Design Challenges for Energy- Constrained Ad Hoc Wireless Network - PowerPoint PPT Presentation

About This Presentation
Title:

Design Challenges for Energy- Constrained Ad Hoc Wireless Network

Description:

(1) pure Aloha. Users transmit no channels whenever they have data to send (2) slotted Aloha ... (2)Busy tone transmission. Listening for busy tone no separate ... – PowerPoint PPT presentation

Number of Views:127
Avg rating:3.0/5.0
Slides: 51
Provided by: homeE
Category:

less

Transcript and Presenter's Notes

Title: Design Challenges for Energy- Constrained Ad Hoc Wireless Network


1
Design Challenges for Energy- Constrained Ad
Hoc Wireless Network
  • 2004. 4. 16
  • ???

2
Agenda
  • 1. Introduction
  • 2. Application
  • 3. Cross layer design
  • 4. Link design issue
  • 5. Medium Access Control Design Issues
  • 6. Network Design Issues
  • 7. conclusion

3
Introduction
ltCellular system architecturegt
  • Divided into cell
  • Mobile terminal communicate directly with a base
    station
  • There is no peer to peer communication between
    mobiles
  • All communication is via base station through
    single-hop routing

4
Introduction
  • Each base station is connected by a high-speed
    link (typically wired) to a mobile switching
    center (MSC) that in turn is connected to PSTN.
  • Base station/MSC/PSTN infrastructure and
    centralized control
  • Most WLAN have similar centralized single hop
    architecture.
  • mobile-centralized access point backbone
    Internet, access point performs all networking
    and control functions for mobile nodes.

5
Introduction
ltAd Hoc network structuregt
  • Ad Hoc wireless network has peer to peer
    communication, distributed networking, control
    functions among all nodes, and multi-hop routing
  • Line width SINR (connectivity)

6
Introduction
  • ltsingle hop routinggt
  • Node send pkt directly to final destination.
  • Excessive interference to surrounding nodes.
  • Low rate, high probability of bit error,
    excessive delay
  • ltmulti hop routinggt
  • Routing via forwarding by intermediate nodes
  • Increase capacity of ad hoc wireless networks
    (but achieving these gains through decentralized
    routing remains elusive)

7
Introduction
  • ltEnergy constrainedgt
  • This is not inherent to all ad hoc wireless
    network. (may be attached to large energy source)
  • However many interesting ad hoc network nodes
    will be powered by batteries with limited
    lifetime.
  • Interest are devices that cannot be recharged,
    sensors that are imbedded in walls or dropped
    into remote region.

8
Introduction
  • All levels of protocol stack minimize power
    requirements
  • Sleep modes for nodes conserve standby energy.
  • Hardware and operating system design in the node
    can also conserve energy.

9
Introduction
  • ltmobilitygt
  • Link layer how fast link characteristics change,
    whether or not link connectivity is stable.
  • MAC layer how scheduling algorithms perform.
  • Network layer performance of different routing
    protocols.

10
-- Application -- 1. data network
  • Ad Hoc can support data exchange between laptops,
    palmtops, PDA .
  • WLAN provide 10100Mb/s like wired LAN for low
    mobility and stationary users
  • 802.11b provide data rates on this order. But
    many user, less rate
  • Commercial LAN,Not based on Ad Hoc

11
-- Application -- 1. data networks
  • Wireless MAN
  • multi hop routing
  • Cover large area
  • High data rate
  • Link quality of each hop is changed
  • Distributed collection of Laptop computer
  • canonical example of Ad Hoc wireless data
    network.
  • Highly limited in battery power, laptop acting as
    a router for other laptops

12
-- Application --2.Home Networks
  • Support communication between PC, laptop, PDA,
    cordless phone, consumer electronics,
    entertainment system anywhere in around home.
  • All without getting up from bed.
  • Sense people and movement, adjust light and
    heating.
  • Alert police or fire department.
  • Internet for remote control, software upgrade,
    schedule maintenance.

13
-- Application --2.Home Networks
  • Design challenges
  • 1. support QoS
  • Data rates delay constraint is quite stringent
    for home entertainment systems
  • 2. need standardization
  • All home network device follow same
    standardization
  • 3. power constraints

14
-- Application --2.Home Networks
  • ltApproachgt
  • Use existing WLAN standard such as 802.11.
  • Limitation of 802.11
  • Use star architecture all device talk directly to
    single access point. Eliminate benefit of multi
    hop routing
  • Limited in data rate (802.11b110Mbps,
    802.11a1070Mbps for pkt data not media streaming

15
-- Application --2.Home Networks
  • Natural choice peer to peer ad hoc network
  • Eliminates overhead of centralized node.
  • Need only communicate with their nearest
    neighbors
  • Most home networking applications involve
    stationary or low-mobility nodes, so protocols
    need not support high mobility.
  • But provide high quality

16
-- Application --3. Device Networks
  • Primarily intended to replace in convenient
    cabled connections with wireless connections.
  • Many of these devices have limited battery life,
    rechargeable.
  • Main topology Bluetooth
  • Short range communications.
  • 10m-1mW, 100m- 100mW
  • 2.4GHz (used without any licensing)
  • Energy constraints
  • 8 Bluetooth devices form star topology with one
    node acting as a master and other nodes acting as
    slaves. Master node is responsible for
    synchronization and scheduling transmissions of
    slave nodes

17
-- Application --4. sensor networks
  • Have enormous potential for both consumer and
    military applications.
  • Enemy targets will be small, mobile.
  • Used to identify and track targets.
  • Electromagnetic, optical, chemical, biological
    sensors.
  • Optical sensor provide navigation, routing
    vehicles around obstacles while guiding them into
    position for defense or attack.

18
-- Application --4. sensor networks
  • Examples of sensor networks for home environment
  • electricity, gas, water meters can be read
    remotely through wireless connections.
  • Smoke detector
  • Detect spread of gas leaks or toxic fumes.

19
-- Application --4. sensor networks
  • Use at the sites of large accidents
  • (collapse of a building)
  • Sensor deployed at the site of an accident used
    to track heat, natural gas, toxic substances
  • Used to detect and locate trapped survivors
  • Low cost, low power sensor being inserted into
    concrete before it is poured
  • Must be robust, self-configuring, long lifetime

20
Cross-Layer design
  • Requirements at each layer
  • (1)link layer
  • Multiple antenna, coding, power control
  • (2) MAC layer
  • Power control, scheduling
  • (3) network layer
  • Energy-constrained, delay constrained routing
  • (4) application layer
  • Application adaptation

21
Cross-Layer design
  • Current ad hoc wireless network protocol design
    is based on this approach.
  • Designed and operated independently.

22
Cross-Layer design
  • Inflexibility of this paradigm result in poor
    performance for ad hoc networks especially when
    energy is constraint or application need high
    bandwidth or stringent delay.
  • So, cross layer design support adaptivity and
    optimization across multiple layers

23
Cross-Layer design
  • Example 1
  • If link connectivity (link SINR) is weak, this
    connectivity information is relayed to higher
    level.
  • Network layer rerouting
  • Applicationreduce rate compression

24
Cross-Layer design
  • Example 2
  • Consider again weak link.
  • Link connectivity can be measured accurately an
    quickly at the link layer.
  • Link layer increasing transmit power or error
    correction coding
  • However, if weak link is caused by something
    difficult to correct at the link layer, (mobile
    in tunnel) it is better for higher layer respond.
    Delaying pkt transmission until mobile leaves
    tunnel

25
Cross-Layer design
  • Example3
  • If nodes are highly mobile
  • Link characteristics and network topology will
    change rapidly.
  • Network layer might change routing strategy from
    uni cast to broad cast in direction of intended
    user
  • summary
  • Integrated approach to adaptive networking each
    layer of protocol stack responding to local
    variations and information from other layer

26
--Link Design Issue--
  • Wireless channels are difficult communication
    medium.
  • Low capacity per unit bandwidth, random
    amplitude, phase fluctuation.
  • Goal of link layer design is to achieve rates
    close to fundamental capacity limits using little
    energy.

27
--Link Design Issue--1. Fundamental Capacity
Limits
  • Fundamental capacity of channel maximum
    data rate that can be transmitted over channel
    with small probability of error.
  • (1) Shannon in 1948
  • Capacity of an additive white Gaussian noise
    (AWGN)
  • CBlog2(1SNR) b/s

28
--Link Design Issue--1. Fundamental Capacity
Limits
  • (2) recent work
  • Capacity of single multiuser fading channels,
    capacity of diversity channels, capacity of
    channels with multiple antennas.
  • (3) new area of research
  • Channel capacity under hard transmit energy
    constraint

29
--Link Design Issue--2. coding
  • Channel coding can reduce power
  • Engineers pursued use of very complex codes to
    minimize power.
  • ?convolutional code
  • ?Trellis code
  • ?Turbo code

30
--Link Design Issue--3. Multiple antenna
  • Multiple antennas at the transmitter and receiver
    improve performance and reduce required power.
  • Multiple antenna system use diversity, beam
    steering, multiple input multiple output(MIMO)
    techniques.

31
--Link Design Issue--3. Multiple antenna
  • ? diversity combining
  • Mitigate flat fading,combine multiple
    independently fading copies of signal
  • Reducing impact of flat fading, diversity
    combining can lead to power savings

32
--Link Design Issue--3. Multiple antenna
  • ? Beam steering
  • Creates effective antenna pattern at the receiver
    with high gain in direction of desired signal and
    low gain in all other directions.
  • Combining arrays of antennas with signal
    processing in both space and time
  • Improve energy efficiency since transmitter power
    is focused in the direction of its intended
    receiver

33
--MAC design Issues--
  • Medium access control protocol how
    different users share available spectrum
  • 2 components to this spectrum allocation
  • How to divide spectrum into different channels
  • How to assign these different channels to
    different users.

34
--MAC design Issues--1. channelization
  • 1. Frequency division
  • Bandwidth is divided into nonoverlapping channels
  • Each active user is assigned different channel
  • Frequency division is the simplest channelization
    technique to implement, but it is rather
    inflexible
  • It is difficult to allocate multiple channels on
    demand to a single user, since this requires
    simultaneous demodulation of multiple channels in
    different frequency bands.

35
--MAC design Issues--1. channelization
  • 2. Time division
  • Time is divided into orthogonal time slots.
  • One difficulty
  • Need for synchronization among all nodes
    transmitting to the same receiver.
  • Different nodes have different propagation delays
    to the receiver

36
--MAC design Issues--1. channelization
  • 3. Code division
  • Time and bandwidth are used simultaneously by
    users.
  • Orthogonal or semi-orthogonal code.
  • There is hard limit on how many users can
    simultaneously occupy system. (orthogonal code)
  • Semi-orthogonal code, number of users is
    interference-limited (no hard limit)

37
--MAC design Issues--1. channelization
  • Semi-orthogonal cod require power control to
    compensate for near-far problem
  • Near-far problem
  • If one user is very close to receiver, another
    user is very far away.
  • If both users transmit ant the same power level,
    interference from close user will swamp signal
    from far user.
  • Power control is used that received signal powers
    from all users are same

38
--MAC design Issues--2. Random Access
  • Most data users do not require continuous
    transmission
  • Most systems have many more total users than
    channel.
  • Assign channels to active users.

39
--MAC design Issues--2. Random Access
  • (1) pure Aloha
  • Users transmit no channels whenever they have
    data to send
  • (2) slotted Aloha
  • Only begin transmitting at the start of a time
    slot.

40
--MAC design Issues--2. Random Access
  • (3) CSMA
  • Hidden terminal
  • Node 3, 5 wish to transmit to node 4.
  • Node 5 starts its transmission
  • Node 3 is too far away to detect this
    transmission
  • Exposed terminal
  • Node 2 wish to transmit to node 1, node 3 to node
    4
  • Node 2 sensing and detect node 3s transmit,
    assume ch is busy
  • Node 2 will not transmit even though no
    collision

41
--MAC design Issues--2. Random Access
  • Technique to avoid hidden and exposed terminals
  • (1)Four-way handshaking (in 802.11 WLAN)
  • (2)Busy tone transmission
  • Listening for busy tone no separate control
    channel
  • Not real busy tone, bit is set in field on
    control channel.

42
--MAC design Issues--3. Scheduling
  • If users have long strings of pkt or continuous
    stream data, random access works poorly
    (collision)
  • Channel must be assigned to users in more
    systematic fashion by transmission scheduling
  • scheduling access protocol is unavailable at
    startup. (use Aloha)

43
--MAC design Issues--3. Scheduling
  • Systematic approach to initialization that
    combines benefits of random access for burst data
    with scheduling for continuous data PRMA
  • PRMA assumes slotted system with both burst,
    continuous users.
  • After one user capture channel, she has dedicated
    channel for remainder of her transmission.
  • Subsequent transmission are not corrupted by
    channel.

44
--MAC design Issues--3. Scheduling
  • Data users with short transmission benefit from
    random access protocol assigned to unused slots,
    users with continuous transmission get scheduled
    periodic transmissions after capturing an initial
    time slot

45
--Network Design Issue--Routing
  • Multihop routing protocol in ad hoc network is
    significant design challenge
  • Most work in multihop routing protocols falls
    into 3 main categories flooding, proactive
    routing, reactive routing

46
--Network Design Issue--Routing
  • (1) flooding
  • Broadcast all nodes within receiving range.
  • Advantage
  • Highly robust to changing network topologies and
    requires little routing overhead.
  • Disadvantage
  • Multiple copies of same pkt traverse through
    network, wasting bandwidth and battery power
  • Just small network

47
--Network Design Issue--Routing
  • (2) proactive routing
  • ?centralized route
  • Information about channel conditions and network
    topology are determined by each nod and forwarded
    to a centralized location that computes routing
    tables for all nodes in network.
  • ? distributed route
  • Most common routing used in ad hoc
  • Nodes send their connectivity information to
    neighboring nodes.
  • Routes are computed from this local info.

48
--Network Design Issue--Routing
  • (3) reactive routing
  • Routes are created only at initiation of source
    node that has traffic to send to a given
    destination.
  • Eliminate overhead of maintaining routing tables
    for routes not currently use.
  • Source node initiate a route discovery process
    when it has data to send.
  • Route are maintained until source has no more
    data.

49
--Network Design Issue--Routing
  • ltEnergy constraints in routing protocolgt
  • Exchange of routing information between nodes
    entails energy cost.
  • Routing protocol under energy constraints must
    somehow balance delay constraints, battery
    lifetime, routing efficiency.
  • 102 compare energy consumption of different
    well-known routing protocols.
  • Their result indicate that reactive routing is
    more energy efficient.
  • 102 J.-C. Cano and P. Manzoni, Evaluating the
    Energyconsumption Reduction in a MANET by
    Dynamically
  • Switching-off Network Interfaces, Proc. IEEE
    Symp. Comp. Commun., 2001.

50
conclusion
  • Recent advances in ad hoc wireless network.
  • There is still work improving link, mac, network,
    app protocols, interactions across these layers.
  • Biggest design challenges in ad hoc networks are
    lack of centralized control, limited node
    capability.
Write a Comment
User Comments (0)
About PowerShow.com