1 / 19

Outline

Outline. Wireless introduction Wireless cellular (GSM, CDMA, UMTS) Wireless LANs, MAC layer Wireless Ad hoc networks routing: proactive routing, on-demand routing, scalable routing, geo-routing wireless Ad hoc multicast TCP in ad hoc networks QoS, adaptive voice/video apps

sage
Download Presentation

Outline

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Outline Wireless introduction Wireless cellular (GSM, CDMA, UMTS) Wireless LANs, MAC layer Wireless Ad hoc networks routing: proactive routing, on-demand routing, scalable routing, geo-routing wireless Ad hoc multicast TCP in ad hoc networks QoS, adaptive voice/video apps Sensor networks

  2. QoS Support • Objective: • Guarantee Quality of Service to real time users • QoS users: • Multimedia • Voice over IP • Time critical applications • QoS metrics: • Bandwidth / data rate • Delay • Delay variance / jitter • Packet loss • Bit Error rate • etc

  3. QoS Wireless Scenarios • Scenarios • Multihop ad hoc network • Wireless Internet access (from W-LAN or ad hoc net) • Wireless mesh networks • The above scenarios are all challenging: • Channel parameter fluctuations (motion, obstacles, interference) • Change in S/N ratio at receiver • Mobility (path breaks, must be repatched, etc) • Multiaccess channel (distributed queue scheduling) • Interference from nodes beyond one hop • QoS support truly a cross-layerdesign: • From physical layer to applications

  4. Physical Layer QoS • As distance between tx and rcv increases, and noise/interference increases, S/N ratio decreases. • This requires a reduction in data rate to maintain acceptable BER • In 802.11 a, the rate can be adjusted in the following steps: 6,9,12,18, 24,48,54 Mbps • In 802.11b: 1,2,5.5, 11 Mbps • ARF (Auto Rate Feedback): • Rate reduced by transmitter if ACK missed twice • Rate restored if 10 consecutive ACKs are correctly received • RBAR (Receiver Based Auto Rate): • Receiver measures S/N ratio from RTS; returns allowed rate in CTS

  5. MAC Layer QoS • IEEE 802.11e is addressing QoS: • DCF -> Enhanced DCF (EDCF); 8 priority levels defined by different AIFS values • PCF -> Hybrid CF (HCF): can “poll” nodes also in contention free period • 802.11e is mainly directed to infrastucture LANs • In ad hoc, multihop network one can use some of the EDCF features • Another problem in ad hoc nets is the simultaneous scheduling of multiple transmissions (spatial reuse) • Several solutions: • TDMA (eg, cluster TDMA) • Non interfering flow graph (SCOPE)

  6. Network layer QoS • QoS routing (either proactive or on demand) • QoS extension of LANMAR, Fisheye • Real time bandwidth measurement in mobile ad hoc networks • Bandwidth allocation (or re-allocation) • Bandwidth reservation • Call Acceptance Control • Mobility adaptation • Scheduling and policing • Coexistence with best effort traffic

  7. Application layer QoS • Rate adaptation of multimedia stream • Network to application layer feedback (with avail bandwidth information) • End to end bandwidth estimation • We will discuss a video adaptation example

  8. Adaptive , renegotiable QoS support • If nodes move, the end to end path capacity can change dynamically: • The connectivity and topology change • The channel propagation characteristics change (eg, multipath fading) • The external interference can change • A possible answer: • Renegotiable QoS • rate adaptation • preemption of lower priority users • Case study: adaptive video streaming

  9. server Scalable, Adaptive Video Streaming FLIR client

  10. Why Adaptive Video? • To prevent traffic congestion • Adjust the stream rate so that it “fairly” shares the available bandwidth • To deal with channel random interference, propagation, jamming • Adjust (reduce) video packet size, strengthen the channel encoding to combat random/burst errors • Challenge: • Distinguish between congestion (must reduce rate) and random errors (keep same rate but strengthen the code)

  11. MPEG H263 server client client server The concept channel fading • End to End Feedback Adaptation Approach • Traditional approach • Transport and Application employ end-to-end measurements to perform flow control • Network Feedback Adaptation Approach • Network layer propagates channel measurements to source • Ability to detect cause of degradation and react properly • Implemented in simulation, testbed and hybrid simulatn mobility internal interference LOSS FROM CONGESTION LOSS FROM ERROR external interference (jamming, environment noise)

  12. Adaptation Techniques • RTP Loss rate Adaptation (Trial And Error) • It constantly ‘tries’ to support higher rates • Backs up when loss is detected • Available Bandwidth (AB Probe) eg. TCPW • Av. Bdw. estimated from inter-packet intervals • Can distinguish error loss from congestion • Network Feedback • Link channel quality and bdw info piggybacked on routing pkts • Gives accurate picture of network path state to source

  13. Network feedback • Goal: dynamic estimation of available bandwidth AB from source to destination • Approach: MAC and network layer cooperation • At the MAC layer: • estimate the permissible throughput from a node to each of its neighbors • estimate the “available bandwidth” AvBd for a node • At the network layer: • Finally, propagate the AvBd estimate to source using routing control packets

  14. MAC layer computation (a) Estimate link throughput: Throughput = S / (Tack - Ttx) where: S = packet size (bits) Ttx : time when packet tx is first attempted Tack : time when ACK is received This estimate accounts for collision avoidance, retransmissions, backoffs etc The estimate is averaged over a measurement window (b) Measure node utilization u = fraction of time node is busy (ie, it has packets in the queue) Available Bdw = (1 – u) x Throughput

  15. Hidden Terminals Hidden Terminals Hidden Terminals Measurement Example 10Mbps 11Mbps 802.11 links B Node utilization=80% A 7Mbps 6Mbps C D

  16. ..... client Network Network client Link Link Network Network Network server Link Link Link Application API Network Link Link Assisted/Network Feedback channel fading MPEG H263 internal interference mobility Multi-hop noise server LOSS FROM CONGESTION LOSS FROM ERROR external interference (jamming, environment noise) Propagate() Measure() Measure() Measure() Measure() Network QoS Feedback Accurate Available Bandwidth Measurement and Advertising

  17. How to propagate to sources? QM-AODV RREQ 1 • Q-AODV – extension of AODV for QoS routing • Finds routes given: minimum bandwidth and maximum delay constraint • Does not deal with measurement aggregation • Does not have to propagate measurements • AvBd propagation - QM-AODV • Use Q-AODV packet format • Add bottleneck hop information • Use an MREP special reply packet • Use Measurement Relay Factor to trigger Measurement updates Update Reverse Route with AB(1) 2 MREP(AB(4)) If Bhop=2 or minimum Update 3 RREP (min(inf,AB(5),0 or 1) RREP (min(inf,AB(5),0 or 1) Significant AB Change 4 RREP (inf,0) 5

  18. 802.11 Multi-hop – Loss Rates • AB-probe performs best initially, then worse than RTP when load is high • RTP adaptation works better than direct PP (Packet Pair) method • Network Feedback is best

  19. Real-Time Protocol (RTP) • Provides standard packet format for real-time applications - a sublayer of the transport layer • Typically runs over UDP • Specifies header fields (see below) • Payload Type: 7 bits, providing 128 possible different types of encoding; eg PCM, MPEG2 video, etc. • Sequence Number: 16 bits; used to detect packet loss • Timestamp: 32 bytes; gives the sampling instant of the first audio/video byte in the packet; used to remove jitter introduced by the network • Synchronization Source identifier (SSRC): 32 bits; an ID for the stream; assigned randomly by the source

More Related