enhanced security (cryptography)
24
47
Prof. Davide Dardari, Ing. Matteo Lucchi
Example: sail surface monitoring
Mainsail shift pressure over the sail section
sensor nodesover the sail
Pressure sensor
Sensed data are time-spatially correlated
Redundancy can be reduced through distributed source coding
Reduced network throughput longer lifetimeRef: Sortron - Senet s.r.l – Wireless Sensor Network Solutions (www.senet.it)
48
Prof. Davide Dardari, Ing. Matteo Lucchi
• Energy scavenging– solar power, thermal energy, wind energy, salinity gradients,
kinetic energy, RF sources
Technology domain
• Ultrawide bandwidth (UWB)– Higher data rate– Robustess to interference– High ranging accuracy (centimeters) localization
Other advanced topics for WSNs (2/2)
25
49
Prof. Davide Dardari, Ing. Matteo Lucchi
Example of new services
context aware applications (e.g., real-time location systems, RTLS)
* R. Das and P. Harrop ”RFID Forecast, Players and Opportunities 2007-2017”, 2007, http://www.idtechex.com.
new market opportunities ($6 Billion in 2017*)
50
Prof. Davide Dardari, Ing. Matteo Lucchi
- Theoretical characterization and modeling- Simulations- Measurements- Implementation on software/hardware
ICTICT
Electronic-Sensors-Embedded sys.-Low-power devices
TLC-Signal processing-Transmission techniques- ad-hoc networks-…………..
Computer Science- agents- distributed systems-…………
….but also
BiologyGeologyBiochemistryStructural EngineeringEnvironmental Engineering…………
Methodologies
26
51
Prof. Davide Dardari, Ing. Matteo Lucchi
Standards
Low power, reliability, low cost, scalability, ranging
Low power, reliability, low cost, scalability
Low costSpeedOptimized for
Star, Tree, MeshStar, Tree, MeshStarStarNetwork architecture
~ 50-100~ 50-100~ 10~ 100Range (meters)
850Kbit/s-27Mbit/s
250Kbit/s1-3Mbit/s54Mbit/sBit rate
Up to 65536Up to 655367~10-30Network size
30ms30msUp to 10 secUp to 3 secEnumeration latency
YearsYearsDaysHoursBattery life
(WPAN/WSN)Monitoring, control, localization
(WPAN/WSN) Monitoring & control
WPAN (Cable replacement)
WLAN (Web, Email, video)
Applications
ZigBee(UWB - IEEE 802.15.4a)
ZigBee(IEEE 802.15.4)
BluetoothWi-Fi(IEEE 802.11
52
Prof. Davide Dardari, Ing. Matteo Lucchi
• Proprietary choices ASK, 1-10 Kbit/s, UHF ISM bands
• Standards
IEEE 802.15.4bit rates below 250 Kbit/s, ISM bands, 2.4 GHz (see later)
IEEE 802.15.4a (2007)Ultrawide bandwidth (UWB)
range 10-50m, data-rate 850-27Mbit/s, low energy consumptionApplications: WPANs and WSNsRanging capability (precision<1m)Promising technology for its localization and high performance capabilitiesin dense multi-path environments
Physical layers for WSNs
both support Zigbee
27
53
Prof. Davide Dardari, Ing. Matteo Lucchi
• Low latency
• Short range
• Low data rate
• Low complexity
• Low power
• Robustness
• Security
• Low cost
Standard IEEE 802.15.4 (1/2)
54
Prof. Davide Dardari, Ing. Matteo Lucchi
Standard IEEE 802.15.4 (2/2)
It defines the physical layer (PHY) and medium access control (MAC) sublayer specifications for LR-WPAN (Low Rate – Wireless Personal Area Networks)
28
55
Prof. Davide Dardari, Ing. Matteo Lucchi
Three license-free frequency bands:
• 868 MHz – Europe – 20 Kbit/s – 1 channel
• 915 MHz – USA, Canada, Japan – 40 Kbit/s – 10 channels
• 2.4 GHz – ISM band – 250 Kbit/s – 16 channels
Physical layer (1/2)
56
Prof. Davide Dardari, Ing. Matteo Lucchi
• Preamble
• SDF - start of frame delimiter
• Packet lenght (max 127 byte)
• PSDU – payload from MAC layer
Two types of modulations:
• 868/915 MHz bands – BPSK – RF bandwidth of 600/1200 KHz respectively
• 2.4 GHz band – O-QPSK – RF bandwidth of 2 MHz -DSSS
Physical layer (2/2)
29
57
Prof. Davide Dardari, Ing. Matteo Lucchi
Frame control: packet identifier;
Data sequence number to ensure that packets are tracked;
Address: 16 or 64 bit
FCS (frame check sequence): cyclic code
Channel access scheme CSMA-CA (carrier sense multiple access with collision avoidance ).
MAC layer (1/2)
58
Prof. Davide Dardari, Ing. Matteo Lucchi
Four MAC frame types:• Beacon frame to synchronize devices• Data frame (max lenght 104 byte)• Acknowledgement frame• Command frame
MAC layer (2/2)
30
59
Prof. Davide Dardari, Ing. Matteo Lucchi
• Network setup – able to establish a new network
• Joining and leaving network – nodes are able to become members of the network as well as quit being members
• Configuration – ability of the node to configure its stack to operate in accordance with the network type
• Addressing – the ability of the coordinator to assign addresses to devices joining the network
• Synchronization – ability of a node to synchronize with another node by listening for beacons or polling for data
• Security – ability to ensure end-to-end integrity of frames
• Routing – nodes can properly route frames to their destination
Network Layer Responsibilities
60
Prof. Davide Dardari, Ing. Matteo Lucchi
Three type of nodes:• PAN coordinator. Initializes the network, manages network nodes
and stores network node information.
• FFD (Full Function Devices). Capable of functioning in any topology, capable of acting as a coordinator, capable of talking to any other device in network and generally line powered, Capable of performing energy detection (ED) and active scans
• RFD (Reduced Function Devices). Capable of communicating in a star topology and only to a FFD, cannot be a coordinator and generally battered powered
ZigBee provides a security toolbox to ensure reliable and securenetworks. Typical standard used AES (Advanced Encryption Standard).
ZigBee networks are primarily intended for low duty cycle sensornetworks (<1%).
Network Layer (1/2)
31
61
Prof. Davide Dardari, Ing. Matteo Lucchi
Three networking topologies:– Star (for very long battery life operation)– Mesh (high levels of reliability and scalability by providing more
than one path through the network)– Cluster tree (hybrid star/mesh topology that combines the
benefits of both for high levels of reliability and support for battery-powered nodes)
Network Layer (2/2)
62
Prof. Davide Dardari, Ing. Matteo Lucchi
We can distinguish three different multiple traffic type of networks:
• Periodic data. Data typically is handled using a beaconing system whereby the sensor wakes up at a set time and checks for the beacon, exchanges data, and goes to sleep (wireless sensor).
• Intermittent data. Data can be handled in a beaconless system or disconnected. In disconnected operation, the device will only attach to the network when communications is required, saving significant energy (e.g., wireless light switch).
• Repetitive low latency data. These applications may use the guaranteed time slot (GTS) capability. GTS is a method of QoS that allows each device a specific duration of time as defined by the PAN coordinator in the Superframe to do whatever it requires withoutcontention or latency (e.g., security system).
ZigBee applications
32
63
Prof. Davide Dardari, Ing. Matteo Lucchi
The IEEE 802.15.4a
As an amendment to 802.15.4 for an alternative PHY to provide (2007):
• High precision ranging/location capability• Ultra low power• Lower cost• High aggregate throughput• Scalability to data rates• Longer range
Asset tracking
Home automation
and.. WSNs, health care monitoring…
64
Prof. Davide Dardari, Ing. Matteo Lucchi
Physical layers
The amendment adds two new PHYs:
• 150 - 650 MHz and 3.1 - 10.6 GHz Ultra-Wide Band (UWB)– Impulse Radio (IR) based signaling– Mandatory nominal data rate of 1 Mbps– Modulation: Combination of BPM (burst position mod.) and BPSK
• 2450 MHz Chirp Spread Spectrum (CSS)
33
65
Prof. Davide Dardari, Ing. Matteo Lucchi
UWB PHY: Channels
A compliant UWB device shall be capable of transmitting in at least one of three specified bands.
• Sub-GHz Band: 150 – 650 MHz
• Low Frequency Band (LFB): 3244 - 4742 MHz
• High Frequency Band (HFB): 5944 – 10234 MHz
66
Prof. Davide Dardari, Ing. Matteo Lucchi
UWB-PHY: Bursts
• Each information-bearing symbol is represented by a burst of short time duration pulses
• The duration of an individual pulse is nominally considered to be the length of a chip
• Chip duration is equal to 2.02429 ns (Bandwidth > 494MHz)
34
67
Prof. Davide Dardari, Ing. Matteo Lucchi
UWB PHY: Modulation
Combination of burst position modulation (BPM) and binary phase shift keying (BPSK)
A UWB PHY symbol carries two bits of information: one bit for the position of a burst and the other bit for the phase (parity) of the burst.
• Each symbol can have 8, 16, 32, 64, and 128 possible slots for aburst.
• A symbol period is divided into two BPM intervals, and a burst occurs one of two intervals.
• A burst is formed by grouping 2n, n=0, 1, 2, ..., 9 consecutive chips.
68
Prof. Davide Dardari, Ing. Matteo Lucchi
The symbol structure
The UWB-PHY supports both coherent and non-coherent receivers.
35
69
Prof. Davide Dardari, Ing. Matteo Lucchi
UWB-PHY: Spreading
• The UWB PHY provides interference suppression through scrambling and time-hopping.
• Scrambling: each symbol contains a single burst of pulses, which are scrambled by a time-varying scrambling sequence.
• Time Hopping: each burst position varies from one symbol to the next according to a time hopping code.
70
Prof. Davide Dardari, Ing. Matteo Lucchi
UWB-PHY: Pulse shape
• The transmitted pulse shall have a cross correlation with a reference pulse r(t), whose main lobe is greater or equal to 1 to 0.8 for a duration of at least 0.5 ns and any side lobe shall be no greater that 0.3.
• The reference pulse r(t) is a root raise cosine pulse with roll-off factor of β = 0.6
• Optional pulse shapes are: Chirp on UWB (CoU) pulses, Continuous Spectrum (CS) pulses, Linear combination of pulses (LCP)
36
71
Prof. Davide Dardari, Ing. Matteo Lucchi
Commercial devices: ChipCon CC2430
Main characteristics:
• µcontroller 8051
• 32 KB flash memory, 8KB RAM, SFR (Special Function Register)
• No TinyOS Operative System OSAL
• 2.4 GHz frequency bands
• O-QPSK modulation
• CSMA-CA tecnique for MAC sublayer
• Index RSSI and LQI to evaluate the quality of communications
• 4 internal timers
• Low consumptions
72
Prof. Davide Dardari, Ing. Matteo Lucchi
• Battery or solar operation
• Soil moisture and temperature monitoring
• Insect monitoring and control
• Quickly reconfigurable, no sophisticated installation required
Agricultural application example
37
73
Prof. Davide Dardari, Ing. Matteo Lucchi
Remote monitoring (1/2)
• Not always it is possible to reach sensor field
• Sensors and sink have limited memory capacity
• It is necessary to guarantee a real-time monitoring
• Read data from Internet reduces the costs of the application
74
Prof. Davide Dardari, Ing. Matteo Lucchi
Remote monitoring (2/2)
38
75
Prof. Davide Dardari, Ing. Matteo Lucchi
Energy efficiency (1/2)
In environmental applications (and in many other applications) the purpose is to increase the lifetime of the network.
Two steps:
1. Creation of low power applications.
2. Use of photovoltaic panels to supply sensors.
76
Prof. Davide Dardari, Ing. Matteo Lucchi
Energy efficiency (2/2)
39
77
Prof. Davide Dardari, Ing. Matteo Lucchi
For further information…………