CMPE 257: Wireless and Mobile Networking

Transcription

1 CMPE 257: Wireless and Mobile Networking Katia Obraczka Computer Engineering UCSC Baskin Engineering Lecture 5 CMPE 257 Winter'11 1

2 Announcements Project proposals. Student presentations. 10 students so we need 5 lectures. 2 students per lecture. Topics: Security. Mobility management. Hybrid networks. Energy management. DTNs CMPE 257 Winter'11 2

3 Today Finish MAC. Unicast routing. CMPE 257 Winter'11 3

4 IEEE Provides 2 types of medium access: DCF: distributed coordination function. PCF: point coordination function. DCF is contention-based. PCF is polling-based. Collision free. Implemented atop DCF. PCF DCF CMPE 257 Winter'11 4

5 IEEE DCF Physical carrier sensing: Stations listen to channel before transmitting (CS of CSMA/CA). Virtual carrier sensing: CA OF CSMA/CA. Reserve channel for transmission. Use RTS/CTS handshake. CMPE 257 Winter'11 5

6 IEEE MAC Protocol: CSMA sender (no CA) 1 if sense channel idle for DIFS then transmit entire frame (no CD) 2 if sense channel busy then start random backoff time timer counts down while channel idle transmit when timer expires if no ACK, increase random backoff interval, repeat receiver DIFS sender data ACK receiver SIFS - if frame received OK return ACK after SIFS (ACK needed due to hidden terminal problem) CMPE 257 Winter'11 6

7 IEEE MAC Protocol: CSMA/CA Physical CS + virtual CS. Sense channel for DIFS. RTS/CTS handshake before sending data. RTS is 20 bytes and CTS is 16 bytes. Maximum data frame is 2,346 bytes. Note: This is only for unicast transmissions. Broadcast Transmissions do not use virtual carrier sensing. CMPE 257 Winter'11 7

8 CSMA-CA Examples F D A C B E Scenario: A wants to transmit to C.. A sends RTS.. D defers.. C sends CTS.. B defers. CMPE 257 Winter'11 8

9 IEEE Wireless LAN b GHz unlicensed spectrum up to 11 Mbps direct sequence spread spectrum (DSSS) in physical layer all hosts use same chipping code a 5-6 GHz range up to 54 Mbps g GHz range up to 54 Mbps n: multiple antennae GHz range up to 200 Mbps All use CSMA/CA for multiple access. All have base-station and ad-hoc network modes. CMPE 257 Winter'11 9

10 CSMA Variants 1-persistent (IEEE 802.3): If medium idle, transmit. If medium busy, keep listening; when medium idle, transmit with probability 1. p-persistent: Same as above but with probability p. Non-presistent: If medium idle, transmit. If medium busy, wait a random period before retrying. CMPE 257 Winter'11 10

11 MAC: A Bird s Eye View CMPE 257 Winter'11 11

12 Solutions to Hidden/Exposed Nodes in CSMA Use only virtual CS: RTS/CTS (Request-To-Send/Clear-To-Send) Used by MACA (Multiple Access Control Avoidance) and MACAW (MACA for Wireless LANs). Use both physical- and virtual CS: CSMA/CA, IEEE CMPE 257 Winter'11 12

13 Dynamic Reservation Approaches: Sender- vs. Receiver-initiated Sender-initiated: A node wanting to send data takes the initiative of setting up the reservation. Most existing schemes. Receiver-initiated: A receiving node polls a potential transmitting node for data. A node can send data after being polled. E.g., MACA-By Invitation. CMPE 257 Winter'11 13

14 Single vs. Multiple Channel Protocols Single channel protocols: control and data use the same channel. Multiple channel protocols: separate channels for control & data transmission; data transmission on separate channels. CMPE 257 Winter'11 14

15 Other criteria for classification Power-aware. E.g., PAMAS. Directional or omnidirectional antennas. QoS-aware End-to-end (E2E) delay Packet loss rate (or the probability) Available bandwidth Challenges: lack of centralized control, limited bandwidth, node mobility, power/computational constraints, error-prone nature of wireless media. CMPE 257 Winter'11 15

16 MACAW [Bharghavan, 1994]. Proposed as improvement to MACA [Karn, 1990]. Note that first IEEE standard (IEEE legacy ) released in CMPE 257 Winter'11 16

17 MACA Proposed as alternative to CSMA. Introduced CA. RTS/CTS handshake (2-way). CMPE 257 Winter'11 17

18 MACA If node A wants to transmit to B, it first sends an RTS packet to B, indicating the length of the data transmission to follow. B returns a CTS packet to A with the expected length of the transmission. A starts transmission when it successfully receives CTS. RTS and CTS packets are much shorter than data packets. A neighboring node overhearing an RTS defers its own transmission until the corresponding CTS would have been finished. A node hearing the CTS defers for the expected length of the data transmission. CMPE 257 Winter'11 18

19 MACA (Cont d) Nodes close to sender: If no CTS heard, OK to transmit. Avoid exposed terminal problem: nodes that hear only RTS can transmit simultaneously with RTS sender. Nodes close to receiver: Upon hearing CTS, defer till after data. Avoid hidden terminal. Binary exponential backoff (BEB). Possible unfair channel allocation (starvation). CMPE 257 Winter'11 19

20 MACAW Inspired basic changes to MACA: Additional signaling. Modified backoff algorithm. CMPE 257 Winter'11 20

21 MACAW Backoff Tries to avoid BEB s unfairness. Proposed fix: sharing congestion information among nodes. Backoff counter information propagated in packet header. After successful transmission, neighbors have the same backoff counter. Tries to prevent large variations of the backoff value. Multiplicative increase (1.5), linear decrease (decremented by 1). CMPE 257 Winter'11 21

22 Data Transmission in MACAW Added ACK. Reliability at layer 2. If ACK not received: Retransmit frame. Increment backoff timer. CMPE 257 Winter'11 22

23 Data Transmission in MACAW Added small Data Sending (DS) control frame. Addresses exposed terminal problem. In MACA, exposed node (received RTS but not CTS) is allowed to transmit. Example: S1->R1 and S2->R2 CTS from R2 may collide with transmission S1->R1. S2 backs-off. Fix: make sure S2 knows RTS-CTS exchange between S1 and R1 was successful. S1 sends small control frame, DS with data exchange duration. When S2 receives DS, defers its transmission. R1 S1 S2 R2 CMPE 257 Winter'11 23

24 Data Transmission in MACAW Added Request for Request-to-Send (RRTS). R2 contends on behalf of S2 if it received RTS from S2 when it could not have responded because deferring due to S1->R1 exchange. When S2 receives RRTS from R2, proceeds with RTS, etc. RRTS RRTS S2 R2 R1 S1 CMPE 257 Winter'11 24

25 FAMA Protocols Floor Acquisition Multiple Access. Floor acquisition = gain control of channel. MACA is an example of a FAMA protocol.. Floor acquisition on packet-by-packet basis. No physical CS; only virtual CS. For collision freedom, RTS needs to be at least 2*channel-propagation-delay. CMPE 257 Winter'11 25

26 FAMA Paper (Garcia-Luna et al.) FAMA non-persistent packet sensing, or FAMA-NPS. No carrier sensing, i.e., MACA. Uses ALOHA to transmit RTS. CMPE 257 Winter'11 26

27 FAMA-NTR FAMA non-persistent transmit request. Sender can send packet bursts. Combines non-persistent CS + RTS/ CTS exchange. Enforces waiting periods at sender and receiver. For both data and control frames. Waiting period proportional to maximum propagation time. CMPE 257 Winter'11 27

28 FAMA-NTR (cont d) Before sending: Node senses channel. If channel busy, backs-off for random period and retries later => nonpersistent. If channel free, node sends RTS. Node waits CTS for 1 RTT. If CTS not received, node backs-off. Otherwise, transmits data burst (up to a maximum size). CMPE 257 Winter'11 28

29 FAMA-NTR To allow bursts, receiving station waits after processing each data packet. Waiting period (T) = maximum propagation time. Transmitting node waits for 2T after any control frame. Allows enough time for RTS-CTS exchange. CMPE 257 Winter'11 29

30 Unicast Routing in MANETs CMPE 257 Winter'11 30

31 Why MANET routing is challenging? No fixed infrastructure. Nodes can have unlimited mobility. So? Multiple hops to destination. Unreliable communication medium. All nodes need to participate in routing/ forwarding. Also, security issues. CMPE 257 Winter'11 31

32 Mobility Mobility patterns may vary widely. Stationary nodes (e.g., sensor nodes). Highly mobile nodes (e.g., vehicles). Discrete versus continuous mobility. Structured versus unstructured mobility. Mobility characteristics: Speed. Direction. Pause time. CMPE 257 Winter'11 32

33 MANET Routing Requirements Dissemination of routing information: Multi-hop paths. Loop free all the time, or almost loop-free. Limited signaling overhead. Self configuring, and adaptive to dynamic topology. Efficiency, e.g.,low consumption of communication bandwidth, energy. Scalable with number of nodes. Localized effect of topology or flow change. CMPE 257 Winter'11 33

34 MANET Unicast Routing Many protocols have been proposed. Many have been invented specifically for MANETs. Many are adapted from protocols for wired networks. Can any one protocol work well in all MANET environments? CMPE 257 Winter'11 34

35 DV or LS? CMPE 257 Winter'11 35

36 DV or LS? Distance-Vector Algorithm: Routers exchange their distances to known destinations; a router uses the distance vectors received from its neighbors to compute its own distances. Distributed computation is problem. Link-State Algorithm: Routers exchange information about the state of the links in the network; a router uses this information to compute its distances to destinations. Distributed database problem. CMPE 257 Winter'11 36

37 Proactive or Reactive? CMPE 257 Winter'11 37

38 MANET Unicast Routing Taxonomy Proactive protocols: A.k.a, table-driven. Traditional routing protocols are proactive. Compute and maintain routes independent on traffic demand/patterns. E.g., OLSR. Reactive protocols: Compute and maintain routes on-demand. E.g., DSR, AODV. Hybrid protocols. E.g., ZRP. CMPE 257 Winter'11 38

39 Tradeoffs? Latency of route discovery. Proactive protocols may have lower latency since routes are maintained at all times. Reactive protocols may have higher latency because a route from X to Y will be found only when X attempts to send to Y. CMPE 257 Winter'11 39

40 Tradeoffs? Overhead of route discovery/maintenance. Reactive protocols may have lower overhead because routes are determined only if needed. Proactive protocols may result in higher overhead due to continuous route updating (depends on rate of changes). Which approach achieves better trade-offs depends on the traffic and mobility patterns. CMPE 257 Winter'11 40

41 What about Flooding? CMPE 257 Winter'11 41

42 Flooding for Data Delivery Sender broadcasts data packet P to all its neighbors. Each node receiving P forwards it to its neighbors. Sequence numbers used to avoid forwarding P more than once. Why? P reaches destination if reachable from source. Destination does not forward P. CMPE 257 Winter'11 42

43 Flooding CMPE 257 Winter'11 43

44 Flooding CMPE 257 Winter'11 44

45 Flooding Advantages: Simplicity. Efficient when rate of information transmission lower than topology changes. Robustness. Disadvantages: High overhead. May result in network congetion. CMPE 257 Winter'11 45

46 Flooding for the Control Plane Many protocols perform flooding of control packets. E.g., route discovery and maintenance. Overhead of control packet flooding may be amortized over data packets transmitted. CMPE 257 Winter'11 46

47 Dynamic Source Routing (DSR) [Johnson96] Reactive protocol. When node S wants to send a packet to D, and does not have a route to D, node S initiates a route discovery. S floods Route Request (RREQ). Each node appends own identifier when forwarding RREQ. CMPE 257 Winter'11 47

48 Route Discovery in DSR Y Z S E A B H C I G F K J D M N L Represents a node that has received RREQ for D from S CMPE 257 Winter'11 48

49 Route Discovery in DSR Y Broadcast transmission [S] Z S E A B H C I G F K J D M N L Represents transmission of RREQ [X,Y] Represents list of identifiers appended to RREQ CMPE 257 Winter'11 49

50 Route Discovery in DSR Y S E [S,E] Z A B H C [S,C] I G F K J D M N L Node H receives packet RREQ from two neighbors: potential for collision CMPE 257 Winter'11 50

51 Route Discovery in DSR Y Z S E A B H C I G F [S,C,G] [S,E,F] K J D M N L Node C receives RREQ from G and H, but does not forward it again, because node C has already forwarded RREQ once CMPE 257 Winter'11 51

52 Route Discovery in DSR Y Z A B H S C I E G F K [S,E,F,J] M J D [S,C,G,K] N L Nodes J and K both broadcast RREQ to node D Since nodes J and K are hidden from each other, their transmissions may collide CMPE 257 Winter'11 52

53 Route Discovery in DSR Y Z A B H S C I E G F K J [S,E,F,J,M] M L D N Node D does not forward RREQ, because node D is the intended target of the route discovery CMPE 257 Winter'11 53

54 Route Discovery in DSR Destination D on receiving the first RREQ, sends a Route Reply (RREP). RREP is sent on a route obtained by reversing the route appended to the received RREQ. RREP includes the route from S to D on which RREQ was received by node D. CMPE 257 Winter'11 54

55 Route Reply in DSR Y S E RREP [S,E,F,J,D] Z A B H C I G F K J M D N L Represents RREP control message CMPE 257 Winter'11 55

56 Route Reply in DSR RREP can be sent by reversing the route in RREQ only if links are guaranteed to be bi-directional If unidirectional (asymmetric) links are allowed, then RREP may need a route discovery for S from D. Unless D already knows a route to S. If a route discovery is initiated by D for a route to S, then the RREP is piggybacked on D s RREQ. CMPE 257 Winter'11 56

57 Processing RREP Node S on receiving RREP, caches the route. When node S sends a data packet to D, the entire route is included in the packet header Hence the name source routing. Intermediate nodes use the source route included in a packet to determine to whom a packet should be forwarded. CMPE 257 Winter'11 57

58 Data Delivery in DSR Y DATA [S,E,F,J,D] Z S E A B H C I G F K J D M N L Packet header size grows with route length CMPE 257 Winter'11 58

59 DSR Optimization: Route Caching Each node caches a new route it learns by any means. When node S finds route [S,E,F,J,D] to node D, node S also learns route [S,E,F] to node F. When node K receives Route Request [S,C,G], K learns route [K,G,C,S] to S. When node F forwards Route Reply RREP [S,E,F,J,D], F learns route [F,J,D] to D. When node E forwards Data [S,E,F,J,D] it learns route [E,F,J,D] to node D Nodes may also learn route when it overhears data. CMPE 257 Winter'11 59

60 Use of Route Caching When S learns that a route to D is broken, it uses another route from its local cache, if such a route to D exists in its cache; otherwise, S initiates route discovery. Node X on receiving a RREQ for some node D can send a RREP if X knows a route to D. Use of route cache Can speed up route discovery. Can reduce propagation of route requests. CMPE 257 Winter'11 60

61 Use of Route Caching [S,E,F,J,D] [E,F,J,D] S E [F,J,D],[F,E,S] A B H C [C,S] I G [G,C,S] F K [J,F,E,S] J D M N L Z [P,Q,R]: Represents cached route at a node CMPE 257 Winter'11 61

62 Route Caching: Speed up Route Discovery, Reduce RREQ Flooding A [S,E,F,J,D] B [C,S] [E,F,J,D] E [G,C,S] When node Z sends a route request for node C, node K sends back a route reply [Z,K,G,C] to node Z using a locally cached route H S C I G [F,J,D],[F,E,S] F [K,G,C,S] K RREQ [J,F,E,S] CMPE 257 Winter'11 62 J Z RREP D M N Route caches at K and J limit the flooding of Z s RREQ. L

63 Route Error (RERR) Y RERR [J-D] Z S E A B H C I G F K J D M N L J sends a route error to S along route J-F-E-S when its attempt to forward the data packet S (with route SEFJD) on J-D fails. Nodes hearing RERR update their route cache to remove link J-D CMPE 257 Winter'11 63

64 Route Caching: Beware! Stale caches can adversely affect performance. With time and host mobility, cached routes may become invalid. A sender host may try several stale routes (obtained from local cache, or replied from cache by other nodes), before finding a good route. CMPE 257 Winter'11 64

65 DSR: Advantages Routes maintained only between nodes who need to communicate. Reduces overhead of route maintenance. Route caching can further reduce route discovery overhead. Single route discovery may yield many routes to the destination, due to intermediate nodes replying from local caches. CMPE 257 Winter'11 65

66 DSR: Disadvantages Packet header size grows with route length. Flood of route requests may potentially reach all nodes in the network. Care must be taken to avoid collisions between route requests propagated by neighboring nodes. Insertion of random delays before forwarding RREQ. CMPE 257 Winter'11 66

67 DSR: Disadvantages Increased contention if too many route replies come back due to nodes replying using their local cache. RREP storm problem. Reply storm may be eased by preventing a node from sending RREP if it hears another RREP with a shorter route. CMPE 257 Winter'11 67

68 DSR: Disadvantages An intermediate node may send RREP using a stale cached route, thus polluting other caches. This problem can be eased if some mechanism to purge (potentially) invalid cached routes is incorporated. Static timeouts. Adaptive timeouts based on link stability. CMPE 257 Winter'11 68

69 AODV Route Requests (RREQ) are forwarded similarly to DSR. When a node re-broadcasts a RREQ, it sets up a reverse path pointing towards the source. AODV assumes symmetric (bi-directional) links. When the intended destination receives a RREQ, it replies by sending a RREP. RREPs travel along the reverse path set-up when RREQ is forwarded. CMPE 257 Winter'11 69

70 Route Requests in AODV Y Z S E A B H C I G F K J D M N L Represents a node that has received RREQ for D from S CMPE 257 Winter'11 70

71 Route Requests in AODV Y Broadcast transmission Z S E A B H C I G F K J D M N L Represents transmission of RREQ CMPE 257 Winter'11 71

72 Route Requests in AODV Y Z S E A B H C I G F K J D M N L Represents links on Reverse Path CMPE 257 Winter'11 72

73 AODV Route Discovery: Observations RREQ contains source and destination IP address, current destination seq. number (incremented as a result of loss of prior route), and broadcast id (incremented for every RREQ). Source IP + bcast id uniquely identifies RREQ: nodes do not forward RREQs they have forwarded recently. RREQ processing: node creates reverse route table entry for RREQ source with TTL. If node has unexpired route to destination in its table with sequence number >= RREQ s, it replies to RREQ with Route Reply (RREP) back to source. Otherwise, broadcast RREQ onward. CMPE 257 Winter'11 73

74 Destination Sequence Number When node D receives route request with destination sequence number N, D sets its sequence number to N, unless it is already larger than N. Node s own sequence number is monotonically increasing. Sequence number is incremented after neighborhood topology change. CMPE 257 Winter'11 74

75 Reverse Path Setup in AODV Y Z S E A B H C I G F K J D M N L Node C receives RREQ from G and H, but does not forward it again, because node C has already forwarded RREQ once CMPE 257 Winter'11 75

76 Reverse Path Setup in AODV Y Z S E A B H C I G F K J D M N L CMPE 257 Winter'11 76

77 Reverse Path Setup in AODV Y Z S E A B H C I G F K J D M N L Node D does not forward RREQ, because node D is the intended target of the RREQ CMPE 257 Winter'11 77

78 Route Reply in AODV An intermediate node has current route to destination, responds to RREQ with RREP. RREP contains source and destination IP, current sequence number, number of hops to destination. If destination, then destination seq. #. Else, node s current record of destination s seq. #. Node receiving RREP sets up forward path to destination. If multiple RREPs received, node forwards first one. Later RREPs discarded unless greater seq. # or smaller # of hops. CMPE 257 Winter'11 78

79 Route Reply Example Y Z S E A B H C I G F K J D M N L Represents links on path taken by RREP CMPE 257 Winter'11 79

80 Forward Path Setup in AODV Y Z S E A B H C I G F K J D M N L Forward links are setup when RREP travels along the reverse path Represents a link on the forward path CMPE 257 Winter'11 80

81 Data Delivery in AODV Y DATA Z S E A B H C I G F K J D M N L Routing table entries used to forward data packet. Route is not included in packet header. CMPE 257 Winter'11 81

82 Timeouts A routing table entry maintaining a reverse path is purged after a timeout interval. Timeout should be long enough to allow RREP to come back. Routing table entry maintaining a forward path is purged if not used for active_route_timeout interval. If no data being sent using a particular routing table entry, that entry will be deleted from the routing table (even if the route may actually still be valid). CMPE 257 Winter'11 82

83 Link Failure Reporting Link failures are propagated by means of Route Error messages, which also update destination sequence numbers. RERR lists destinations now unreachable. If upstream node has neighbors as precursors for the affected destinations, it broadcasts RERR. Nodes receiving the RERR update cost to destination to infinity and forward RERR if needed. Upon receiving RERR, source will initiate route discovery if still needs route. CMPE 257 Winter'11 83

84 Route Error When node X is unable to forward packet P (from node S to node D) on link (X,Y), it generates a RERR message. Node X increments the destination sequence number for D cached at node X. The incremented sequence number N is included in the RERR. When node S receives the RERR, it initiates a new route discovery for D using destination sequence number at least as large as N. CMPE 257 Winter'11 84

85 Link Failure Detection Hello messages: neighbor nodes periodically exchange hello messages. Absence of hello message is used as an indication of link failure. Alternatively, failure to receive several MAC-level ACKs may be used as an indication of link failure. CMPE 257 Winter'11 85

86 AODV Packet Header RREQ: RREQ id. Destination IP address. Destination sequence number. Originator IP address. Originator sequence number. RREQ id + originator IP uniquely identifies the RREQ. Originator sequence number. Destination sequence number. CMPE 257 Winter'11 86

87 Destination Sequence Number Avoid using stale routes. Node updates its destination seq. # when: It generates a RREQ. Prevents conflicts previously established reverse routes. It generates a RREP. New-seq-# = max(current seq #, RREQ dest. seq #). CMPE 257 Winter'11 87

88 Sequence Numbers in AODV To prevent formation of loops A B C D E A B C D E Assume that A does not know about failure of link C-D because RERR sent by C is lost Now C performs a route discovery for D. A receives the RREQ (say, via path C-E-A) A will reply since it knows a route to D via B. Results in a loop (for instance, C-E-A-B-C ) CMPE 257 Winter'11 88

89 Optimization: Expanding Ring Search RREQs are initially sent with small Time-to-Live (TTL) field, to limit their propagation. DSR also includes a similar optimization. If no RREP is received, then larger TTL tried. CMPE 257 Winter'11 89

90 Does The Sequence Numbering Work? To some extent: Sequence numbering scheme is not very efficient. Scheme requires that any given node A either never forgets a destination sequence number it learns, or is able to wait long enough so that it cannot possibly attempt to reach a destination D through a path involving a node B that uses A to reach D. CMPE 257 Winter'11 90

91 Summary: AODV Routes not included in packet header. Nodes maintain routing table entries for active routes. At most one route (next hop) maintained at each node. Unused routes expire even if topology does not change. CMPE 257 Winter'11 91

92 Optimized Link State Routing (OLSR) [RFC 3626] Overhead of flooding link state information reduced by having fewer nodes forward the information. Broadcast from X only forwarded by its multipoint relays (MPRs). Overhead is also reduced as the size of the LS updates is reduced: LS updates contain only info on MPRs. CMPE 257 Winter'11 92

93 OLSR OLSR floods information through MPRs. Flooded information contains links connecting nodes to respective MPRs. I.e., node sends info on nodes that selected it as their MPR. Periodic HELLO messages inform nodes which other nodes selected it as their MPR. Routes used by OLSR only include multipoint relays as intermediate nodes. CMPE 257 Winter'11 93

94 MPRs Multipoint relays of node X are its neighbors such that each two-hop neighbor of X is a one-hop neighbor of at least one multipoint relay of X. Each node transmits its neighbor list in periodic beacons, so that all nodes know their 2-hop neighbors. MPRs of X are 1-hop neighbors of X covering X s 2-hop neighbors. CMPE 257 Winter'11 94

95 Optimized Link State Routing (OLSR) C and E are multipoint relays of A. B F J G C A D E H K Node that has broadcast state information from A CMPE 257 Winter'11 95

96 Optimized Link State Routing (OLSR) Nodes C and E forward information received from A. B F J G C A D E H K Node that has broadcast state information from A CMPE 257 Winter'11 96

97 Optimized Link State Routing (OLSR) E and K are multipoint relays for H. K forwards information received from H. E has already forwarded the same information once. B F J G C A D E H K L Node that has broadcast state information from A CMPE 257 Winter'11 97