• Changing modulation mode depending on the conditions in the radio channel:
• Interferences from other radio transmitters, change of received signal strength, signal-to-noise ratio etc.
• Transmitter uses information measured in the channel
• Example:
• Channel is noisy: QPSK
• Less noisy channel: 16QAM
2.43. "Analog" modulation
• Modulating a sinusoidal carrier by an analog modulating signal
• Amplitude and angle modulation (frequency or phase modulation, depending on which parameter of the carrier is proportional to the modulating signal’s magnitude)
• Will not be dealt with in detail
• Amplitude modulation is being used for multiplexing several signals onto a single transmission channel, because it is
• Produced by multiplication in the time domain...
• which results in spectrum shift in the frequency domain
2.44. 4 SHARING A COMMON RADIO CHANNEL: MULTIPLEXING 2.45. Multiplexing: FDM and TDM
2.46. TDM - time division multiplexing
2.47. Wavelength division multiplexing
• Mentioning for completeness; it is a technique used in fibre optic communications channels
• It is essentially an FDM technique but called WDM in optical communications
2.48. CDM - code division multiplexing
• Neither frequency nor time separation
• How can it be?
• How can then the sources be distinguished?
• And what is it good for?
• Let us answer these questions by the example of direct sequence CDM-et!
2.49. Principle of code division multiplexing
• Analogy: humans communicating in a room
• TDM: talk one after the other, not the same time
• CDM: talk the same time but speak different languagues
• Principle: assigning individual specific codes to the sources (chip code)
• Chip codes are transmitted, resulting in spectrum spreading, signal uses the whole frequency band.
• Detection based on orthogonality of the chip codes
2.50. Principle of CDM, example
2.51. CDMA: advantages and disadvantages
• +:
• Eavesdropping is difficult, looks like a noise
• Not sensitive to narrow-band interference and multipath fading
• Frequency reuse in cellular systems
• -:
• Larger bandwidth proportionally to the chip code length
• Exact synchronization is needed at chip boundaries
2.52. 5 SHARING A COMMON RADIO CHANNEL: MULTIPLE ACCESS
2.53. Multiplexing multiple access
• Multiplexing:
• Common channel
• Many sources
• All at the same place, at the input of the channel
• Multiple access:
• Common channel
• Many sources
• Sources are spread over a certain territory which can be large
• In network protocol architectures the name is MAC (Medium Access Control)
2.54. Ways to divide a common channel ("fixed" methods)
• Multiple access companions of multiplexing techniques:
• FDMA TDMA CDMA
• FDMA
• Orthogonal, but difficult to implement
• Efficiency lower due to guard bands between adjacent channels inorder to ensure separability
• Applications for real-time traffic
• TDMA
• Orthogonal, good for flexible division, easy to create sub-channels of different size
• CDMA
• ...
• Interesting comment: what is SDMA (Space ...)?
• Separation of users in space
• Using directed antenna beams (smart antennas: e.g. MIMO - multiple input multiple output)
2.55. Example of a common radio channel: using a satellite repeater
2.56. Random access (random = free)
• Simple (pure) Aloha: entirely free access
2.57. The simple Aloha protocol
• Rules:
• Each terminal transmits when it has message to transmit
• It doesn't check whether the channel is idle
• Collisions are unavoidable, if a collision occurs no participant's transmission is successful
• They try again
• After a random delay!
• Simple and works: no coordination is needed, terminals don't have to know how many they are etc.
• Detecting a collision: automatic acknowledgements (sending on uplink-receiving - hearing back their own transmission on downlink)
2.58. "Slotted" Aloha
• Similar but time axis is "slotted", terminals can only start transmission at the beginning of a slot
• Consequently, either full overlap of packets or no overlap
• Probability of collisions smaller, utilization higher Performance of Aloha protocols
• Utilization: relatively low ( ) because channel cannot be heavily loaded as in this case it tends to become unstable (saturated)
• Delay: not bounded, but acceptable when traffic is not high
• Fairness: yes
2.59. Reservation Aloha
• By reservation, the channel throughput and delay can be improved
• Users send reservation messages using a (small) part of the channel bandwidth
• All users listen to the channel and hear about the reservation requests of others
• This way a reserved channel can be granted to users who successfully bidded for the channel
2.60. Applications of the Aloha protocol
• In a wireless environment, in cases where channel state cannot be sensed before transmitting
• No "carrier sensing", see later
• For bursty traffic, it is advantageous
• fixed allocation schemes are not efficient, cause delays (see later)
• but the traffic should be below 10%
• If the traffic grows beyond that, throughput falls quickly
• Application areas (mainly slotted Aloha):
• Satellite systems (tactical)
• RFID
• Together with reservation: GSM-GPRS
2.61. Carrier Sensing Multiple Access
• Using additional information from the channel: if it is idle or busy (commonly called carrier sensing)
• If the channel is idle:
• transmit
• If the channel is busy:
• Try later (non-persistent)
• Wait for the channel becoming idle (persistent):
• When idle, transmit immediately (1-persistent)- If collision: wait until a randomly chosen delay expires
• When idle, transmit with probability p (p-persistent)
2.62. Performance of carrier sensing multiple access
• A bit more complicated but still relatively simple implementation (only channel status - idle/busy - needs to be sensed)
• Significantly higher throughput can be achieved, in theory, it can be close to 100%
• Sensitive to the propagation delay, signal has to reach a station so that it can sense the channel busy.
Otherwise the information about the channel status can be outdated
• Can be unstable
• Fairness is granted
2.63. Centralized multiple access methods
• Polling: right to transmit is granted by a controller which polls the stations
• Probing: controller polls a group of stations
• If collision occurs (more than one station in the group want to transmit) it needs to be resolved
• Reservation: controller grants transmission right to a station for a specified time.
• Requests for the channel can be collected:
• Using a separate channel or part of a common channel for requests
• By competition
2.64. Roll-call polling
• Controller polls all stations one after the other
• If a station has message to transmit it sends
• Others are not allowed to transmit thus communication is contention-less
• Part of the time (of the available channnel throughput) is used for organizing the communication
• Throughput is the ratio of the to parts
• Access delay: bounded
• Stable and fair method
2.65. Probing
• Group polling:
• When collision, group is divided into subgroups
• Better channel utilization
• Lower acess delay
• Stability and
• Fairness granted
2.66. Reservation
• For large round-trip time, polling is not efficient
• Better to divide the channel into:
• Reservation part
• Transmission part
• The reservation part (constituting a small fraction of the total channel capacity):
• Dedicated to requests, based on
• Fixed allocation (e.g. TDMA)
• Contention (e.g. Aloha)
• Controller receives successful bids and grants access
2.67. Reservation: dividing the channel
2.68. 6 NETWORKING ASPECTS IN WIRELESS COMMUNICATIONS
• Using a satellite repeater
• Self organizing topologies
• The cellular principle
2.70. Centralized topologies
• Communication between the nodes goes through a central station (hub or base station)
• The central station controls the nodes and the transmission from/to each node
• The central station manages the access by nodes to the network resources, allocates bandwidth
• Examples:
• Cellular networks
• WLAN, WMAN (in most cases)
• VSAT networks
2.71. Centralized topologies: advantages and disadvantages
• Advantages
• Nodes use less power to reach their peers compared with fully connected peer-to-peer since the communication goes through the hub
• Complex hub but simple nodes
• Hub can be placed on an optimal location
• Hub can provide connection to the backbone
• Disadvantages
• Single point of failure
• Delay in node-to-node communication due to two hops
• Cannot cover large areas
2.72. Decentralized topologies
• Also called peer-to-peer topology
• Fully connected network
• Any pair of nodes "see" each other (in radio comm. sense)
• Nodes communicate with each other directly
• Multi-hop networks (also called mesh topologies)
• Nodes cannot reach each other
• Intermediate nodes forward messages to the destinations
2.73. Fully connected network: advantages and disadvantages
• Advantages
• There is no single point of failure
• No store-and-forward delay
• No routing is needed, reduced complexity in nodes
• Disadvantages
• In large networks performance degradation
• Nodes need to increase transmit power to reach remote destinations
• Interferences
2.74. Multi-hop peer-to-peer network: advantages and disadvantages
• Advantages
• Nodes can operate at low transmitting power as they need to reach only the neighbors
• No infrastructure is needed, therefore
• Wide use in military, disaster recovery and similar areas
• Fault-tolerant, nodes can leave the network (because of failure or other reasons) of join, a kind of organic property
• Routing algorithms restore communication paths
• Total throughput of the network is high
• Use in ad hoc and sensor networks
• Disadvantages
• Complex routing algorithms are needed
• Performance degradation when traffic travels over multiple hops (nodes are mostly busy with forwarding)
• Therefore, in large networks a kind of infrastructure (backbone) is needed
2.75. Broadcast networks
• A special case of centralized topology
• Central station sends messages to all nodes
• Classic computer networks example: ALOHANET developed at the Univ. of Hawaii by Prof. Abramson
2.76. Literature
• Andreas F. Molisch: Wireless Communications, Wiley, 2011
• Theodore S. Rappaport: Wireless Communications. Theory and Practice. Edition. Prentice Hall, 2002
• ECE 245/445: Wireless Communications and Networking, Lecture Notes, Rochester University.
3. 3 WIRELESS LOCAL AND METROPOLITAN AREA NETWORKS
3.1. Contents
• Introduction
• Wireless LANs - Wi-Fi - IEEE 802.11
• 802.11 MAC layer
• Wi-Fi Mesh
• WiMAX - IEEE 802.16
• Application areas
• MAC and QoS
• Wi-Fi - WiMAX comparison
3.2. BWA- BROADBAND WIRELESS ACCESS: WLAN, WMAN
3.3. Family of wireless network technologies according to coverage area
3.4. WLANs and WMANs: coverage and data rates
3.5. Our objectives
• BWA is a big family of technologies characterized by:
• A number of different physical layer (radio) technologies
• Many medium access (MAC) protocols
• Numerous standards developed by several stadardization organizations
• Organizations supporting the introduction and penetration of BWA technologies.
• Objectives of our survey lecture?
• Overview of most important LAN-MAN standardized technologies,
• Concentrate on architectural aspects, and
• on the main characteristics of these technologies, and
• on the most important application areas to see how and to what extend the individual technologies meet the requirements of the given application
3.6. WIRELESS LANS - IEEE 802.11
WLAN - Wireless Local Area Network Wi-Fi - Wireless Fidelity
3.7. Wireless LANs?
• Main characteristics:
• Coverage of a few hundred meters
• Data rates from 1-2 Mbps to 54 Mbps, high speed WLANs: hundreds of Mbps
• Operation in the ISM (industrial-scientific-medical) unlicensed frequency band
• Typical applications, originally:
• Segments of LANs within building, in particular: hospitals, shopping centers, hotels, university campuses, protected buildings
• Connections between nearby buildings, e.g. across a street
• Temporary network installations for internet access in exhibition and conference areas
• Today:
• SOHO - small office - home office
• Public internet access points (hotspots)
3.8. WLAN devices
AccessPoint (AP)
3.9. 802.11 standards
• A family of sub-standards
• 802.11 - 1-2 Mbps, 2.4 GHz, FHSS
• 802.11a - 54 Mbps, 5 GHz, OFDM
• 802.11b - 11 Mbps, 2.4 GHz, DSSS, 11-13 channels
• 802.11g - 54 Mbps, 2.4 GHz, OFDM / DSSS, 13 channels
• 802.11n - up to 600 Mbps, 2.4 GHz, OFDM MIMO
• Additional sub-standards specifying important additional functionalities
• 802.11e - QoS support
• 802.11h - automatic power control (ATPC)and dynamic frequency selection (DFS)
• 802.11i - data security, encryption (e.g.: AES encryption)
• 802.11j - 802.11a - HiperLAN2 co-existence
• 802.11s - mesh mode operation
3.10. Channels (2.4 GHz band)
• E.g.: in the 2.4000-2.4835 GHz band, thirteen 22 MHz channels, separated by 5 MHz
• Spectral mask for every channel (because of overlappig)
• 30 dB attenuation at +/-11 MHz from the channel central frequency
• As an example, channels 1, 6 and 11 can be used without overlapping
• Interferences with home applicances and other communication devices: microwave oven, cordless phone, Bluetooth
• In the 5 GHz band: 23 non-overlapping channels
3.11. Physical layer technologies for wireless communications
3.12. 802.11 sub-standards corresponding to different physical
layer technologies
FHSS: frequency hopping spread spectrum DSSS: direct sequence spread spectrum
OFDM: orthogonal frequency division multiplexing
3.13. DSSS: detection, noise protection
• A bit is mapped to a -bit "chip", pseudo-random bit sequence of length
• Detection: correlation receiver, uses the pseudo-random bit sequence of the transmitter, synchronized
• Correlation detector yields maximum within the given chip time and low values at other places
• No correlation with noises and interferences
• Low correlation with interfering signals caused by multipath propagation
• Selection of the chip frequency ("processing gain"):
• Long code: good interference supression
• Short code: consumes less bandwidth
• IEEE 802.11 WLAN: 11 bit spreading
• FCC prescription: min. 10 in the ISM band
• This value is not too high: moderate interference supression, but acceptable bandwidth occupation
3.14. Frequency Hopping Spread Spectrum (FHSS)
• Uses multiple frequencies
• Example:
• 10 MHz sub-bands in the 50 MHz band
• First bit (or group of bits) is transmitted on 2.44 GHz, second bit on 2.41 GHz, third bit on 2.30 GHz
3.15. FHSS: main parameters
• Slow and fast frequency hopping, depending on whether
• the bit rate greater or smaller than the hopping rate
• In the 2.4 GHz ISM band:
• Using min. 75 frequencies
• Max. 400 ms on one frequency, if exactly 400 then 2.5 hops per second;
• FHSS DSSS
• Different bandwidths and data rates
• FHSS provides more protection against narrow-band interferences, because signal is spread over the whole ISM band, while DS occupies only a part of it
3.16. OFDM
• Ortogonal Frequency Division Multiplexing
• Band is divided into many sub-bands
• Parallelized signal streams are transmitted in these sub-bands
• Within sub-bands, QAM (16-ary or 64-ary) or PSK (BPSK or QPSK) is used
• Efficient use of the spectrum: "orthogonal" carriers
• Sub-bands are not disjoint, the spectra are overlapping
• But can be separated, if the Nyquist principle, known from the time domain, is satisfied in the frequency domain
3.17. Channel allocation
3.18. OFDM: advantages and disadvantages
• Advantages:
• Good protection against ISI (intersymbol interference), narrow-band interference and multipath fading
• High spectral efficiency
• Disadvantages:
• Sensitivity to the Doppler effect and frequency synchronization
3.19. WLAN: main operating modes
3.20. WLAN topology (BSS and ESS)
Basic Service Set (BSS) - one cell Extended Service Set (ESS) - several cells
Distribution System (DS) - backbone, e.g. a cable-based Ethernet
3.21. Wireless Access Point (WAP) - Bridge
3.22. 802.11 frame format
3.23. Frame control field
• Protocol version (2 bits)
• Type (2 bits): distinguishes data, control or management frames
• ToDS-FromDS (1-1 bit): from/to distribution network
• Communication within BSS: 00
• Additional bits:
3.25. 802.11 MAC layer - access methods
• Carrier Sense Multiple Access/Collision Avoidance - CSMA/CA
• Not CSMA/CD (802.3), CS, but not CD. Why?
• In wireless LANs it is impossible to detect collisions in a reliable way
• Two access methods:
• Distributed Coordination Function (DCF)
• Point Coordination Function (PCF)
3.26. Distributed Coordination Function (DCF)
• CSMA/CA
• CSMA with Collision Avoidance
• No collision detection
• Stations are not capable of sensing collisions
• Are collisions really "avoided"?
• No, collisions are still possible, but there are measures to decrease the probability of collisions
• What are these measures?
• Different timings called IFS - InterFrame Space
• Short IFS - for control messages
• PCF IFS (PIFS)
• DCF IFS (DIFS) - for data frames
3.27. SIFS-PIFS-DIFS
• SIFS:
• High priority messages can access the channels first
• Example: ACK, RTS, CTS
• Fixed value (10-16 s)
• Reduced IFS (RIFC): in the 802.11n standard (high data rate WLANs)
• PIFS: for APs capable of operating PCF mode (see later)
• PIFS = SIFS + slot time
• DIFS: for data frames
• DIFS = SIFS + ( slot time)
3.28. DCF algorithm
• If the channel is sensed idle, station waits for (D)IFS. If it is still free, transmits.
• If the channel is sensed busy (it is busy at the beginning or it becomes busy during IFS), station continues listening.
• When the channel becomes idle, station still waits for (D)IFS, then continues waiting for an additional randomly chosen time. When that expires, starts trasmitting.
• If the transmission is successful, recevier sends ACK.
3.29. CSMA/CA (DCF)
DIFS: DCF IFS PIFS: PCF IFS
SIFS: Short IFS
3.30. Exponential backoff algorithm
• Backoff_Time=INT(CW RND()) Slot_Time
• CW - Contention Window
• Initially 31, then 63, 127 and so on. Delay is measured in (Slot_Time)
• Slot time
• Chosen in such a way that, during this time, the station can reliably sense if the channel is busy- 20 s (DSSS), 50 s (FHSS)
• RND()
• Function generating random numbers between 0 and 1
3.31. Problem of the "hidden" station
A is communicating with B. C is not aware of it, therefore it too starts transmitting messages to B, causing collision at B.
3.32. Solution: RTS/CTS handshaking
RTS - Request To Send (asking for permission to send) CTS - Clear To Send
(allowed to transmit) ACK - ACKnowledgement
• Optional procedure
3.33. Network Allocation Vector (NAV)
• Every RTS frame contains the duration, for which the station wants to occupy the channel
• NAV: counter at the other stations, they must wait for NAV before sensing if the channel is idle
• When a station (WS1) sends an RTS-t (or CTS), the other stations (WS2 and WS3) starts the NAV counter
3.34. RTS/CTS + NAV: solution to the hidden station problem
3.35. Busy channel
• Two possibilities:
• Is "physically" busy
• Station senses the channel busy
• Is "virtually" busy
• Station receives RTS or CTS which means that the channel will be busy for a duration equal to NAV
3.36. Evaluation of RTS/CTS
• Optional procedure in WLANs
• Good to use when:
• There is a possibility of a hidden station problem
• Competition for accessing the channel is too high:
• Many stations transmitting large amount of traffic
• Collisions of short (RTS/CTS) messages not of the long data frames
• Disadvantages:
• Lower data rate(overhead due to RTS/CTS handshaking)
• Delay can be significant
3.37. Point Coordination Function (PCF)
1. Optional, if exists, it operates "above" DCF, the two are implemented together 2. A single AP controls the access
3. When this AP sends a beacon message, all stations stop operating the DFC mode
4. AP polls stations Consequence: maximum delay is guaranteed 5. Station is allowed to transmit only when it is polled
6. Priorities can be assigned to the stations thus time-critical applications can be supported
3.38. PCF operation
3.39. DCF PCF
• Disadvantages of DCF:
• Too many collisions when traffic is high
• No QoS provisioning
• If a station seizes the channel, it can keep it for a long time
• Advantage of PCF:
• Provides QoS
• Rarely used, usually not implemented in off-the-shelf devices
3.40. Quality of Service (QoS) assurance in WLANs: the 802.11e standard
• HCF - Hybrid Coordination Function
• Traffic classes (Traffic Category): e.g. low priority for e-mail, high priority for Voice over WLAN
• EDCA - Enhanced Distributed Channel Access
• Contention mode: high priority traffic waits for a shorter time
• Contention free period: for voice and video
• HCCA - HCF Controlled Channel Access
• Similar to PCF, but an AP can initiate a contention-free period, otherwise EDCA contention
• Not only stations, traffic classes can also be scheduled
• Higher priority class first (per-session service)
• Most complex CF, operates also above DCF. Requires a different scheduler in APs. Rarely implemented.
3.41. WLAN theoretical and measured performance
Test was conducted in laboratory environment at a distacne of less than 10 m.
3.42. Mesh networks based on WLAN- devices: the 802.11s standard
Source: Proxim Wireless, Mesh Technology Primer, Position Paper, 2005
3.43. Mesh networks based on WLAN- devices: the 802.11s standard
Reconfiguration capability is illustrated in the figure
Source: Proxim Wireless, Mesh Technology Primer, Position Paper, 2005
3.44. Too many nodes to reach the backbone: access capacity goes down
3.45. Dual-radio mesh (IEEE802.11b/g/n for access, IEEE802.11a
for backhaul)
Source: BelAir Networks, Capacity of Wireless Mesh Networks, White Paper, 2006
3.46. Multi-radio mesh (IEEE802.11b/g/n for access, IEEE802.11a for backhaul)
Source: BelAir Networks, Capacity of Wireless Mesh Networks, White Paper, 2006
3.47. Very high speed WLANs: the 802.11n standard
• Up to 600 Mbps
• 4 data streams, 64QAM modulation,
• For 40 MHz channel bandwidth
• MIMO technology
• Multiple transmitters and receivers
• 2X2:2, 3X3:3, 4X4:4 (trsmtr/rcvr antennas/no of data streams
• Frame aggregation to make overhead smaller
3.48. Security in WLANs
• Evolution of solutions
• WEP (Wired Equivalent Privacy): encryption used initially, easy to break
• WEP2, WPA (Wi-Fi Protected Access): WEP improvements, slightly better, no new hardware needed
• WPA2: strong encryption on a new hardware,defined by 802.11i sub-standard
• Some pieces of advice
• Turn off unused WLAN devices
• No advertisement of SSID
• Filtering based on MAC address
• Enabling WEP, WEP2, WPA, WPA2
• Preferably using individual key
3.49. WIMAX - IEEE 802.16
A WMAN technology
3.50. WMAN and WiMAX
• WMAN - Wireless Metropolitan Area Network - IEEE 802.16
• WiMAX - Worldwide Interoperability for Microwave Access
3.51. WMAN - WiMAX
• WLAN: IEEE 802.11x
• Wi-Fi - Wi-Fi Alliance (Wi-Fi - "wireless fidelity")
• WMAN: IEEE 802.16x
• WiMAX - WiMAX Forum - "Worldwide Interoperability of Microwave Access"
• Wi-Fi Alliance and WiMAX Forum: collaboration among vendors and other participants of the market:
• To introduce new technologies,
• To certify standard compatibility,
• To ensure interoperability
3.52. WIMAX application areas
3.53. WIMAX application areas: "hot zone"
3.54. Example of a city backbone: Trento, Italy
3.55. LOS - NLOS environment
• LOS (line-of-sight)
• Direct path
• No obstruction of the Fresnel-zone
• NLOS (non-line-of-sight)
• No direct path only reflected waves
• Thus multipath propagation
• WiMAX:
• NLOS reception is ensured by a specific technology, using the reflected waves, if there is no direct path
LOS and NLOS conditions in urban and rural environment
3.56. WiMAX standards, versions
• "Fixed" WiMAX
• 802.16d or IEEE 802.16-2004
• Physical layer: OFDM
• "Mobile" WiMAX
• 802.16e
• +MIMO
3.57. MAC
• Unlike WLAN: scheduling algorithm
• Unlike WLAN: scheduling algorithm