UNIT –II Routing for Ad hoc Networks

Routing in Ad hoc Networks –Design Issues –Proactive, Reactive and

Hybrid Routing Protocols
Issues in Designing a Routing Protocol for Ad Hoc Wireless Networks

Routing protocols for ad hoc wireless networks face several significant challenges due to the
inherent characteristics of these networks, such as node mobility, limited resources, and
environmental constraints. Below is a detailed discussion of these challenges:

Mobility

 Dynamic Topology: The frequent movement of nodes leads to highly dynamic network
topologies, causing ongoing sessions to suffer from frequent path breaks.
 Path Disruption: Path breaks occur due to the movement of intermediate or end nodes,
unlike in wired networks with stationary nodes.
 Inefficiency of Wired Protocols: Traditional wired network routing protocols are
inadequate for ad hoc networks as they cannot adapt quickly to topology changes. Ad hoc
routing protocols must provide effective mobility management.

Bandwidth Constraint

 Limited Bandwidth: Unlike wired networks, wireless networks have constrained

bandwidth due to the limited radio spectrum.
 High Overhead: Frequent topology changes increase control overhead for maintaining
consistent topological information, leading to bandwidth wastage.
 Optimization Needs: Efficient routing protocols in ad hoc networks must minimize
overhead to optimize bandwidth usage, as complete topology information maintenance is
impractical.

Error-Prone Shared Broadcast Radio Channel

 Time-Varying Links: Wireless links have variable capacities and high error
probabilities, requiring the routing protocol to interact with the MAC layer for selecting
better-quality routes.
 Collisions: The shared nature of the wireless medium leads to frequent collisions of data
and control packets, necessitating protocols to address congestion and ensure reliable data
transfer.

Hidden and Exposed Terminal Problems

 Hidden Terminal Problem:

o Occurs when nodes outside the sender's range but within the receiver's range
transmit simultaneously, causing packet collisions at the receiver.
 o Example: If nodes A and C transmit to node B simultaneously, their packets
collide as A and C are unaware of each other.
o Solutions: Medium access protocols like MACA, MACAW, FAMA, and
DBTMA aim to mitigate this issue using mechanisms such as RTS-CTS
handshake and acknowledgments.
o Limitations: Despite these solutions, issues like improper CTS reception by
hidden nodes (e.g., node A in certain scenarios) may still disrupt communication.
 Exposed Terminal Problem:
o Happens when a transmitting node (e.g., node B) prevents another nearby node
(e.g., node C) from initiating transmission to another target (e.g., node D),
reducing spectrum reuse.
o Resolution: Requires separate transmitting and receiving frequencies for
simultaneous operations.

Resource Constraints

 Battery Life and Processing Power:

o Nodes in ad hoc networks are typically portable and constrained by limited
battery life and processing capabilities.
o Trade-offs: Enhancing power or processing increases size and reduces
portability, making efficient resource management essential.
 Protocol Adaptation: Routing protocols must minimize energy consumption and
processing demands to prolong the operational life of nodes.

CLASSIFICATIONS OF ROUTING PROTOCOLS

Routing protocols for ad hoc wireless networks can be classified into several types based on
different criteria.

 Routing Information Update Mechanism:

 Proactive (Table-driven) Routing Protocols: Maintain routing tables by exchanging

network topology information periodically. Suitable for consistent topology but may
introduce overhead in dynamic environments.
 Reactive (On-demand) Routing Protocols: Establish paths only when needed, avoiding
periodic updates and reducing overhead but potentially introducing delays.
 Hybrid Routing Protocols: Combine proactive methods for nearby nodes and reactive
methods for distant nodes, balancing efficiency and resource usage.

 Use of Temporal Information for Routing:

 Past Temporal Information: Decisions rely on historical or current link status, which
may not account for rapid topology changes.
  Future Temporal Information: Incorporates predictions of link or node status (e.g.,
battery life, mobility) to improve stability and efficiency.

 Routing Topology:

 Flat Topology: Employs simple addressing, similar to traditional LANs, with no

hierarchy.
 Hierarchical Topology: Introduces logical hierarchy based on geography or hop
distances, optimizing scalability and efficiency.

 Utilization of Specific Resources:

 Power-aware Routing: Focuses on conserving battery life, critical in portable ad hoc

devices.
 Geographical Information-assisted Routing: Utilizes location data to enhance routing
decisions and reduce control overhead.

TABLE-DRIVEN ROUTING PROTOCOLS

 These protocols are extensions of the wired network routing protocols.
 They maintain the global topology information in the form of tables at every node.
 These tables are updated frequently in order to maintain consistent and accurate network
state information.
 The destination sequenced distance-vector routing protocol (DSDV), wireless routing
protocol (WRP), source-tree adaptive routing protocol (STAR), and cluster-head gateway
switch routing protocol (CGSR) are some examples for the protocols that belong to this
category.

Destination Sequenced Distance-Vector Routing Protocol

The Destination Sequenced Distance-Vector (DSDV) routing protocol is a proactive, table-

driven routing protocol tailored for ad hoc wireless networks. It is an enhanced version of the
Bellman-Ford algorithm, adapted to prevent common issues like loops and the count-to-infinity
problem, and to ensure faster convergence. Below is a summary of the key points:

Operation of DSDV:

1. Routing Table Maintenance:

o Each node maintains a routing table with the shortest distance and the first node
on the shortest path to every other node.
o The routing tables are periodically exchanged between neighboring nodes to
maintain an updated view of the network topology.
o Incremental Updates: Used for minor changes in the topology and transmitted in
a single network data packet unit (NDPU).
 o Full Dumps: Triggered when significant topology changes occur and may require
multiple NDPUs.
2. Sequence Numbers:
o Sequence numbers are associated with routing table updates to ensure freshness
and prevent loops.
o Even sequence numbers are generated by destinations, while odd sequence
numbers are used to indicate link breaks.
3. Route Setup:
o Routes are readily available as all nodes maintain a global topology view.
o Example: If node 1 wants to send data to node 15, the routing table shows that the
shortest path is through node 5 with a distance of 4 hops.
4. Route Maintenance:
o Link breaks trigger updates with a distance set to infinity (∞) and a higher
sequence number.
o The broken link information propagates throughout the network to update all
nodes.
o Nodes receiving these updates modify their routing tables and rebroadcast the
information.

Consider the example as shown in Figure 7.5 (a). Here node 1 is the source node and node 15
is the destination. As all the nodes maintain global topology information, the route is already
available as shown in Figure 7.5 (b). Here the routing table of node 1 indicates that the
shortest route to the destination node (node 15) is available through node 5 and the distance
to it is 4 hops, as depicted in Figure 7.5 (b).
Consider the case when node 11 moves from its current position, as shown in Figure 7.6. When a
neighbor node perceives the link break, it sets all the paths passing through the broken link with
distance as ∞. For example, when node 10 knows about the link break, it sets the path to node 11
as ∞ and broadcasts its routing table to its neighbors. Those neighbors detecting significant
changes in their routing tables rebroadcast it to their neighbors. In this way, the broken link
information propagates throughout the network. Node 1 also sets the distance to node 11 as ∞.
When node 14 receives a table update message from node 11, it informs the neighbors about the
shortest distance to node 11. This information is also propagated throughout the network. All
nodes receiving the new update message with the higher sequence number set the new distance
to node 11 in their corresponding tables. The updated table at node 1 is shown in Figure 7.6,
where the current distance from node 1 to node 11 has increased from three to four hops.

Advantages:

1. Immediate Route Availability:

o Routes to all destinations are pre-computed, resulting in minimal delay during
route setup.
2. Adaptability:
o Sequence number tagging and incremental updates make DSDV adaptable to the
dynamic nature of ad hoc networks while building on existing wired network
protocols.
 3. Global Topology View:
o Nodes maintain an up-to-date understanding of the network topology.

Disadvantages:

1. High Control Overhead:

o Frequent table updates, especially during high mobility, generate significant
overhead.
o This overhead can overwhelm limited bandwidth, reducing the protocol’s
scalability.
2. Stale Information:
o Nodes may rely on outdated routing information if they must wait for the next
table update.
3. Bandwidth Consumption:
o The protocol’s requirement for regular table updates can choke bandwidth in
dynamic or large networks.

Wireless Routing Protocol

Overview and Key Features

1. Purpose:
Like DSDV, WRP ensures each node has an up-to-date view of the network with the
shortest distance to all destination nodes.
2. Count-to-Infinity Problem Prevention:
WRP counters the count-to-infinity problem and achieves faster convergence by
maintaining additional detailed information about the network.
3. Tables Maintained by Nodes:
Unlike DSDV, which uses a single topology table, WRP uses four types of tables for
accuracy:
o Distance Table (DT): Stores neighbor-reported distances and penultimate nodes
for each destination.
o Routing Table (RT): Maintains the shortest distance, predecessor, successor, and
path status for all destinations.
o Link Cost Table (LCT): Tracks the cost of links and monitors the number of
update intervals since the last successful update.
o Message Retransmission List (MRL): Maintains messages to be retransmitted
and their acknowledgment status.
4. Retransmission Mechanism:
o Nodes retransmit update messages until acknowledgments are received or the
retransmission counter expires.
o Missed updates trigger retransmission to ensure convergence.
 5. Convergence Speed:
By simultaneously considering the distances reported by other neighbors during an
update, WRP converges faster than DSDV.

Route Maintenance

1. Detecting Link Breaks:

o When a link fails, nodes set the link cost to infinity (∞) and propagate this update
to neighbors.
2. Reaction to Updates:
o Nodes receiving the update delete routes involving the broken link and seek
alternative paths if available in the DT.
o Affected nodes update routing tables only if the new route is better.

Route Establishment Example

Consider the example shown in Figure 7.7, where the source of the route is node 1 and the
destination is node 15. As WRP proactively maintains the route to all the destinations, the route
to any destination node is readily available at the source node. From the routing table shown in
Figure 7.7, the route from node 1 to node 15 has the next node as node 2. The predecessor node
of 15 corresponding to this route is node 12. The predecessor information helps WRP to
converge quickly during link breaks.

Route Maintenance Example

Consider the scenario shown in Figure 7.8. When the link between nodes 12 and 15 breaks, all
nodes having a route to the destination with predecessor as node 12 delete their corresponding
routing entries. Both node 12 and node 15 send update messages to their neighbors indicating
that the cost of the link between nodes 12 and 15 is ∞. If the nodes have any other alternative
route to the destination node 15, they update their routing tables and indicate the changed route
to their neighbors by sending an update message. A neighbor node, after receiving an update
message, updates its routing table only if the new path is better than the previously existing
paths. For example, when node 12 finds an alternative route to the destination through node 13, it
broadcasts an update message indicating the changed path. After receiving the update message
from node 12, neighboring nodes 8, 14, 15, and 13 do not change their routing entry
corresponding to destination 15 while node 4 and node 10 modify their entries to reflect the new
updated path. Nodes 4 and 10 again send an update message to indicate the correct path to the
destination for their respective neighbors. When node 10 receives node 4's update message, it
again modifies its routing entry to optimize the path to the destination node (15) while node 4
discards the update entry it received from node 10.

Advantages of WRP

1. Faster Convergence:
o The use of predecessor information accelerates recovery from link breaks.
2. Reduced Table Updates:
o Fewer table updates compared to DSDV.

Disadvantages of WRP

1. Resource Intensive:
o Requires more memory and processing power due to multiple tables.
2. Unsuitability for High Dynamics:
o At high mobility, control overhead becomes comparable to DSDV, limiting its
scalability.
3. Not Ideal for Large Networks:
 o Control overhead increases with network size, making WRP unsuitable for very
large ad hoc networks.

Cluster-Head Gateway Switch Routing Protocol

The Cluster-Head Gateway Switch Routing Protocol (CGSR) is a hierarchical routing
protocol designed for ad hoc wireless networks, where nodes are organized into clusters, and
each cluster has a cluster-head responsible for coordinating the communication within that
cluster. The cluster-heads are dynamically elected using the Least Cluster Change (LCC)
algorithm, which ensures minimal changes in the network topology. Nodes communicate
through these cluster-heads, and inter-cluster communication happens via gateways, which are
nodes that belong to multiple clusters.

Key Features:

1. Cluster-Head Election:
o Elected dynamically using the Least Cluster Change (LCC) algorithm.
o Ties are resolved using lowest ID or highest connectivity.
o A node stops being a cluster-head if it comes under the range of another cluster-
head.
2. Cluster Coordination:
o Cluster-heads manage bandwidth allocation and channel access using token-
based polling protocols.
o Communication between clusters occurs via gateways, which are nodes
belonging to multiple clusters.
3. Routing Mechanism:
o Extends the Destination Sequenced Distance-Vector (DSDV) protocol.
o Each node maintains a cluster member table and a routing table.
o Packets are routed through cluster-heads and gateways, ensuring hierarchical
path establishment.
4. Route Maintenance:
o Managed by table updates for stale entries.
o LCC algorithm handles cluster-head changes to minimize disruptions.

Figure 7.9 shows the clusterheads, cluster gateways, and normal cluster member nodes in an
ad hoc wireless network. A path between node 11 and node 16 would follow 11 - 6 - 12 - 5 -
16. Since the cluster-heads gain more opportunities for transmission, the clusterheads, by
means of a dynamic scheduling mechanism, can make CGSR obtain better delay
performance for real-time flows. Route reconfiguration is necessitated by mainly two factors:
firstly, the change in cluster-head and secondly, the stale entries in the cluster member table
and routing table.
Advantages:

 Efficient Bandwidth Utilization: Clustering allows better allocation and reuse of

bandwidth.
 Priority Scheduling: Enables dynamic scheduling with token and code management.
 Reduced Routing Overhead: Clustering minimizes routing complexity compared to flat
topology.

Disadvantages:

 Path Length Increase: Routing through cluster-heads and gateways can lengthen paths.
 Instability at High Mobility: Frequent cluster-head changes lead to instability.
 Resource Demands: Gateways require additional interfaces to avoid conflicts.
 Battery Drain: High energy consumption at cluster-heads causes frequent changes.
ON-DEMAND ROUTING PROTOCOLS
Unlike the table-driven routing protocols, on-demand routing protocols execute the path-finding
process and exchange routing information only when a path is required by a node to
communicate with a destination.

Dynamic Source Routing Protocol

The Dynamic Source Routing (DSR) protocol is an on-demand routing protocol specifically
designed for ad hoc wireless networks. The primary goal of DSR is to reduce the bandwidth
consumed by control packets by eliminating the need for periodic table-update messages, which
are typical in table-driven protocols. Here’s a detailed explanation:

Key Concepts and Operation of DSR:

1. On-demand Routing: DSR operates by creating routes only when needed, rather than
maintaining constant routes like in table-driven protocols. This is achieved through the
exchange of RouteRequest (RREQ) and RouteReply (RREP) packets.
2. Beacon-less Operation: Unlike other on-demand protocols, DSR does not rely on
periodic beacon or hello packets (which are used to inform neighbors of a node's
presence), reducing overhead.
3. Route Discovery Process:
o When a source node has data to send to a destination but doesn’t have a route, it
initiates a RouteRequest (RREQ) packet.
o This RREQ is flooded throughout the network. When a node receives a RREQ, it
rebroadcasts it to its neighbors, unless it is the destination node or has already
forwarded the packet.
o Each RREQ contains a sequence number and the path traversed by the request,
which helps to avoid routing loops and duplicate transmissions. A node forwards
the RREQ only if it has not received the same request before.
o The destination node, upon receiving the RREQ, responds with a RouteReply
(RREP) packet. This RREP includes the full path back to the source.
4. Route Cache:
o Each node keeps a route cache, which stores previously learned routes. This
cache is used to reply to RREQ when a node has an existing route to the
destination.
 oIf a node has a route in its cache to the destination, it can immediately send a
RREP to the source, skipping the route discovery process.
5. Promiscuous Mode:
o In promiscuous mode, nodes can overhear packets not addressed to them. This
allows nodes to gather information about the routes that other data packets are
following. This helps in maintaining an up-to-date route cache and facilitates
quicker route replies.

Optimizations in DSR:

1. Route Cache Utilization: Intermediate nodes can use the route cache to respond to
RREQ packets, thereby reducing the need to flood the network with route discovery
requests.
2. Route Maintenance and Error Handling:
o If a link in the route breaks (due to node movement), a RouteError (RERR)
message is generated by the adjacent node and sent to the source, informing it of
the broken link. The source node then reinitiates the route discovery process.
o Route Cache Updates: Intermediate nodes and the source can update their cache
based on new route information received via RREQ or RREP packets.
3. Exponential Backoff: To prevent frequent flooding of RREQ packets during network
partitions (when the destination is unreachable), exponential backoff is used to reduce
the frequency of route requests.
4. Piggybacking: DSR allows piggybacking of data packets onto the RREQ. This means
that a data packet can travel along with the route request, improving efficiency.

Route establishment Example

In Figure 7.10, source node 1 initiates a RouteRequest packet to obtain a path for destination
node 15.
If optimization is not allowed in the DSR protocol, the route construction phase is very simple.
All the intermediate nodes flood the RouteRequest packet if it is not redundant. For example,
after receiving the RouteRequest packet from node 1 (refer to Figure 7.10), all its neighboring
nodes, that is, nodes 2, 5, and 6, forward it. Node 4 receives the RouteRequest from both nodes 2
and 5. Node 4 forwards the first RouteRequest it receives from any one of the nodes 2 and 5 and
discards the other redundant/duplicate RouteRequest packets. The RouteRequest is propagated
till it reaches the destination which initiates the RouteReply. For example, in Figure 7.11, if the
intermediate node 10 has a route to the destination via node 14, it also sends the RouteReply to
the source node. The source node selects the latest and best route, and uses that for sending data
packets. Each data packet carries the complete path to its destination.
 Figure 7.10. Route establishment in DSR.

Route Maintenance Example

When an intermediate node in the path moves away, causing a wireless link to break, for
example, the link between nodes 12 and 15 in Figure 7.11, a RouteError message is generated
from the node adjacent to the broken link to inform the source node. The source node reinitiates
the route establishment procedure. The cached entries at the intermediate nodes and the source
node are removed when aRouteError packet is received. If a link breaks due to the movement of
edge nodes (nodes 1 and 15), the source node again initiates the route discovery process.
Advantages of DSR:

1. Reduced Overhead: Since routes are only discovered when needed, there’s no
continuous table updating as in table-driven protocols, leading to lower overhead.
2. Efficient Route Cache: Intermediate nodes use the route cache to respond to route
requests, reducing the need for full-scale route discovery every time a source needs to
send data.
3. Flexible: It works well in environments with relatively low mobility or static networks,
where route discovery is infrequent.

Disadvantages of DSR:

1. No Local Repair: When a link in the route breaks, DSR cannot locally repair the route.
The source must restart the route discovery process. This can cause delays, especially in
highly dynamic networks.
2. Stale Cache Information: Route cache can become outdated, leading to potential route
inconsistencies. If cache entries are stale, it can result in routing failures or inefficiencies
during route reconstruction.
3. Higher Setup Delay: In comparison to table-driven protocols, DSR has a higher
connection setup delay, as routes must be discovered during data transmission.
4. Scalability Issues: DSR may not scale well in large or highly mobile networks because
the routing overhead increases with the number of hops and the dynamic nature of the
nodes. As mobility increases, the overhead and delay due to route discoveries become
more significant.
5. Routing Overhead: DSR uses source routing, which means the complete route is
included in the header of each data packet. This increases the packet size, especially in
networks with long routes, leading to higher overhead proportional to the path length.

Ad Hoc On-Demand Distance-Vector Routing Protocol

The Ad Hoc On-Demand Distance Vector (AODV) routing protocol is an on-demand routing
mechanism used in ad hoc networks, where routes are established only when needed by a source
node for data transmission. AODV employs destination sequence numbers to ensure the most
recent path is used, distinguishing it from protocols like Dynamic Source Routing (DSR), which
uses source routing. Below is a comprehensive understanding:

Key Features of AODV:

1. On-Demand Route Discovery:

o Routes are established only when a source node requires them.
o The source floods the network with a RouteRequest (RREQ) if no existing route
to the destination is available.
2. Destination Sequence Number:
o Ensures freshness of the route.
o Nodes only update their route if the received sequence number is greater than the
stored one.
 3. RouteRequest Packet:
o Contains fields like source ID, destination ID, sequence numbers, broadcast ID,
and time-to-live (TTL).
o Prevents duplicate processing of RouteRequest using the unique combination of
broadcast ID and source ID.
4. RouteReply Packet:
o Sent by intermediate nodes or the destination if a valid route exists.
o Intermediate nodes forward RouteReply while updating their routing tables with
the most recent sequence number.
5. Route Maintenance:
o AODV does not locally repair broken links.
o Link failures trigger RouteError messages to inform end nodes.
o Source node initiates a new path-finding process upon link break.

Example Scenarios:

1. Route Establishment:
o A source floods a RouteRequest in the network.
o Intermediate nodes forward the request or reply if they have valid routes.
o If multiple RouteReplies reach the source, the most recent and shortest path is
selected.
2. Route Maintenance:
o When a link breaks, RouteError messages notify nodes.
o End nodes remove broken path entries, and the source starts a new route
discovery.

Route Establishment
Consider the example depicted in Figure 7.12. In this figure, source node 1 initiates a path-
finding process by originating a RouteRequest to be flooded in the network for destination
node 15, assuming that the RouteRequest contains the destination sequence number as 3 and
the source sequence number as 1. When nodes 2, 5, and 6 receive the RouteRequest packet,
they check their routes to the destination. In case a route to the destination is not available,
they further forward it to their neighbors. Here nodes 3, 4, and 10 are the neighbors of nodes
2, 5, and 6. This is with the assumption that intermediate nodes 3 and 10 already have routes
to the destination node, that is, node 15 through paths 10-14-15 and 3-7-9-13-15,
respectively. If the destination sequence number at intermediate node 10 is 4 and is 1 at
intermediate node 3, then only node 10 is allowed to reply along the cached route to the
source. This is because node 3 has an older route to node 15 compared to the route available
at the source node (the destination sequence number at node 3 is 1, but the destination
sequence number is 3 at the source node), while node 10 has a more recent route (the
destination sequence number is 4) to the destination. If the RouteRequest reaches the
destination (node 15) through path 4-12-15 or any other alternative route, the destination also
sends a RouteReply to the source. In this case, multipleRouteReply packets reach the source.
All the intermediate nodes receiving a RouteReply update their route tables with the latest
destination sequence number. They also update the routing information if it leads to a shorter
path between source and destination.
Route Maintenance:

Consider the example illustrated in Figure 7.13. When a path breaks, for example, between nodes
4 and 5, both the nodes initiate RouteError messages to inform their end nodes about the link
break. The end nodes delete the corresponding entries from their tables. The source node
reinitiates the path finding process with the new BcastID and the previous destination sequence
number.

Figure 7.13. Route maintenance in AODV.

Advantages:

1. On-Demand Routing: Routes are established only when needed, reducing overhead
during idle periods.
2. Fresh Routes: Destination sequence numbers ensure the use of up-to-date paths.
3. Connection Setup Efficiency: Quick route establishment due to on-demand nature.

Disadvantages:

1. Inconsistent Routes: Intermediate nodes with outdated sequence numbers may create
stale routes.
2. Control Overhead: Multiple RouteReply packets in response to a single RouteRequest
increase control traffic.
3. Beaconing Overhead: Periodic beacons consume bandwidth unnecessarily.

Temporally Ordered Routing Algorithm

The Temporally Ordered Routing Algorithm (TORA) is a source-initiated, on-demand

routing protocol that employs a link-reversal mechanism to provide loop-free multipath routes
to a destination. Below is a summary:

Key Features of TORA:

1. Local Topology Information:

Each node maintains information about its one-hop neighbors and can detect network
partitions.
2. Loop-Free Multipath Routing:
TORA ensures multiple routes to the destination and avoids routing loops.
3. Control Packet Limitation:
During reconfiguration after a path break, control packets are confined to a small region
of the network, reducing overhead.
4. Distance Metric:
o The height of a node, denoted as H(N)H(N)H(N), indicates its distance from the
destination.
o Routing decisions are made based on the relative heights of nodes.

TORA Operations:

1. Route Establishment:
o Triggered when a source node needs a path to the destination but does not have
one.
 oUses a Query/Update mechanism:
 The source sends a Query packet containing the destination address.
 Intermediate nodes forward the Query until it reaches either the
destination or a node with a valid route to the destination.
 The destination replies with an Update packet, setting its height to zero.
Each node that receives this packet sets its height to a value higher than
the sender's, creating a directed acyclic graph (DAG).
2. Route Maintenance:
o If an intermediate node detects a broken link:
 It increases its height value and originates an Update packet.
 Neighboring nodes receiving the Update packet adjust their links
accordingly to maintain the DAG.
o If no valid path exists, the source initiates a new Query/Update process.
o In case of network partition detection:
 A Clear message is broadcasted to erase all route information in the
affected partition.
3. Route Erasure:
o When a partition is detected, TORA erases outdated routing information by
sending Clear messages.

Figure 7.14 shows the distance metric used in TORAwhich is nothing but the length of the
path, or the height from the destination. H(N) denotes the height of node N from the
destination.
Figure 7.14. Illustration of temporal ordering in TORA.

Consider the network topology shown in Figure 7.14. When node 1 has data packets to be sent to
the destination node 7, a Query packet is originated by node 1 with the destination address
included in it. This Query packet is forwarded by intermediate nodes 2, 3, 4, 5, and 6, and
reaches the destination node 7, or any other node which has a route to the destination. The node
that terminates (in this case, node 7) the Query packet replies with an Update packet containing
its distance from the destination (it is zero at the destination node). In the example, the
destination node 7 originates an Update packet. Each node that receives the Update packet sets
its distance to a value higher than the distance of the sender of the Update packet. By doing this,
a set of directed links from the node which originated the Query to the destination node 7 is
created. This forms the DAGdepicted in Figure 7.14. Once a path to the destination is obtained, it
is considered to exist as long as the path is available, irrespective of the path length changes due
to the reconfigurations that may take place during the course of the data transfer session.

When an intermediate node (say, node 5) discovers that the route to the destination node is
invalid, as illustrated in Figure 7.15, it changes its distance value to a higher value than its
neighbors and originates an Update packet. The neighboring node 4 that receives the Update
packet reverses the link between 1 and 4 and forwards the Update packet. This is done to update
the DAG corresponding to destination node 7. This results in a change in the DAG. If the source
node has no other neighbor that has a path to the destination, it initiates a fresh Query/Update
procedure. Assume that the link between nodes 1 and 4 breaks. Node 4 reverses the path between
itself and node 5, and sends an update message to node 5. Since this conflicts with the earlier
reversal, a partition in the network can be inferred. If the node detects a partition, it originates a
Clear message, which erases the existing path information in that partition related to the
destination.

Figure 7.15. Illustration of route maintenance in TORA.

Advantages of TORA:

1. Localized Control Overhead:

Control packets are limited to a small area during reconfiguration, reducing network-wide
congestion.
2. Partition Detection:
TORA can identify network partitions and ensure invalid routes are cleared.
3. Multipath Routing:
Provides alternate paths to the destination, increasing robustness.

Disadvantages of TORA:

1. Non-Optimal Routes:
Local reconfiguration may result in suboptimal paths between source and destination.
 2. Transient Loops and Oscillations:
Concurrent partition detection and route deletions can temporarily cause routing loops or
instabilities.

Location-Aided Routing (LAR)

Key Concepts:

1. Purpose:
LAR uses geographical location information to optimize routing by reducing control
overhead, improving efficiency, and limiting the propagation of control packets.
2. Mechanisms Used:
o Global Positioning System (GPS): Essential for determining nodes' locations.
o ExpectedZone: Region where the destination node is expected to be based on
past location and mobility.
o RequestZone: Defined area where control packets are forwarded to reduce
unnecessary flooding.
3. Routing Algorithms:
o LAR1:
 The source node specifies the RequestZone explicitly, often as a
rectangular region that includes both the source and ExpectedZone.
 Nodes outside the RequestZone discard packets.
 Destination sends a RouteReply with location, time, and optionally, speed.
o LAR2:
 The source node specifies its distance to the destination and the
destination's coordinates.
 Intermediate nodes forward packets only if their distance to the destination
is shorter than the distance between the source and the destination (with
allowance for an error margin δ\deltaδ).
4. RouteReply:
o Destination node responds with a RouteReply packet containing routing
information for future data transmission.

As per LAR1, as illustrated in Figure 7.16, the RequestZone is the smallest rectangle that
includes the source node (S) and the ExpectedZone, the sides of which are parallel to the X
and Y axes, when the node S is outside the ExpectedZone. When node S is within the
ExpectedZone,then the RequestZone is reduced to the ExpectedZone itself. Every
intermediate node that receives theRouteRequest packet verifies the RequestZone information
contained in the packet and forwards it further if the node is within the RequestZone;
otherwise, the packet is discarded. In Figure 7.16, the source node (node 1) originates a
RouteRequest, which is broadcast to its neighbors (2, 5, and 6). These nodes verify their own
geographical locations to check whether they belong to the ExpectedZone.Nodes 2 and 5 find
that they are inside the ExpectedZone and hence they forward the RouteRequest.But node
6 discards the packet. Finally, when the RouteRequest reaches the destination node (node 4),
it originates a RouteReply that contains the current location and current time of the node.
Also, as an option, the current speed of movement can be included in the RouteReply if that
 information is available with the node. Such information included in the RouteReply packet is
used by the source node for future route establishment procedures.

Figure 7.16. RequestZone and ExpectedZone in LAR1.

In LAR2 algorithm (Figure 7.17), the source node S (node 5) includes the distance between itself
and the destination node D (node 8) along with the (X, Y) coordinates of the destination node D
in theRouteRequest packet instead of the explicit information about the Expected Region. When
an intermediate node receives this RouteRequest packet, it computes the distance to the node D.
If this distance is less than the distance from S to node D + δ, where δ is a parameter of the
algorithm decided based on the error in location estimation and mobility, then the RouteRequest
packet is forwarded. Otherwise, the RouteRequest is discarded.
Consider the example illustrated in Figure 7.17. Assume that the value of δ is 0 here. The
RouteRequest packet originated by node 5 is received by nodes 1, 2, 4, 6, and 10. Only nodes 4
and 10 find that the distance between
them and the destination is less than the distance between the node 5 and the destination node;
other nodes discard the RouteRequest. ARouteRequest packet is forwarded only once and the
distance between the forwarding node and D is updated in the RouteRequest packet for further
relaying. When node 4 forwards the RouteRequestpacket, it updates the packet with the distance
between itself and the destination node D. This packet, after being received at neighbor node 3,
is discarded due to the fact that the distance between node 3 and the node 8 is greater than the
distance between nodes 4 and 8. Once the RouteRequest reaches node 8, it originates a
RouteReply packet back to the source node 5, containing the path through which future data
packets are to be propagated. In order to compensate for the location error (due to the inaccuracy
of GPS information or due to changes in the mobility of the nodes), a larger RequestZonethat can
accommodate the amount of error that occurred is considered.
Advantages:

o Reduced control overhead through limited flooding.

o Efficient use of bandwidth and geographical position information.

Disadvantages:

o Requires GPS or a similar infrastructure for accurate location data.

o Inapplicable in scenarios without access to location information.

Signal Stability-Based Adaptive Routing Protocol

Signal Stability-Based Adaptive Routing Protocol (SSA) is an on-demand routing protocol that
prioritizes stable routes based on signal strength. It is beacon-based, using signal stability as the
main factor for route selection.

Key Components:

1. Signal Strength Measurement:

o Links are classified as stable or unstable based on the strength of received
beacons.
o Nodes maintain a Signal Stability Table (SST) to track beacon counts and signal
strengths from neighbors.
 2. Protocols:
o Dynamic Routing Protocol (DRP): Maintains the routing table and interacts
with DRP processes on other nodes.
o Forwarding Protocol (FP): Forwards packets to their destinations based on
routes.

Working Mechanism:

1. Route Discovery:
o RouteRequest packets are broadcasted by the source.
o Only RouteRequest packets received over stable links are processed and
forwarded by intermediate nodes.
o Weak link packets are dropped.
o The destination selects the first RouteRequest received over a stable path and
sends a RouteReply.
2. Route Maintenance:
o When a link breaks, end nodes notify the source and destination.
o The source rebroadcasts a RouteRequest to establish a new stable path.
o If no stable path is found, weak links are considered after multiple failed attempts.

Consider Figure 7.20, where the source node (node 1) broadcasts a RouteRequest for finding
the route to the destination node (node 15). In Figure 7.20, solid lines represent the stable
links, while the dotted lines represent weak links. Unlike ABR, SSA restricts intermediate
nodes from forwarding aRouteRequest packet if the packet had been received over a weak
link. It forwards only RouteRequestpackets received over stable links. In Figure 7.20, when
the RouteRequest is initiated by the source, it is to be processed by all its neighbors. But
before processing, each neighbor node checks whether theRouteRequest packet was received
through a stable link. If the RouteRequest had been received through a stable link and had not
been sent
already (i.e., it is not a duplicate RouteRequest), it is forwarded by the node; otherwise, it is
dropped. For example, when the RouteRequest from node 1 reaches nodes 2, 5, and 6, it is
forwarded only by nodes 2 and 5 as the link between nodes 1 and 6 is weak. Similarly, the
RouteRequest forwarded by node 2 is rejected by nodes 3 and 5, while node 4 forwards it to
its neighbors, provided it has not already been forwarded. In this way, nodes forward
theRouteRequest until it reaches the destination. In this example, only one route is
established based on the strong link criterion. A RouteRequest propagated through path 1-5-
10-14-15 is rejected by the destination as it receives a RouteRequest from node 14 on a weak
link. The stable path consisting of strong links is 1- 2 (or 5)-4-8-13-15. The first
RouteRequest packet that reaches the destination over a stable path is selected by the
destination.
 Figure 7.20. Route establishment in SSA.

As shown in Figure 7.21, when a link breaks, the end nodes of the broken link (i. e., nodes 2 and
4) notify the corresponding end nodes of the path (i. e., nodes 1 and 15). A source node, after
receiving a route break notification packet, rebroadcasts the RouteRequest to find another stable
path to the destination. Stale entries are removed only if data packets that use the stale route
information fail to reach the next node. If the link between nodes 2 and 4 breaks, a new strong
path is established through 1-5-4-8-13-15. If no strong path is available when a link gets broken
(e.g., link 8-13), then the new route is established by considering weak links also. This is done
when multiple RouteRequest attempts fail to obtain a path to the destination using only the stable
links.

Figure 7.21. Route maintenance in SSA.

Advantages:

1. Selects stable routes, reducing frequent route failures.

2. Accommodates temporal stability using beacon counts.
3. Improves reliability compared to shortest-path protocols like DSR and AODV.

Disadvantages:

1. Strict strong link criterion leads to frequent RouteRequest failures.

2. Multiple flooding attempts increase bandwidth usage and path setup time.
3. Preference for stable paths can increase route length, ignoring shorter paths.

Hybrid Routing Protocols

Core Extraction Distributed Ad Hoc Routing Protocol
The Core Extraction Distributed Ad Hoc Routing Protocol (CEDAR) is a specialized routing
protocol designed for ad hoc networks that combines routing with Quality of Service (QoS)
support. Here's a simplified breakdown:

1. Core Nodes and Dominating Set:

o Core nodes (dominator nodes) represent a subset of network nodes ensuring all
others are either part of this subset or neighbors to it.
o This subset approximates the minimum dominating set, ensuring efficient
network coverage.
2. Core Node Role:
o Core nodes broadcast packets efficiently (core broadcast) to minimize the number
of nodes involved.
o Core nodes also maintain local topology information to manage and optimize
routing decisions.
3. Link State Propagation:
o Increase-waves: Slowly propagate high-bandwidth, stable link information across
the network.
o Decrease-waves: Quickly notify nearby nodes about low-bandwidth or unstable
links, keeping this information localized.
4. Route Establishment:
o Phase 1 (Core Path Discovery): A path is established between the source and
destination core nodes using virtual links.
o Phase 2 (QoS Path Selection): The shortest-widest path (shortest path with
maximum bandwidth) is iteratively identified until the destination is reached.
5. Handling Route Failures:
o When a link breaks, the node at the break point attempts to locally recompute the
route.
 o Notifications are sent to the source node, which stops transmitting and initiates a
new route discovery if needed.

Consider Figure 7.24 where the source is node 1 and the destination is node 15. The core
nodes in the network are nodes 3, 5, 11, 12, and 13. In this figure, node 5 is the dominator of
nodes 1 and 6. Similarly, node 12 is the dominator of node 15. When node 1 initiates a
RouteRequest to be flooded throughout the network, it intimates its core node the <source id,
destination id> pair information. If the core node 5 does not have any information about the
dominator of node 15, which is the destination node, it initiates a core broadcast. Due to this,
all nearby core nodes receive the request in the unicast transmission mode. This unicast
transmission is done on the virtual links. For core node 5, the virtual link with core node 3
comprises of the links 5-2 and 2-3, while the virtual link between core nodes 5 and 13 is
represented by path 5-4-8-13. When a core node receives the core broadcast message, it
checks whether the destination is its core member. A core node having the destination as one
of its core members replies to the source core node. In our case, core node 12 replies to core
node 5. The path between core nodes 12 and 5 constitutes a core path. Once a core path is
established, the feasibility of the path in terms of the availability of the required bandwidth is
checked. If the required bandwidth is available on the path, the connection is established;
otherwise, the core path is rejected.
Consider the network topology shown in Figure 7.25. When the link between nodes 12 and 15
breaks, node 12 tries to reconnect to the destination using an alternate path that satisfies the
bandwidth requirement. It also notifies the source node about the link break. The source node
tries to reconnect to the destination by reinitiating the route establishment process. In case node
12 does not have any other feasible path, then the alternate path 1-5-4-8-13-15 found by the
source node is used for the further routing of packets.

6. Advantages:
o Efficient routing with QoS support.
o Reliable core broadcast and propagation of stable, high-bandwidth link
information.
7. Disadvantages:
o High control overhead due to core node updates.
o Protocol performance depends on the stability of core nodes; their movement can
disrupt the network.

Zone Routing Protocol (ZRP)

Zone Routing Protocol (ZRP) is a hybrid routing protocol designed to combine the advantages
of both proactive and reactive routing methods. Here's an overview of its key features and
working principles:
Key Concepts:

1. Hybrid Approach:
o ZRP utilizes both proactive and reactive routing schemes. It works by having a
proactive routing protocol within a local zone (the r-hop neighborhood) and a
reactive routing protocol for nodes beyond this zone.
2. Routing Zones:
o Every node in the network has a routing zone defined by a radius (r). The zone
consists of nodes that are reachable within r hops from the node.
o The zone is split into two parts:
 Interior nodes: Nodes within the zone radius (e.g., nodes that are less
than or equal to r hops away).
 Peripheral nodes: Nodes that are exactly r hops away from the central
node.
3. Intra-Zone Routing Protocol (IARP):
o Inside its zone, a node uses the IARP to maintain route information to all other
nodes within the zone. This is done by periodic route updates.
o Larger zones generate more control traffic because more route updates need to be
exchanged.
4. Inter-Zone Routing Protocol (IERP):
o For nodes outside the local routing zone, IERP is used. IERP relies on the
information available within the node's routing zone.
o If a node wants to send data to a node outside its zone, it sends a RouteRequest
to its peripheral nodes. If any peripheral node can find the destination within its
own zone, it responds with a RouteReply.
o The process of finding the destination can involve bordercasting, where the
RouteRequest is passed along to the neighboring peripheral nodes until the
destination is located.

5. Path Finding and Optimization:

o The source node, upon receiving multiple RouteReply packets, can choose the
best path based on criteria such as the shortest path or the least delay path.
o In case of a link failure, a local reconfiguration is performed, where a new
alternate path bypasses the broken link. This results in sub-optimal paths initially,
but global path-finding can be triggered later to find optimal routes.

Figure 7.26 illustrates routing zones of node 8, with r = 1 hop and r = 2 hops. With zone
radius = 2, the nodes 7, 4, 12, and 13 are interior nodes, whereas nodes 2, 3, 5, 9, 10, 13, and
15 are peripheral nodes (nodes with the shortest distance equal to the zone radius). Each node
maintains the information about routes to all nodes within its routing zone by exchanging
periodic route update packets (part of IARP). Hence the larger the routing zone, the higher
the update control traffic.

Figure 7.26. Routing zone for node 8 in ZRP.

The IERP is responsible for finding paths to the nodes which are not within the routing zone.
IERPeffectively uses the information available at every node's routing zone. When a node s
(node 8 in Figure 7.27) has packets to be sent to a destination node d (node 15 in Figure 7.27), it
checks whether node d is within its zone. If the destination belongs to its own zone, then it
delivers the
packet directly. Otherwise, node s bordercasts (uses unicast routing to deliver packets directly to
the border nodes) theRouteRequest to its peripheral nodes. In Figure 7.27 node 8 bordercasts
RouteRequests to nodes 2, 3, 5, 7, 9, 10, 13, 14, and 15. If any peripheral node finds node d to be
located within its routing zone, it sends a RouteReply back to node s indicating the path;
otherwise, the node rebordercasts theRouteRequest packet to the peripheral nodes. This process
continues until node d is located. Nodes 10 and 14 find the information about node 16 to be
available in their intra-zone routing tables, and hence they originate RouteReply packets back to
node 8. During RouteRequest propagation, every node that forwards the RouteRequest appends
its address to it. This information is used for delivering theRouteReply packet back to the source.
Advantages:

 Reduced Overhead: By combining proactive and reactive approaches, ZRP reduces

control overhead compared to purely on-demand or table-driven routing protocols.
 Efficient Route Maintenance: The proactive approach within the zone helps maintain
up-to-date routing information, while the reactive method is employed for nodes outside
the zone, making the system efficient for routing over a wide area.

Disadvantages:

 Higher Overhead in Large Zones: In cases where the routing zone is large, more
control traffic is required to exchange route updates, leading to higher overhead.
 Duplicate RouteRequests: Without proper query control, there could be redundant or
duplicate RouteRequests forwarded within overlapping zones, causing additional control
traffic.
 Impact of Zone Radius: The performance of the protocol depends heavily on the chosen
zone radius. If the radius is too large or too small, it can negatively affect the
performance by increasing control traffic or failing to maintain optimal routes.

Zone-Based Hierarchical Link State Routing Protocol

The Zone-Based Hierarchical Link State Routing Protocol (ZHLS) is a hybrid hierarchical
routing protocol that combines proactive and reactive routing techniques. It leverages the
geographical location information of nodes to form non-overlapping zones, with each zone
having its own topology for routing. Here's a breakdown of how it works:

Key Concepts of ZHLS:

1. Hierarchical Addressing: ZHLS uses a hierarchical addressing scheme where each node
has a unique zone ID and a node ID. Nodes are grouped into zones based on their
geographical location, which is determined using GPS or a similar positioning system.
2. Intra-Zone and Inter-Zone Routing:
o Intra-Zone Routing: Within a given zone, a proactive routing protocol is used,
which relies on exchanging link state packets to maintain up-to-date node-level
topologies. These updates are distributed using a shortest path algorithm within
the zone.
o Inter-Zone Routing: Between zones, a reactive routing protocol is employed. If
a node wants to communicate with a destination node outside its zone, it initiates
a location request that is forwarded to other zones. The gateway nodes of the
relevant zones provide the routing information.
3. Location Requests and Responses:
o When a source node needs to send a packet to a destination outside its zone, it
sends a location request packet. The request is propagated to other zones. If a
gateway node in another zone knows the location of the destination, it responds
with a location response packet, indicating the zone information of the
destination.
4. Gateway Nodes: Each zone has gateway nodes that serve as points of communication
between zones. These nodes help relay data and location information, ensuring routing
continuity even if some gateway nodes move away.
5. Route Maintenance: ZHLS is resilient to mobility, as multiple gateway nodes per zone
allow for continued routing even if one gateway node moves away. If a route between
two zones breaks due to mobility, the other gateways can continue the communication.

The zone-level topology is shown in Figure 7.28 (b). The zone link state packets originated by
every zone are shown inTable 7.1. Using the information obtained from zone link state packets, a
node can build the zone topology. The zone routing table can be formed for any destination zone
by executing the shortest path algorithm on the zone-level topology. The zone link state packets
are source sequence numbered and a time-stamp field is included to avoid stale link state
packets.
The association of the nodes to the respective zones helps in reducing routing overhead as in
ZRP, but it includes the additional requirement of determining a given destination node's present
location. If a source node Src wants to communicate with a destination node Dest, Src checks
whether Dest resides in its own zone. If Dest belongs to the same zone, then packets are
delivered to Dest as per the intra-zone routing table. If the destination Dest does not belong to the
zone, then the node Src originates a location request packet containing the sender's and
destination's information. This location request packet is forwarded to every other zone. The
gateway node of a zone at which the location request packet is received verifies its routing table
for the destination node for which the location request was originated. The gateway node that
finds the destination
node required by a location request packet originates a location response packet containing the
zone information to the sender.
Advantages:

 Reduced Storage and Communication Overhead: The hierarchical approach

minimizes the need for maintaining complete network-wide routing tables, which is
especially beneficial in large and mobile networks.
 Resilience to Mobility: The zone-based structure makes the network less vulnerable to
disruptions caused by node mobility.
 Efficient Intra-Zone Routing: Intra-zone routing is fast and does not involve network-
wide communication, reducing overhead.

Disadvantages:

 Creation of Zone-Level Topology: The process of creating and maintaining the zone-
level topology incurs additional overhead, especially in terms of geographical data and
periodic updates.
 Sub-optimal Path Selection: Routes are sometimes sub-optimal because the protocol
relies on zone-level topology, which may not always provide the most efficient routing
paths.
 Dependence on Geographical Information: The protocol requires that nodes have
accurate geographical information (e.g., via GPS), which may not be available in all
environments.