UNIT 3

ISSUES IN DESIGNING A ROUTING PROTOCOL

1. Mobility of Nodes

 Dynamic Network Topology: In ad hoc networks, the positions of nodes are

constantly changing due to the mobility of devices. As nodes move, the
network topology changes frequently, which can cause interruptions in ongoing
communications, such as path breaks.
 Unlike in wired networks, where nodes are stationary, ad hoc networks must
deal with frequent changes in connectivity.
 Requirement for Efficient Mobility Management: Routing protocols must
be able to detect and recover from path breaks quickly by finding new routes
to ensure uninterrupted communication despite mobility.

2. Bandwidth Constraints

 Limited Bandwidth: In contrast to wired networks, where fiber optics and

wavelength division multiplexing (WDM) provide abundant bandwidth, ad hoc
wireless networks have limited bandwidth due to the constraints of radio
spectrum availability. This limitation forces routing protocols to be bandwidth
efficient.
 Overhead of Topology Maintenance: Maintaining the network's topology in
a frequently changing environment consumes valuable bandwidth. If the
topology is shared globally, the control overhead increases, which in turn
wastes bandwidth.
 Optimization of Bandwidth: Routing protocols must minimize data overhead
to ensure efficient use of available bandwidth, while still maintaining the
integrity of routing information and communication links.

3. Error-Prone Shared Broadcast Radio Channel

 Shared Radio Channel: Wireless communication uses a shared medium,

which means multiple nodes can transmit simultaneously. However, this leads
to the risk of data collisions, especially in an ad hoc setting where nodes may
be unaware of each other's transmissions.
 Time-Varying Link Quality: The quality of wireless links is not static—
interference, signal attenuation, and fading can affect the link's reliability,
making the communication channel error-prone.

4.Hidden and Exposed Terminal Problem

 Hidden Terminal Problem: This problem arises when two nodes (A and C in
the figure) cannot detect each other's transmissions but are both within range
of a third node (B). This leads to packet collisions at node B, as neither node A
nor node C is aware of the other’s activity.
 Exposed Terminal Problem: Conversely, this problem occurs when a node
(C) cannot transmit because it assumes that its neighbour (B) is already
transmitting, even though its transmission would not interfere with B’s ongoing
communication. This reduces the efficient utilization of the available radio
spectrum.
 Solutions: Protocols like MACA, MACAW, FAMA, and DBTMA aim to reduce the
occurrence of these issues by implementing mechanisms such as Request-to-
Send (RTS) and Clear-to-Send (CTS) handshakes to control access to the shared
medium and reduce the chance of collision.
5. Resource Constraints

 Battery Life and Processing Power: Ad hoc network nodes are typically
portable devices with limited battery life and processing power. Increasing
these resources would make the devices heavier and less portable, which
contradicts the fundamental nature of ad hoc networks.

 Efficient Resource Management: Routing protocols must be designed to

minimize power consumption and processing load. This includes reducing the
frequency of control packet transmission, minimizing the complexity of route
computations, and ensuring efficient use of available resources.
CHARACTERISTICS OF AN IDEAL ROUTING PROTOCOL

 Fully Distributed: Centralized routing protocols are not feasible in ad hoc

networks due to high overhead and scalability issues. A distributed approach,
where each node independently makes routing decisions, is more fault-tolerant
and scalable.

 Adaptive to Topology Changes: Since nodes are constantly moving, the

routing protocol must quickly adapt to topology changes and reconfigure
routes to maintain network connectivity.

 Efficient Route Computation and Maintenance: The protocol should

minimize the number of nodes involved in route computation and
maintenance, reducing the control overhead and improving the speed of route
discovery.

 Localized: Rather than maintaining global state information, which is costly in

terms of control messages and bandwidth usage, routing protocols should
focus on localized information related to each node’s immediate neighbours
and network segment.

 Loop-Free and Stale-Free Routes: To ensure reliable communication,

routing protocols must avoid routing loops and stale (outdated) routes that
could lead to inefficiency or network failure.

 Minimize Packet Collisions: By minimizing broadcast messages and

ensuring reliable transmissions, the protocol should reduce packet collisions
that would waste bandwidth and battery power.

 Fast Convergence to Optimal Routes: Once the network topology

stabilizes, the protocol should quickly converge to the optimal route, ensuring
minimal delay and packet loss.

 Efficient Resource Use: The routing protocol must effectively utilize available
resources such as bandwidth, battery life, and processing power to avoid
overburdening any one resource.

 Local Topology Updates: Nodes should only update their routing information
based on local topology changes to reduce unnecessary overhead and
minimize the impact of distant topology changes.

 Quality of Service (QoS): The protocol should support applications with

specific quality-of-service requirements, such as real-time video or voice,
ensuring minimal delay, jitter, and packet loss.

CLASSIFICATION OF ROUTING PROTOCOLS

1. Based on Routing Information Update Mechanism

 Proactive (Table-Driven) Routing Protocols:

o Nodes maintain a routing table with up-to-date network topology
information by periodically exchanging routing information.
o When a node needs a path, it simply refers to its maintained table to
find it.
o Examples: Routing Information Protocol (RIP), Open Shortest Path First
(OSPF).
 Reactive (On-Demand) Routing Protocols:
o These protocols do not maintain network topology. Instead, they find
routes when needed, by initiating a route discovery process.
o No periodic updates are exchanged.
o Examples: Ad hoc On-Demand Distance Vector (AODV), Dynamic Source
Routing (DSR).
 Hybrid Routing Protocols:
o Combine features of proactive and reactive protocols.
o Nodes within a specific range (or region) use proactive methods, while
nodes outside this range use reactive methods.
o Example: Zone Routing Protocol (ZRP).

2. Based on Use of Temporal Information for Routing

 Using Past Temporal Information:

o These protocols base routing decisions on the past status of links, such
as link availability at the time of routing.
o Although the path may be stable initially, it could break due to changing
network conditions, requiring expensive reconfiguration.
 Using Future Temporal Information:
o These protocols use predictions about the future status of links and
nodes, such as battery life, link availability, or node mobility.
o The routing decision aims to avoid potential disruptions and optimize
path stability in the future.

3. Based on Routing Topology

 Flat Topology Routing Protocols:

o Use a flat addressing scheme (like in LANs), where every node is treated
equally and there is no hierarchy in addressing.
o Suitable for small networks with fewer nodes.
  Hierarchical Topology Routing Protocols:
o Use a logical hierarchy to organize the network. The addressing and
routing mechanisms are based on this hierarchy.
o Can reduce the overhead of maintaining routing tables across large
networks.

4. Based on Utilization of Specific Resources

 Power-Aware Routing Protocols:

o Focus on minimizing energy consumption by considering battery levels
in routing decisions, as nodes in ad hoc networks are often battery
powered.
o Examples include protocols that attempt to find energy-efficient paths
and reduce unnecessary transmissions to conserve power.
 Geographical Information-Assisted Routing:
o Use geographical information (e.g., location of nodes) to enhance
routing decisions.
o This can help reduce control overhead and improve routing efficiency by
directing data packets towards the destination using the shortest or
least congested path.

DESTINATION SEQUENCED DISTANCE-VECTOR ROUTING PROTOCOL (DSDV)

DSDV is an early routing protocol developed for ad hoc wireless networks. It's an
enhanced version of the distributed Bellman-Ford algorithm, where each node
maintains a routing table that contains:

 The shortest distance to every other node.

 The first node on the shortest path to each destination.

It uses sequence numbers to prevent routing loops and to ensure faster

convergence. Here's how it works:

1. Table Updates: Each node periodically exchanges routing tables with its
neighbours to maintain an up-to-date view of the network topology. If the
topology changes significantly, a full dump of the table is sent. Otherwise,
incremental updates are sent to keep the update process efficient.

2. Sequence Numbers: Each update includes a sequence number that is

incremented with every update. This sequence number is crucial for ensuring
the accuracy of the routing information. When a node receives an update with
a sequence number greater than the previous one, it updates its routing table
accordingly. The sequence number prevents routing loops and avoids the
count-to-infinity problem (a problem where routes continuously increment in
distance due to network instability).

3. Broken Link Propagation: If a link breaks (e.g., a node moves or becomes

unreachable), the node at the end of the broken link initiates an update with
the distance set to infinity (∞). This update is disseminated throughout the
network, informing all nodes that the link is no longer viable. The protocol then
recalculates the new path to the affected destination.

4. Handling Path Reconfiguration: If a path is no longer valid (due to node

movement or link failure), nodes update their tables with the new paths and
propagate these updates through the network. This process ensures that the
network adapts quickly to changes in topology.
Example of Route Calculation:

 Suppose Node 1 wants to send data to Node 15. Node 1 already has a route
to Node 15 stored in its table (as the topology is available globally at all times).
The table indicates that the shortest path to Node 15 is through Node 5, and
the distance to Node 15 is 4 hops.

 If a link fails (say between Node 10 and Node 11), Node 10 will set the path to
Node 11 as ∞ and propagate this information throughout the network. This
process ensures that all nodes, including Node 1, are aware of the broken link
and can update their routing tables accordingly.

Advantages of DSDV:

1. Low Latency for Route Setup: Since routes to all destinations are always
available at every node, there is little delay when setting up a route.
2. Wired Network Adaptation: The protocol is an enhancement of the Bellman-
Ford algorithm, making it easier to adapt wired network protocols for use in
wireless ad hoc networks.
3. Maintaining Up-to-date Network Topology: Periodic table updates ensure
that all nodes in the network have an accurate and current view of the network
topology.

Disadvantages of DSDV:

1. High Control Overhead: Frequent updates, generate significant control

traffic. The updates required when a link break can lead to excessive control
overhead. This overhead can consume bandwidth and reduce the overall
efficiency of the network.
2. Scalability Issues: The overhead increases with the number of nodes in the
network, making DSDV less scalable for large networks or networks with high
mobility.
3. Stale Routing Information: If a node has to wait for a routing table update
from a destination node, there might be a delay in obtaining the most current
routing information. This can result in using outdated or stale routing
information until the next update is received.
WIRELESS ROUTING PROTOCOL (WRP)

Key Differences from DSDV:

1. Table Maintenance: Unlike DSDV, which uses a single table for routing, WRP
employs multiple tables for more accurate routing information:

o Distance Table (DT): This table holds the distance and penultimate
node (the second-to-last node) reported by a neighbour for a particular
destination. It helps determine the shortest path and facilitates faster
convergence.
o Routing Table (RT): This table maintains the best routes, with entries
for the shortest distance, predecessor node (penultimate node),
successor node (next node on the path), and path status (correct, error,
or null).
o Link Cost Table (LCT): This table tracks the cost (such as the number
of hops) of relaying messages through each link. A broken link is marked
with a cost of infinity (∞). It also tracks the time since the last successful
update, which helps in detecting link failures.
o Message Retransmission List (MRL): This table keeps track of
messages that need to be retransmitted and their respective
retransmission counters. When the counter reaches zero, the message is
deleted if no acknowledgments are received, helping detect missing
updates or link failures.
 2. Faster Convergence: WRP speeds up convergence by checking not only the
distance of transmitted neighbours but also the distances of other neighbours.
This helps the network quickly identify alternative routes when a link break
occurs.

How WRP Works:

1. Route Maintenance: Like DSDV, WRP ensures that each node has routes to
all destinations, keeping them readily available. For instance, if Node 1 wants
to send data to Node 15, the routing table shows that the next hop is Node 2
and the predecessor node is Node 12.

2. Link Failure and Route Update: When a link breaks (e.g., between Nodes 12
and 15):

o The nodes detect the failure and set the link cost to ∞.
o They broadcast update messages to inform their neighbours of the
broken link.
o Affected nodes update their routing tables and find alternative paths if
possible, ensuring the data continues to flow through the network
without disruption.

3. Alternative Paths: If a node finds an alternative route (for example, through

Node 13 instead of Node 12), it broadcasts an update to its neighbours. The
neighbours update their tables only if the new path is better (shorter distance)
than the old path. This minimizes unnecessary updates and keeps the routing
information efficient.
 4. Update Propagation: After updating their routing tables, nodes propagate
these updates to their neighbours. Nodes that receive the updates adjust their
tables, accordingly, optimizing the paths to destination nodes.

Advantages of WRP:

1. Faster Convergence: WRP converges faster than DSDV due to its proactive
maintenance of multiple routing tables and the use of penultimate node
information, which helps identify alternative routes more quickly when a link
breaks.
2. Fewer Updates: The use of multiple tables allows for more accurate and
efficient routing, leading to fewer updates and minimizing the control
overhead compared to DSDV.
3. Accurate Routing Information: WRP provides more precise routing
information because it maintains multiple tables, including the distance table
(DT) and link cost table (LCT), to track changes and link failures in more
detail.

Disadvantages of WRP:

1. Complexity: WRP requires maintaining multiple tables (DT, RT, LCT, and
MRL), which increases the memory and processing power requirements of the
nodes. This could be challenging for devices with limited resources.
2. High Mobility Overhead: Although WRP has faster convergence, it still
suffers from high control overhead in highly dynamic environments (such as
when nodes are frequently moving), similar to DSDV. This can be problematic
in large, highly mobile networks.
3. Not Suitable for Large Networks: WRP's reliance on multiple table
updates and its complexity make it less suitable for large-scale networks with
a lot of mobility or nodes.