Transcript Kein Folientitel

CPSC 689: Discrete Algorithms
for Mobile and Wireless Systems
Spring 2009
Prof. Jennifer Welch
Lecture 14
 Topic:
 Routing in Mobile Wireless Networks
 Sources:




Karp. Multi-hop wireless networks.
Vaidya, Chapter 6.
Schiller, Chapter 8.
Johnson, Maltz. Dynamic source routing in ad hoc wireless
networks.
 Perkins, Royer. Ad hoc on-demand distance-vector routing.
 Chen, Murphy. Disconnected transitive communication.
 MIT 6.885 Fall 2008 slides
Discrete Algs for Mobile Wireless Sys
2
Point-to-Point Routing
 New problem: Deliver messages from a particular source node
S to a particular destination node D, in a mobile ad hoc
network.
 Main issue: S, and other nodes, don’t (initially) know where D
is.
 Similar to the basic IP routing service provided in the Internet:
send message to a particular IP address.
 Easier in Internet: fixed infrastructure, with wires, routers,
backbone networks,…
 But still important in wireless setting.
 Very widely studied, proposals currently being put forth for
practical routing methods.
Discrete Algs for Mobile Wireless Sys
3
Overview of Routing

Internet-style approaches, not taking advantage of the wireless setting:





A simple algorithm that tries to take advantage of mobility.
Link-reversal algorithms



Routing using some substitute for geography.
Geographical routing without location information, Gradient Routing
(GLIDER), Beacon Vector Routing
Location services



[Gafni, Bertsekas 81]
Temporally-Ordered Routing Algorithm (TORA)
Location-free routing



Destination-Sequenced Distance Vector (DSDV)
Dynamic Source Routing (DSR)
Ad-Hoc On-Demand Distance Vector Routing (AODV)
Keeping track of geographical locations of mobile nodes.
[Awerbuch, Peleg 91], GRID, LLS
Location-based routing


Routing using actual geography.
Geocasting, location-aided routing (LAR), compass routing, GPSR,…
Discrete Algs for Mobile Wireless Sys
4
The Setting
 Wireless networks, no infrastructure, no
hierarchical structure.
 Dynamic:
 Stationary and/or mobile nodes.
 Failure/recoveries
 Scale: Small or large, even Internet-scale.
 Typical examples:
 Rooftop networks, of customer-owned radios;
stationary, large-scale.
 Floating sensors: mobile, large-scale.
 Ad-hoc meetings, rescue workers in disaster
areas, soldiers in battle…
Discrete Algs for Mobile Wireless Sys
5
The Routing Problem
 Sender S anywhere in the network wants to send a
message to a particular node D somewhere else in
the network.
 Doesn’t know where D is.
 Subproblem: Set up a route (end-to-end path) to
that node.
 Routes may need to change, in response to
network changes.
 Some routing strategies don't set up routes, just
send messages one step at a time.
 Real goal is delivering messages, not setting up
routes.
Discrete Algs for Mobile Wireless Sys
6
Criteria
 Criteria:
 Accommodate large scale, mobility
 Minimize communication overhead for sending data
messages
 Maximize packet delivery success rate
 Minimize latency of packet delivery
 Minimize route length (e.g., number of hops)
 Minimize amount of state that must be maintained at each
node
 Usual strategies for coping with scale:
 Hierarchy: Good when level boundaries are relatively
fixed, can be chosen by administrative authority.
 Caching (e.g., for routes)
 But these don't work so well for dynamic or unstructured
networks.
Discrete Algs for Mobile Wireless Sys
7
Reuse?
 Could use traditional protocols for wired networks in
wireless networks:
 Establish abstract network with point-to-point links.
 Define links using Neighbor Discovery protocol, based on
periodic Hello messages.
 Run traditional protocols over the abstract network.
 But resulting protocols may not perform very well:
 “Links” may be unreliable, failing and recovering often.
 May cause routes to break frequently, hurting performance.
 Point-to-point link abstraction doesn’t fit wireless networks
well, since they use local broadcast with interference.
 Need to adapt protocols, design new protocols.
Discrete Algs for Mobile Wireless Sys
8
Proactive vs. Reactive Protocols
 Proactive:
 Maintain routes in the background, regardless of current
data-sending activity.
 High overhead.
 Low latency for message forwarding.
 Reactive:
 Establish and maintain routes only when needed for
data-sending.
 Low overhead, if not too much data activity.
 May have high latency for message forwarding.
Discrete Algs for Mobile Wireless Sys
9
Proactive Protocols
 Main methods: Link-State Routing, DistanceVector Routing
 Adaptations of wired protocols, treating every node
as a router.
 Try to establish and maintain routes with fewest
hops (“shortest paths”).
 Use O(n) local state per node = router.
 Require that, periodically, or in response to link
changes, information must be propagated to all
nodes.
Discrete Algs for Mobile Wireless Sys
10
Link-State Routing
 Used in Open Shortest Path First (OSPF) Internet
protocol.
 Can adapt to use in wireless network, using local
broadcast.
 Each node periodically broadcasts information about
the status of its adjacent links, throughout the
network.
 Every node maintains a complete picture of the
network, to use for message routing.
Discrete Algs for Mobile Wireless Sys
11
Link-State Routing, More Details
 Each node periodically constructs a link state packet
describing the status of all its adjacent links, up or down.
 Attaches sequence numbers, to keep track of which packets
are more recent.
 Broadcasts the link state packets throughout the network,
using a network-wide flooding algorithm.
 Each node collects all this information, assembles a complete
picture of the network locally.
 Uses a centralized algorithm, e.g., Dijkstra’s algorithm, to
determine shortest paths.
 Uses these shortest paths to fill in a next-hop routing table,
indicating the next hop for a message for each destination D.
Discrete Algs for Mobile Wireless Sys
12
Link-State Routing, Optimizations
 Limit dissemination of link state packets (LSPs).
 Hazy-sighted LSP dissemination:
 Propagate LSPs more frequently to shorter distances, less frequently to
longer distances.
 Nodes keep more up-to-date about nearby links.
 Uncoordinated selective forwarding:
 Propagate LSPs with some probability p.
 Don’t propagate LSPs if overhear someone else sending it.
 Unreliable, since some neighbors might not have received it.

Coordinated selective forwarding:
 Decide cooperatively who should propagate.
 E.g., each node identifies a Relay Set, a subset of its neighbors such that
each 2-hop neighbor is a neighbor of some node in the Relay Set.
 Propagate LSP to Relay Set nodes only.
 Requires Neighbor Discovery protocol to learn about 2-hop neighbors.
Discrete Algs for Mobile Wireless Sys
13
Distance-Vector Routing
 Bellman-Ford style
 Each node keeps track, for each possible destination D, of the
shortest distance (number of hops) to D that it knows, R[D].cost,
together with the first hop along a shortest path, R[D].nexthop.
 Each node periodically broadcasts to its neighboring nodes its
distance vector, giving its current distances to all nodes.
 R[D].cost for all D.
 Each node adjusts its distance estimates and first hops, based on
information from its neighbors:
 When Z receives distance d for D  Z from neighbor X:
 if R[D].nexthop = X then R[D].cost := d + costZX
 else if d + costZX < R[D].cost then
 R[D].cost := d + costZX
 R[D].nexthop := X
Discrete Algs for Mobile Wireless Sys
14
Distance-Vector Routing
 When Z receives distance d for D  Z from X:
 if R[D].nexthop = X then R[D].cost := d + costZX
 Update the cost estimate with newer information.
 else if d + costZX < R[D].cost then
R[D].cost := d + costZX
R[D].nexthop := X
 If a different neighbor provides a better path, update both the cost
estimate and the nexthop.
 In addition, if Z does not hear from R[D].nexthop for a
while:
R[D].cost := 
R[D].nexthop := 
Discrete Algs for Mobile Wireless Sys
15
Counting to Infinity
 Basic Distance-Vector algorithm allows “counting
to ”, if links fail:
 With all links up, RA[D].cost and RB[D].cost
stabilize to 1 and 2, respectively.
 Then link AD fails, setting RA[D].cost := .
 Then A hears from B, so RA[D].cost := 2 + 3 = 5,
RA[D].nexthop := B (correct).
 Then link BD fails, setting RB[D].cost := .
 Then B hears from A, so RB[D].cost := 5 + 3 = 8,
RB[D].nexthop := A (not correct).
 A hears from B, RA[D].cost := 8 + 3 = 11.
 B hears from A, RB[D].cost := 11 + 3 = 14.
 Etc.
Discrete Algs for Mobile Wireless Sys
D
2
1
1
A
3
3
A
A
2
B
D


5
B
D
2

5
2
3
8
B
16
DSDV
 Destination-Sequenced Distance-Vector [Perkins, Bhagwat 94].
 Add mechanism to DV to prevent counting to .
 Source of problem in example:
 A’s estimate of 5 was based on B’s old estimate of 2.
 B’s estimate is now .
 But B accepts A’s estimate.
 Mechanism to avoid using stale information in determining routes:
Associate a (destination sequence number, distance) tag with each
routing table entry and each distance vector entry.
 (seqno1, dist1) better than (seqno2, dist2) if either:
 seqno1 > seqno2 (newer information), or
 seqno1 = seqno2 and dist1 < dist2 (better distance).
Discrete Algs for Mobile Wireless Sys
17
Some DSDV Details
 Somewhat complicated rules for manipulating tags,
see [Vaidya, Section 6.2.2]. Roughly:
 Node Z increases its own seqno (for Z itself) by 2
each time it sends out a new distance vector.
 When node Z detects that a link to X has failed, Z
increases its sequence number for X by 1.
 When processing a received distance vector from X,
Z accepts new information for D only if its associated
tag is better than the tag stored in RZ[D].

Larger seqno, or same seqno and shorter distance.
Discrete Algs for Mobile Wireless Sys
18
DSDV Guarantees
 Valid route information is always associated with
even seqnos.
 Tags are monotonic: Starting from any node, and
following nexthop entries for a particular D, the tags
in the R tables increase monotonically until we either
reach D or find R[D].nexhop = .
 Implies no routing loops.
 Avoids counting to : In example, B would not
accept the info from A because the tag is not better
than the entry in RB[D].
Discrete Algs for Mobile Wireless Sys
19
Proactive Algorithms, Summary




Link-State Routing, Distance-Vector Routing
Adaptations of wired protocols, treat every node as a router.
Try to establish and maintain routes with fewest hops.
High overhead:




Maintain routes in the background, regardless of current datasending activity.
O(n) local state per node.
Require that, periodically, or in response to link changes,
information must be propagated throughout the network.
Frequent network changes result in instability, global updates.
 Suggests reactive algorithms might work better, in dynamic
mobile ad hoc networks:


Establish and maintain routes only when needed for data-sending.
Low overhead, if not too much data activity.
Discrete Algs for Mobile Wireless Sys
20
Reactive approaches
 Dynamic Source Routing (DSR)
[Johnson, Maltz 96]
 Ad-Hoc On-Demand Distance Vector
Routing (AODV) [Perkins, Royer 99]
Discrete Algs for Mobile Wireless Sys
21
DSR Overview
 Reactive, uses on-demand routing.
 When S has a message for D, it tries to find a good
route to D.
 Performs route discovery, by flooding a query packet to
find a route to D.
 S learns the entire route.
 Attaches the entire route to each data packet it
sends (source routing).
 Caches the route so it doesn't have to search for
new routes for each message (or each packet).
 But, if the network is large and or/changes rapidly, the
routes tend to break frequently, which makes caching less
effective.
Discrete Algs for Mobile Wireless Sys
22
DSR Performance Highlights
 Claim good performance:
 Low communication overhead, high reliability
and low latency of message delivery, and
nearly-optimal routes, under a variety of
conditions.
 During stable periods: Little overhead, can
reuse already-determined routes. Yields lowlatency, low-communication reliable message
delivery.
 When network changes: Adapts reasonably
quickly. Diagnoses broken routes and finds
new ones.
Discrete Algs for Mobile Wireless Sys
23
DSR Assumptions
 Nodes are willing to “participate fully”: forward traffic on
behalf of others, etc.
 Network diameter:
 Examples in paper are fairly small.
 Suggests that this may not scale well.
 Mobility:
 Hosts may move at any time
 Speed much lower than time for packet transmission and
local processing.
 Don't move so fast as to make route computation useless.
 Hosts have a “promiscuous receive mode”, that allows
them to receive every packet they hear, not filtered by
destination address.
Discrete Algs for Mobile Wireless Sys
24
DSR, Basic Operation
 Message forwarding: Source routing

Sender puts entire route in packet header, forwards packet to
first host in the route, successive hosts forward.
 Each host maintains a route cache, caching some source
routes it has learned, each with an expiration time.
 When sender S wants to send a packet to destination D:




S first checks its route cache for a route to D.
If it finds one, it uses it.
If not, uses route discovery protocol to try to find one.
Processes other packets in the meantime.
 Route maintenance:


Source S monitors correct operation of its cached routes; if a
route isn't working, removes it from cache.
If a currently-used route is discovered not to be working, then S
may do route discovery again, resend on the new route.
Discrete Algs for Mobile Wireless Sys
25
DSR Route Discovery
 S broadcasts “route request” (RREQ) packet, containing id pair (S,D).
 Should result in “route reply” (RREP) containing full path.
 Route discovery works by simply flooding the packet, looking for a path
that works.
 RREQ packet contains:
 (S,D)
 request id, unique id for the request, allows pruning out duplicates
 route record, accumulates a record of the sequence of hops the packet
takes during discovery
 Processing a RREQ:
 Duplicate detection: Discard if duplicate, according to request id.
 Loop detection: If your own address already appears in the route record,
this is a loop, discard.
 If you are D, the route has been found; send RREP with the final route
record back to S.
 Otherwise, append your own address and re-broadcast.
Discrete Algs for Mobile Wireless Sys
26
Sending the RREP from D to S



If D has a route to S in its cache, sends the RREP on that.
Else, if communication is bidirectional, D sends the RREP along
the reverse of the route in the RREQ packet.
Else, D initiates a route discovery for (D,S), flooding a new RREQ
message, with the RREP piggybacked on it.



We're not counting on this second route discovery to provide D
with a route from D to S---just using the flooding facilities to send
the RREP back to S.
When S receives this new RREQ, it puts into its route cache the
route from S to D that appears in the piggybacked RREP.
This wouldn’t work without piggybacking: If D just did route
discovery for (D,S) without piggybacking the RREP on the new
RREQ message, then S would not have a path to use to return the
new RREP to D!
Discrete Algs for Mobile Wireless Sys
27
DSR Route Maintenance
 Since DSR is reactive, it does not send routing updates continually
 Instead, while a route is actually being used to send a message, a
route maintenance protocol monitors the operation of the route,
informs sender of any problems.
 Some mechanisms used to diagnose the failure of individual hops
along the path:
 Link-level acks, with non-occurrence reported to higher layers.
 Passive acks, trying to overhear the next transmission by the next node
along the path.
 Some application-level ack information, if available.
 Explicit acks added to the message-forwarding protocol.
 A node discovering such a local failure then sends a Route Error
(RERR) message to the sender S of the current message, indicating
the particular hop that failed.
 When S receives the RERR, it truncates all the paths in its cache at
the failed hop.
Discrete Algs for Mobile Wireless Sys
28
DSR Route Maintenance
 Q: How does the diagnosing node X get the RERR message back
to the sender S?
 A: As for the RREP:
 If X has a route to S in its cache, sends the RERR on that.
 Or, reverse the route from the packet (if communication is bidirectional).
 Or piggyback RERR on a new route discovery.
 Another option: X performs an entire new route discovery for (X,S),
with no piggybacking, before sending the RERR message.
 Q: Why is this not circular?
 A: Because S should have a route to X in its cache.
 End-to-end route maintenance:
 Use explicit ack for data packet, from destination D to source S.
 Gives “coarser” information---source could delete only one route, not
truncate many routes as for single-hop diagnosis.
Discrete Algs for Mobile Wireless Sys
29
DSR Optimizations
 Storing routes in the route cache:
 List of routes, indexed by destination.
 Tree of routes, to different destinations.
 When to add a route to the cache?
 As a result of a successful route discovery (of course).
 Maybe also:
 When forwarding a data packet along a route, add the rest of the route.
 When forwarding an RREP containing a route, add the relevant portion.
 Some overheard routes.
 Use route cache to avoid propagating an RREQ:
 If you already have a route to the intended target D of the RREQ (that would not
introduce a loop), then just append your route to the route that already appears
in the RREQ, to obtain a candidate route from S to D, and send this in a special
RREP message back to S.
 Could also gather route info from neighbors’ caches.
 Set bound on route length, truncate search at that bound.
Discrete Algs for Mobile Wireless Sys
30
DSR Performance






Packet-level simulation.
Varied parameters in the protocol, number of nodes, pattern/speed of
movement, distribution of nodes in space.
Typically: Mobile hosts in a large room, moving at walking speeds,
with 3-meter transmission/reception range.
Claim the results are also valid for vehicles, which move faster, over
a larger region, because parameters scale appropriately (?).
Random initial placements, random pause, random new locations,
random speed.
Results:



For all but the shortest pause times, the total number of packets
transmitted is very close to optimal (for communicating actual data).
Nearly all data packets are successfully delivered, (because route
maintenance detects when routes are no longer working).
The protocol finds and uses close to optimal routes.
Discrete Algs for Mobile Wireless Sys
31
Ad Hoc On-Demand Distance Vector (AODV)
Routing


Another reactive protocol, like DSR.
But not a source routing algorithm:





Nodes maintain distance-vector routing tables, route messages
according to these tables.
Tables populated on-demand: Next hops added only as routes are
needed, time out if not used, or if detected as broken.
Gives usual advantages of reactive strategies:



Does not maintain the entire route anywhere, but distributes its
representation among the nodes on the route.
Like distance-vector algorithms.
Low communication overhead, small storage
Lower local storage requirements than DSR, since only next hop
information is stored, not full routes.
Combination of performance advantages makes AODV more
scalable than DSR and DSDV.
Discrete Algs for Mobile Wireless Sys
32
AODV Assumptions




Basically the same as DSR:
Nodes are willing to “participate fully”.
Mobility is slower than routing.
Nodes in promiscuous receive mode.
 However, unlike for DSR, the network is not
assumed to be small.
 AODV is intended to be scalable.
Discrete Algs for Mobile Wireless Sys
33
AODV, Basic Operation
 S wants to send a message to D, but has
no entry for D in its routing table.
 S initiates route discovery by locally
broadcasting a RREQ, containing:
 (S,D)
 request_id
S
 source_seqno, dest_seqno
 hop_count
 (S, request_id) uniquely identify RREQ.
H
I
G
A
F
E
B
C
Discrete Algs for Mobile Wireless Sys
D
34
AODV, Basic Operation
 When a node receives an RREQ:
 If has already seen it (same S,
request_id), discard.
 If have relevant information,
then reply.
 Else, increment hop_count,
S
rebroadcast, and record:
 (S,D), request_id,
source_seqno
 Who forwarded the RREQ to
you,
 Expiration time
Discrete Algs for Mobile Wireless Sys
H
I
G
A
F
E
B
C
D
35
AODV, Route Discovery
 Relevant information:
 If you are the destination D, or have a
route to D with the same or greater
destination sequence number.
 Reply: Next slide…
 For now, assume no node knows any S
route to D.
 As the RREQ propagates through the
network, the information nodes record
about whom they received it from sets up
temporary reverse pointers.
 Temporary because they get discarded
after expiration time.
Discrete Algs for Mobile Wireless Sys
H
I
G
A
F
E
B
C
D
36
Replying to RREQs

Node X with relevant information replies to an RREQ by generating
a RREP, containing:








(S,D), dest_seqno
hop_count, number of hops in route from X to D.
lifetime
This gets passed back along the reverse pointers, unicast fashion.
RREP's handled separately from RREQs, no request_id.
For a given (S,D), X passes RREP along the reverse pointer iff X
hasn't already passed along such an RREP for a larger dest_seqno,
or for the same dest_seqno with the same or smaller hop_count.
Requires keeping track of RREPs handled for each (S,D).
Assumes X has at most one reverse link for each (S,D):




A reverse link should mean X is on an active path from S to D.
Two reverse links would mean two active paths.
If S sends out a new RREQ for D, it means the previous route was
broken.
So X should discard the old reverse link in favor of the new one.
Discrete Algs for Mobile Wireless Sys
37
Forward Path Setup
 As the RREP is passed back, each
node that handles it sets up the
appropriate forward pointer in its
routing table entry for D.
 Each routing table entry holds, for
each D:




Number of hops to D
Next hop
dest_seqno
Active neighbors for this route
(neighbors who have sent a
message recently that used it)
 Expiration time for the route table
entry (if not used for a while)
H
I
G
S
Discrete Algs for Mobile Wireless Sys
A
F
E
B
C
D
38
Local Connectivity Management
 Suppose D moves.
 Suppose A wants to send a
message to D, broadcasts
RREQ.
 A gets RREP's from S, B, and
D; chooses D’s because it has
the largest dest_seqno.
 B overhears D’s RREP,
changes its route table entry
because this RREP has a
larger dest_seqno.
Discrete Algs for Mobile Wireless Sys
H
I
G
S
A
F
D
E
B
C
39
Local Connectivity Management
 Q: How does C determine its path to D is
broken?
 A: D is supposed to send Hello messages to
C periodically.
 In general:
 Nodes on active paths send Hello messages
S
to their prececessors.
 If a predecessor doesn’t receive one for a
while, it assumes the successor is gone,
propagates back a special RREP with
hop_count =  and one greater dest_seqno.
 The effect: Anyone who has not found a
more recent route to the destination
removes the route from its table.
Discrete Algs for Mobile Wireless Sys
H
I
G
A
F
D
E
B
C
40
Local Connectivity Management
 Here, C removes its entries for
this route, propagates RREP
back to B, who already has a
better route.
 If D had moved far away, then
C's infinite RREP might have
propagated all the way back to
S, forcing S to rediscover from
scratch.
Discrete Algs for Mobile Wireless Sys
H
I
G
S
A
F
D
E
B
C
41
AODV Simulation Results
 Simulated AODV using PARSEC event-driven, packet-level simulator.
 Not very realistic:
 Strict range cut-off.
 Perfect physical carrier sensing.
 Always a collision (losing both messages) if two neighbors broadcast
at the same time.
 Ran experiments with 50, 100, 500, and 1000 nodes, to test
scalability.
 1000 is very large compared to most MANET simulations.
 Delivery success rate:
 For few nodes, very high success rate.
 For more nodes: About 25% of messages lost.
 Main reason is collisions.
 Collisions are very important factor with MANET routing protocols.
Discrete Algs for Mobile Wireless Sys
42
AODV Real-World Experiments

Dartmouth experiments:









Approximately 40 students carrying laptops, moving randomly.
Playing field, 225 x 365m.
Traffic generation: 1200 byte packets, 5.5 packets per stream, 3 seconds
between packets, 15 seconds between streams (numbers derived from a
military application prototype).
Around 425 bytes per second – modest.
Message delivery success rate: 50%
Good latency, best compared to other algorithms tested (ODMRP, APRL,
STARA).
Second best for route costs (hop count).
Smallest amount of overhead.
Issues:



Collisions.
Fluctuating links: you might think you are somone’s neighbor because you
got a message, but the link quality is bad.
Control traffic: Aggressively seeking new routes can make things worse.
Discrete Algs for Mobile Wireless Sys
43
Disconnected Transitive Communication
[Chen, Murphy]
 Routing algorithms we’ve discussed assume a network that
stays connected, though the network topology can change.
 Now assume the network may not always be connected.
 Nodes in clusters, clusters intermittently connected.
 For applications that can tolerate highly asynchronous
communication.
 Minutes or hours delay, not milliseconds or seconds.
 Don't use to run a ssh connection.
 But perhaps to update two military camps about a change in
the plans for a battle coming up in a few days.
E
B
D
A
F
C
Discrete Algs for Mobile Wireless Sys
44
Disconnected Transitive Communication
[Chen, Murphy]
 Routing algorithms we’ve discussed assume a network that
stays connected, though the network topology can change.
 Now assume the network may not always be connected.
 Nodes in clusters, clusters intermittently connected.
 For applications that can tolerate highly asynchronous
communication.
 Minutes or hours delay, not milliseconds or seconds.
 Don't use to run a ssh connection.
 But perhaps to update two military camps about a change in
the plans for a battle coming up in a few days.
E
B
A
C
D
Discrete Algs for Mobile Wireless Sys
F
45
The Basic Idea
Algorithm maintains no long-term or
medium-term state, not even on-demand
routes.
Just pass the message on, hop-by-hop.
Pass it to the node in your current cluster
that is “closest” to the destination.
Figuring out who is closest is the main trick;
might require application-layer information.
Discrete Algs for Mobile Wireless Sys
46
More Detail
 If X has a message to send or pass on to D:
 X broadcasts probe throughout its component
(multi-hop).
 Recipients test their utility for D; respond if high
enough.
 Both broadcasting and response use low-level
protocol like DSR or AODV, within the cluster.
 X passes message on to node with highest utility.
Discrete Algs for Mobile Wireless Sys
47
Utility Measures
 History-based utility measures:


Most Recently Noticed: Higher utility if you have noticed
(encountered) D recently.
Most Frequently Noticed: Higher utility if you have noticed D
frequently.
 Future-based:

Future Plans: Assumes the application at your node will tell you
next time it expects to see D (e.g., by reading a calendar).
 Status-based:


Power: More power, better utility.
Rediscovery Interval: Smaller RDI means a better chance of
discovering D or someone closer to D.
 Weighted combinations of these.
 Sender can choose its own metric, based on special scenariospecific information.
Discrete Algs for Mobile Wireless Sys
48
Rediscovery
 Choosing a good rediscovery interval (RDI) is
important.
 If it’s too big, you may miss potential connections.
 Too small leads to too much traffic.
 Their solution:
 Double RDI after each unsuccessful probe for new
connections.
 If new node discovered, reset RDI to initial value.
 Discover new nodes using Hello messages.
Discrete Algs for Mobile Wireless Sys
49