Module 2

Ad-hoc network routing & TCP issues

Ad Hoc Network Routing

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Introduction
• An ad hoc wireless network consists of a set of mobile nodes (hosts) that are connected by
wireless links. The network topology (the physical connectivity of the communication network) in
such a network may keep changing randomly. Routing protocols that find a path to be followed by
data packets from a source node to a destination node used in traditional wired networks cannot
be directly applied in ad hoc wireless networks due to their highly dynamic topology, absence of
established infrastructure for centralized administration (e.g., base stations or access points),
bandwidth-constrained wireless links, and resource (energy)-constrained nodes.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Classifications of routing protocols

Fig. 1

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

The routing protocols for ad hoc wireless networks can be broadly classified into four categories
based on
• Routing information update mechanism
• Use of temporal information for routing
• Routing topology
• Utilization of specific resources

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Based on the Routing Information Update
Mechanism
Ad hoc wireless network routing protocols can be classified into three major categories based on the
routing information update mechanism. They are:
• Proactive or table-driven routing protocols: In table-driven routing protocols, every node maintains
the network topology information in the form of routing tables by periodically exchanging routing
information. Routing information is generally flooded in the whole network. Whenever a node
requires a path to a destination, it runs an appropriate path-finding algorithm on the topology
information it maintains.
• Reactive or on-demand routing protocols: Protocols that fall under this category do not maintain
the network topology information. They obtain the necessary path when it is required, by using a
connection establishment process. Hence these protocols do not exchange routing information
periodically.
• Hybrid routing protocols: Protocols belonging to this category combine the best features of the
above two categories. Nodes within a certain distance from the node concerned, or within a
particular geographical region, are said to be within the routing zone of the given node. For routing
within this zone, a table-driven approach is used. For nodes that are located beyond this zone, an on-
demand approach is used.
Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi
Based on the Use of Temporal Information
for Routing
This classification of routing protocols is based on the use of temporal information used for routing.
Since ad hoc wireless networks are highly dynamic and path breaks are much more frequent than in
wired networks, the use of temporal information regarding the lifetime of the wireless links and the
lifetime of the paths selected assumes significance. The protocols that fall under this category can
be further classified into two types:
• Routing protocols using past temporal information: These routing protocols use information
about the past status of the links or the status of links at the time of routing to make routing
decisions. For example, the routing metric based on the availability of wireless links (which is the
current/present information here) along with a shortest path-finding algorithm, provides a path
that may be efficient and stable at the time of path-finding. The topological changes may
immediately break the path, making the path undergo a resource-wise expensive path
reconfiguration process.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Routing protocols that use future temporal information: Protocols belonging to this category use
information about the expected future status of the wireless links to make approximate routing
decisions. Apart from the lifetime of wireless links, the future status information also includes
information regarding the lifetime of the node (which is based on the remaining battery charge
and discharge rate of the non-replenishable resources), prediction of location, and prediction of
link availability.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Based on the Routing Topology
Routing topology being used in the Internet is hierarchical in order to reduce the state information
maintained at the core routers. Ad hoc wireless networks, due to their relatively smaller number of
nodes, can make use of either a flat topology or a hierarchical topology for routing.
• Flat topology routing protocols: Protocols that fall under this category make use of a flat
addressing scheme similar to the one used in IEEE 802.3 LANs. It assumes the presence of a
globally unique (or at least unique to the connected part of the network) addressing mechanism
for nodes in an ad hoc wireless network.
• Hierarchical topology routing protocols: Protocols belonging to this category make use of a
logical hierarchy in the network and an associated addressing scheme. The hierarchy could be
based on geographical information or it could be based on hop distance.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Based on the Utilization of Specific
Resources
• Power-aware routing: This category of routing protocols aims at minimizing the consumption of a
very important resource in the ad hoc wireless networks: the battery power. The routing decisions
are based on minimizing the power consumption either locally or globally in the network.
• Geographical information assisted routing: Protocols belonging to this category improve the
performance of routing and reduce the control overhead by effectively utilizing the geographical
information available.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Hierarchical Routing Protocols
The use of routing hierarchy has several advantages, the most important ones
being reduction in the size of routing tables and better scalability.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Hierarchical State Routing Protocol
• The hierarchical state routing (HSR) protocol is a distributed multi-level hierarchical routing
protocol that employs clustering at different levels with efficient membership management at
every level of clustering. The use of clustering enhances resource allocation and management. For
example, the allocation of different frequencies or spreading codes to different clusters can
improve the overall spectrum reuse. HSR operates by classifying different levels of clusters.
Elected leaders at every level form the members at the immediate higher level.
• Different clustering algorithms are employed for electing leaders at every level. The first level of
physical clustering is done among the nodes that are reachable in a single wireless hop. The next
higher level of physical clustering is done among the nodes that are elected as leaders of each of
these first-level clusters. In addition to the physical clustering, a logical clustering scheme has
been proposed in HSR, which is based on certain relations among the nodes rather than on their
geographical positions, as in the case of physical clustering.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 2

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Figure 2 illustrates the multilayer clustering defined by the HSR protocol. At the lowest level (L =
0), there are six cluster leaders (nodes 1, 2, 3, 4, 5, and 6). Nodes are classified as cluster leaders,
or gateway nodes, or normal member nodes. A cluster leader is entrusted with responsibilities
such as slot/frequency/code allocation, call admission control, scheduling of packet
transmissions, exchange of routing information, and handling route breaks.
• In Figure 2, node 5 is a clusterhead marked as L0-5, which refers to the level of clustering (L = 0)
and node ID (5). Similarly, each of the higher-level cluster leaders is also marked (e.g., L1 − 6, L2 −
6, and L3 − 6 refer to the same node 6, but acting as leader with the given leader IDs at levels 1, 2,
and 3, respectively).
• The spectrum reuse schemes, including spreading code assignment, can be used among the
cluster leaders of the L = 0 clusters. For the nodes under the leadership of node 6 at level 0, the
cluster members are nodes 9, 10, 11, 12, and 17. Those nodes that belong to multiple clusters are
referred to as cluster gateway nodes. For the level 0 cluster whose leader is node 6, the cluster
gateways are nodes 10, 12, and 17. The second level of clustering is done among the leaders of
the first level, that is, the leaders of 0th level clusters, L0 - 1, L0 - 2, L0 - 3,L0 - 4, L0 - 5, and L0 - 6,
form the members of the first-level cluster.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Every node maintains information about all its neighbors and the status of links to each of them.
This information is broadcast within the cluster at regular intervals. The cluster leader exchanges
the topology and link state routing information among its peers in the neighborhood clusters,
using which the next higher-level clustering is performed. This exchange of link state information
is done over multiple hops that consist of gateway nodes and cluster-heads. The path between
two cluster-heads which is formed by multiple wireless links is called a virtual link. The link status
for the virtual link(otherwise called tunnel) is obtained from the link status parameters of the
wireless links that constitute the virtual link.
• In Figure 2, the path between first-level clusters L1 – 6 and L1 - 4 includes the wireless links 6 - 12
- 5 - 16 - 4. The clustering is done recursively to the higher levels. At any level, the cluster leader
exchanges topology information with its peers. After obtaining information from its peers, it
floods the information to the lower levels, making every node obtain the hierarchical topology
information.
• This hierarchical topology necessitates a hierarchical addressing which helps in operating with less
routing information against the full topology exchange required in the link state routing. The
hierarchical addressing defined in HSR includes the hierarchical ID (HID) and node ID. The HID is a
sequence of IDs of cluster leaders of all levels starting from the highest level to the current node.
This ID of a node in HSR is similar to the unique MAC layer address. The hierarchical addresses are
stored in an HSR table at every node that indicates the node's own position in the hierarchy. The
HSR table is updated whenever routing update packets are received by the node.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• The hierarchical address of node 11 in Figure 7.31 is < 6, 6, 6, 6, 11 >, where the last entry (11) is
the node ID and the rest consists of the node IDs of the cluster leaders that represent the location
of node 11 in the hierarchy. Similarly, the HID of node 4 is < 6, 4, 4, 4 >. When node 11 needs to
send a packet to node 4, the packet is forwarded to the highest node in the hierarchy (node 6).
Node 6 delivers the packet to node 4, which is at the top-most level of the hierarchy.
Advantages and Disadvantages
• The HSR protocol reduces the routing table size by making use of hierarchy information. In HSR,
the storage required is O(n × m) compared to O(nm) that is required for a flat topology link state
routing protocol (n is the average number of nodes in a cluster and m is the number of levels).
Though the reduction in the amount of routing information stored at nodes is appreciable, the
overhead involved in exchanging packets containing information about the multiple levels of
hierarchy and the leader election process make the protocol unaffordable in the ad hoc wireless
networks context. Besides, the number of nodes that participate in an ad hoc wireless network
does not grow to the dimensions of the number of nodes in the Internet where the hierarchical
routing is better suited.
• In the military applications of ad hoc wireless networks, the hierarchy of routing assumes
significance where devices with higher capabilities of communication can act as the cluster
leaders.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Fisheye State Routing Protocol
• The table-driven routing protocols generate routing overhead that is dependent on the size of the
network and mobility of the nodes, whereas the routing overhead generated by on-demand
routing protocols are dependent on the number of connections present in the system in addition
to the above two factors. Hence, as the number of senders in the network increases, the routing
overhead also increases. ZRP uses an intra-zone proactive approach and an inter-zone reactive
approach to reduce control overhead.
• The fisheye state routing (FSR) protocol is a generalization of the GSR protocol. FSR uses the
fisheye technique to reduce information required to represent graphical data, to reduce routing
overhead.
• The basic principle behind this technique is the property of a fish's eye that can capture pixel
information with greater accuracy near its eye's focal point. This accuracy decreases with an
increase in the distance from the center of the focal point. This property is translated to routing in
ad hoc wireless networks by a node, keeping accurate information about nodes in its local
topology, and not-so-accurate information about far-away nodes, the accuracy of the network
information decreasing with increasing distance.
Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi
• FSR maintains the topology of the network at every node, but does not flood the entire network
with the information, as is done in link state routing protocols. Instead of flooding, a node
exchanges topology information only with its neighbors. A sequence numbering scheme is used
to identify the recent topology changes. This constitutes a hybrid approach comprising of the link-
level information exchange of distance vector protocols and the complete topology information
exchange of link state protocols.
• The complete topology information of the network is maintained at every node and the desired
shortest paths are computed as required. The topology information exchange takes place
periodically rather than being driven by an event. This is because instability of the wireless links
may cause excessive control overhead when event-driven updates are employed. FSR defines
routing scope, which is the set of nodes that are reachable in a specific number of hops. The
scope of a node at two hops is the set of nodes that can be reached in two hops.
• Figure 3 shows the scope of node 5 with one hop and two hops. The routing overhead is
significantly reduced by adopting different frequencies of updates for nodes belonging to
different scopes.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 3

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• The link state information for the nodes belonging to the smallest scope is exchanged at the
highest frequency. The frequency of exchanges decreases with an increase in scope. This keeps
the immediate neighbourhood topology information maintained at a node more precise
compared to the information about nodes farther away from it.
• Thus the message size for a typical topology information update packet is significantly reduced
due to the removal of topology information regarding the far-away nodes. The path information
for a distant node may be inaccurate as there can be staleness in the information. But this is
compensated by the fact that the route gets more and more accurate as the packet nears its
destination. FSR scales well for large ad hoc wireless networks because of the reduction in routing
overhead due to the use of the above described mechanism, where varying frequencies of
updates are used.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Figure 4 illustrates an example depicting the network topology information maintained at nodes
in a network. The routing information for the nodes that are one hop away from a node are
exchanged more frequently than the routing information about nodes that are more than one
hop away. Information regarding nodes that are more than one hop away from the current node
are listed below the dotted line in the topology table.

Fig. 4
Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi
Advantages and Disadvantages
• The notion of multi-level scopes employed by FSR significantly reduces the bandwidth consumed
by link state update packets. Hence, FSR is suitable for large and highly mobile ad hoc wireless
networks. The choice of the number of hops associated with each scope level has a significant
influence on the performance of the protocol at different mobility values, and hence must be
carefully chosen.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Power-aware Routing Protocols
In a deviation from the traditional wired network routing and cellular wireless network
routing, power consumption by the nodes is a serious factor to be taken into consideration
by routing protocols for ad hoc wireless networks. This is because, in ad hoc wireless
networks, the routers are also equally power constrained just as the nodes are.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Power Aware Routing Metrics
• The limitation on the availability of power for operation is a significant bottleneck, given the requirements of
portability, weight, and size of commercial hand-held devices. Hence, the use of routing metrics that
consider the capabilities of the power sources of the network nodes contributes to the efficient utilization of
energy and increases the lifetime of the network. Singh et al. proposed a set of routing metrics that supports
conservation of battery power. The routing protocols that select paths so as to conserve power must be
aware of the states of the batteries at the given node as well as at the other intermediate nodes in the path.

Minimal Energy Consumption per Packet

• This metric aims at minimizing the power consumed by a packet in traversing from source node to the
destination node. The energy consumed by a packet when traversing through a path is the sum of the
energies required at every intermediate hop in that path. The energy consumed at an intermediate hop is a
function of the distance between the nodes that form the link and the load on that link. This metric does not
balance the load so that uniform consumption of power is maintained throughout the network.
• The disadvantages of this metric include selection of paths with large hop length, inability to measure the
power consumption at a link in advance when the load varies, and the inability to prevent the fast
discharging of batteries at some nodes.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Maximize Network Connectivity
• This metric attempts to balance the routing load among the cut-set (the subset of the nodes in the
network, the removal of which results in network partitions). This assumes significance in
environments where network connectivity is to be ensured by uniformly distributing the routing load
among the cut-set. With a variable traffic origination rate and unbounded contention in the network, it
is difficult to achieve a uniform battery draining rate for the cut-set.
Minimum Variance in Node Power Levels
• This metric proposes to distribute the load among all nodes in the network so that the power
consumption pattern remains uniform across them. This problem is very complex when the rate and
size of data packets vary. A nearly optimal performance can be achieved by routing packets to the
least-loaded next-hop node.
Minimum Cost per Packet
• In order to maximize the life of every node in the network, this routing metric is made as a function of
the state of the node's battery. A node's cost decreases with an increase in its battery charge and vice
versa. Translation of the remaining battery charge to a cost factor is used for routing. With the
availability of a battery discharge pattern, the cost of a node can be computed. This metric has the
advantage of ease in the calculation of the cost of a node and at the same time congestion handling is
done.
Minimize Maximum Node Cost
• This metric minimizes the maximum cost per node for a packet after routing a number of packets or
after a specific period. This delays the failure of a node, occurring due to higher discharge because of
packet forwarding. Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi
Multicast Routing

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Introduction
• Multicast routing protocols play an important role in ad hoc wireless networks to provide this
communication. It is always advantageous to use multicast rather than multiple unicast, especially
in the ad hoc environment, where bandwidth comes at a premium.
• Conventional wired network Internet protocol (IP) multicast routing protocols, such as DVMRP,
MOSPF, CBT, and PIM, do not perform well in ad hoc networks because of the dynamic nature of
the network topology. The dynamically changing topology, coupled with relatively low bandwidth
and less reliable wireless links, causes long convergence times and may give rise to formation of
transient routing loops which rapidly consume the already limited bandwidth.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• In a wired network, the basic approach adopted for multicasting consists of establishing a routing
tree for a group of routing nodes that constitute the multicast session. Once the routing tree (or
the spanning tree, which is an acyclic connected subgraph containing all the nodes in the tree) is
established, a packet sent to all nodes in the tree traverses each node and each link in the tree
only once. Such a multicast structure is not appropriate for ad hoc networks because the tree
could easily break due to the highly dynamic topology.
• Multicast tree structures are not stable and need to be reconstructed continuously as
connectivity changes. Maintaining a routing tree for the purpose of multicasting packets, when
the underlying topology keeps changing frequently, can incur substantial control traffic. The
multicast trees used in the conventional wired network multicast protocols require a global
routing substructure such as a link state or a distance vector sub-structure.
• The frequent exchange of routing vectors or link state tables, triggered by continuous topology
changes, yields excessive control and processing overhead. Further, periods of routing table
instability lead to instability of the multicast tree, which in turn results in increased buffering time
for packets, higher packet losses, and an increase in the number of retransmissions. Therefore,
multicast protocols used in static wired networks are not suitable for ad hoc wireless networks.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Classifications of Multicast Routing
Protocols

Fig. 5

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Multicast routing protocols for ad hoc wireless networks can be broadly classified into two types:
application-independent/generic multicast protocols and application-dependent multicast
protocols.
• While application-independent multicast protocols are used for conventional multicasting,
application-dependent multicast protocols are meant only for specific applications for which they
are designed. Application-independent multicast protocols can be classified along three different
dimensions.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Based on topology: Current approaches used for ad hoc multicast routing protocols can be
classified into two types based on the multicast topology: tree-based and mesh-based.
In tree-based multicast routing protocols, there exists only a single path between a source-receiver
pair, whereas in mesh-based multicast routing protocols, there may be more than one path
between a source-receiver pair. Tree-based multicast protocols are more efficient compared to
mesh-based multicast protocols, but mesh-based multicast protocols are robust due to the
availability of multiple paths between the source and receiver.
Tree-based multicast protocols can be further divided into two types: source-tree based and shared-
tree-based. In source-tree-based multicast protocols, the tree is rooted at the source, whereas in
shared-tree-based multicast protocols, a single tree is shared by all the sources within the multicast
group and is rooted at a node referred to as the core node.
The source tree-based multicast protocols perform better than the shared-tree-based protocols at
heavy loads because of efficient traffic distribution. But the latter type of protocols are more
scalable. The main problem in a shared tree-based multicast protocol is that it heavily depends on
the core node, and hence, a single point failure at the core node affects the performance of the
multicast protocol.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Based on initialization of the multicast session: The multicast group formation can be initiated by
the source as well as by the receivers. In a multicast protocol, if the group formation is initiated
only by the source node, then it is called a source-initiated multicast routing protocol, and if it is
initiated by the receivers of the multicast group, then it is called a receiver-initiated multicast
routing protocol. Some multicast protocols do not distinguish between source and receiver for
initialization of the multicast group. We call these source-or-receiver-initiated multicast routing
protocols.
• Based on the topology maintenance mechanism: Maintenance of the multicast topology can be
done either by the soft state approach or by the hard state approach. In the soft state approach,
control packets are flooded periodically to refresh the route, which leads to a high packet delivery
ratio at the cost of more control overhead, whereas in the hard state approach, the control
packets are transmitted (to maintain the routes) only when a link breaks, resulting in lower
control overhead, but at the cost of a low packet delivery ratio.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Tree based Multicast routing
protocols

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Introduction
• Tree-based multicasting is a well-established concept used in several wired multicast protocols to
achieve high multicast efficiency. In tree-based multicast protocols, there is only one path
between a source-receiver pair. The main drawback of these protocols is that they are not robust
enough to operate in highly mobile environments. Tree-based multicast protocols can be
classified into two types: source-tree-based multicast routing protocols and shared-tree based
multicast routing protocols.
• In a source-tree-based protocol, a single multicast tree is maintained per source, whereas in a
shared tree-based protocol, a single tree is shared by all the sources in the multicast group.
Shared-tree-based multicast protocols are more scalable compared to source-tree-based
multicast protocols. By scalability, we mean the ability of the protocol to work well without any
degradation in performance when the number of sources in a multicast session or the number of
multicast sessions is increased.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• In source-tree-based multicast routing protocols, an increase in the number of sources gives rise
to a proportional increase in the number of source trees.
• This results in a significant increase in bandwidth consumption in the already-bandwidth-
constrained network. But in a shared-tree-based multicast protocol, this increase in bandwidth
usage is not as high as in source-tree-based protocols because, even when the number of sources
for multicast sessions increases, the number of trees remains the same.
• Another factor that affects the scalability of source-tree-based protocols is the memory
requirement. When the multicast group size is large with a large number of multicast sources, in a
source-tree-based multicast protocol, the state information that is maintained per source per
group consumes a large amount of memory at the nodes. But in a shared-tree-based multicast
protocol, since the state information is maintained per group, the additional memory required
when the number of sources increases is not very high. Hence shared-tree-based multicast
protocols are more scalable compared to source-tree-based multicast protocols.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Bandwidth Efficient Multicast Routing
Protocol
• Ad hoc networks operate in a highly bandwidth-scarce environment, and hence bandwidth
efficiency is one of the key design criteria for multicast protocols. Bandwidth efficient multicast
routing protocol (BEMRP) tries to find the nearest forwarding node, rather than the shortest path
between source and receiver. Hence, it reduces the number of data packet transmissions. To
maintain the multicast tree, it uses the hard state approach, that is, to rejoin the multicast group,
a node transmits the required control packets only after the link breaks. Thus, it avoids periodic
transmission of control packets and hence bandwidth is saved. To remove unwanted forwarding
nodes, route optimization is done, which helps in further reducing the number of data packet
transmissions. The multicast tree initialization phase and the tree maintenance phase are
discussed in the following sections.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Tree Initialization Phase
• In BEMRP, the multicast tree construction is initiated by the receivers. When a receiver wants to
join the group, it initiates flooding of Join control packets. The existing members of the multicast
tree, on receiving these packets, respond with Reply packets. When many such Reply packets
reach the requesting node, it chooses one of them and sends a Reserve packet on the path taken
by the chosen Reply packet. When a new receiver R3 (Fig. 6) wants to join the multicast group, it
floods the Join control packet. The nodes S, I1, and R2 of the multicast tree may receive more
than one Join control packet. After waiting for a specific time, each of these tree nodes chooses
one Join packet with the smallest hop count traversed. It sends back a Reply packet along the
reverse path which the selected Join packet had traversed. When tree node I1 receives Join
packets from the previous nodes I9 and I2, it sends a Reply packet to receiver R3 through node I2.
The receiver may receive more than one Reply packet. In this case, it selects the Reply packet
which has the lowest hop count, and sends a Reserve packet along the reverse path that the
selected Reply packet had traversed. Here, in Figure 6, receiver R3 receives Reply packets from
source S, receiver R2, and intermediate node I1. Since the Reply packet sent by intermediate node
I1 has the lowest hop count (which is 3), it sends a Reserve packet to node I3, and thus joins the
multicast group.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 6

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Tree Maintenance Phase
• To reduce the control overhead, in BEMRP, tree reconfiguration is done only when a link break is
detected. There are two schemes to recover from link failures.
• 1. Broadcast-multicast scheme: In this scheme, the upstream node is responsible for finding a
new route to the previous downstream node. This is shown in Figure 7. When receiver R3 moves
from A to B, it gets isolated from the remaining part of the tree. The upstream node I3 now floods
broadcast-multicast packets (with limited TTL). After receiving this packet, receiver R3 sends a
Reserve packet and joins the group again.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 7

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• 2. Local rejoin scheme: In this scheme, the downstream node of the broken link tries to rejoin the
multicast group by means of limited flooding of the Join packets. In Figure 8, when the link
between receiver R3 and its upstream node I3 fails (due to movement of node R3), then R3 floods
the Join control packet with a certain TTL value (depending on the topology, this value can be
tuned). When tree nodes receive the Join control packet, they send back the Reply packet. After
receiving the Reply packet, the downstream node R3 rejoins the group by sending a Reserve
packet to the new upstream node I4.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 8

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Route Optimization Phase
• When a tree node or a receiver node comes within the transmission range of other tree nodes,
then unwanted tree nodes are pruned by sending the Quit message. In Figure 9, when receiver R3
comes within the transmission range of the intermediate node I2, it will receive a multicast packet
from node I2 earlier than from node I5. When node R3 receives a multicast packet from node I2,
it sends a Reserve packet to node I2 to set up a new route directly to node I2, and sends a Quit
packet to node I5. Since node R3 is no more its downstream node, node I5 sends a Quit packet to
node I4, node I4 sends a Quit packet to node I3, and node I3 in turn sends a Quit packet to node
I2. Thus unnecessary forwarding nodes are pruned. This mechanism helps to reduce the number
of data packet transmissions.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 9

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Advantages and Disadvantages
• The main advantage of this multicast protocol is that it saves bandwidth due to the reduction in
the number of data packet transmissions and the hard state approach being adopted for tree
maintenance. Since a node joins the multicast group through its nearest forwarding node, the
distance between source and receiver increases. This increase in distance increases the
probability of path breaks, which in turn gives rise to an increase in delay and reduction in the
packet delivery ratio. Also, since the protocol uses the hard state approach for route repair, a
considerable amount of time is spent by the node in reconnecting to the multicast session, which
adds to the delay in packet delivery.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Mesh based multicast routing
protocols
In ad hoc wireless networks, wireless links break due to the mobility of the nodes. In the case of
multicast routing protocols, the path between a source and receiver, which consists of multiple
wireless hops, suffers very much due to link breaks. Multicast routing protocols which provide
multiple paths between a source-receiver pair are classified as mesh-based multicast routing
protocols. The presence of multiple paths adds to the robustness of the mesh-based protocols at
the cost of multicast efficiency.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

On Demand Multicast Routing Protocol
• In the on-demand multicast routing protocol (ODMRP), a mesh is formed by a set of nodes called
forwarding nodes which are responsible for forwarding data packets between a source-receiver
pair. These forwarding nodes maintain the message-cache which is used to detect duplicate data
packets and duplicate JoinReq control packets.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Mesh Initialization Phase
• In the mesh initialization phase, a multicast mesh is formed between the sources and the
receivers. To create the mesh, each source in the multicast group floods the JoinReq control
packet periodically. Upon reception of the JoinReq control packet from a source, potential
receivers can send JoinReply through the reverse shortest path. The route between a source and
receiver is established after the source receives the JoinReply packet. This is illustrated in Figure
10. For initializing the mesh, sources S1 and S2 in the multicast group flood the JoinReq control
packets. The nodes that receive a JoinReq control packet store the upstream node identification
number (ID) and broadcast the packet again. When receivers R1, R2, and R3 receive the JoinReq
control packet, each node sends a JoinReply control packet along the reverse path to the source.
Here in Figure 10, receiver R2 receives JoinReq control packets from sources S1 and S2 through
paths S1-I2-I3-R2 and S2-I6-I4-I5-R2, respectively.
• The JoinReply packet contains the source ID and the corresponding next node ID (the upstream
node through which it received the JoinReq packet). When node I2 receives the JoinReply control
packet from receiver R1, it sets a forwarding flag and becomes the forwarding node for that
particular multicast group. After waiting for a specified time, it composes a new JoinReply packet
and forwards it. The format of the JoinReply packet sent by the node R2 is shown in Table 1. In
this way, subsequent forwarding of JoinReply packets by the intermediate nodes along the
reverse path to the source establishes the route.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Table-1

Fig.10

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Mesh Maintenance Phase
• In this phase, attempts are made to maintain the multicast mesh topology formed with sources,
forwarding nodes, and receivers. To some extent, the multicast mesh protects the session from
being affected by mobility of nodes. For example, due to movement of the receiver R3 (from A to
B), when the route S2-I9-I10-R3 breaks (Figure 11), R3 can still receive data packets through route
S2-I6-I4-I7-I8-R3 and this contributes to the high packet delivery ratio. ODMRP uses a soft state
approach to maintain the mesh, that is, to refresh the routes between the source and the
receiver, the source periodically floods the JoinReq control packet. In Figure 11, when receiver R3
receives a new JoinReq control packet from node I11 (sent by the source S2), it sends a JoinReply
on this new shortest path R3-I11-I10-I9-S2, thereby maintaining the mesh structure.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 11

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Advantages and Disadvantages
• Since ODMRP uses the soft state approach for maintaining the mesh, it exhibits robustness. But
this robustness is at the expense of high control overhead.
• Another disadvantage is that the same data packet (from source S2 to receiver R3) propagates
through more than one path to a destination node, resulting in an increased number of data
packet transmissions, thereby reducing the multicast efficiency.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

TCP in Ad Hoc Networks

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Introduction
• The objectives of a transport layer protocol include the setting up of an end-to end connection,
end-to-end delivery of data packets, flow control, and congestion control. There exist simple,
unreliable, and connection-less transport layer protocols such as UDP, and reliable, byte-stream-
based, and connection oriented transport layer protocols such as TCP for wired networks. These
traditional wired transport layer protocols are not suitable for ad hoc wireless networks due to
the inherent problems associated with the latter.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Issues in designing a transport layer protocol
for ad hoc networks
• Induced traffic: Unlike wired networks, ad hoc wireless networks utilize multi-hop radio relaying.
A link-level transmission affects the neighbor nodes of both the sender and receiver of the link. In
a path having multiple links, transmission at a particular link affects one upstream link and one
downstream link. This traffic at any given link (or path) due to the traffic through neighboring
links (or paths) is referred to as induced traffic. This is due to the broadcast nature of the channel
and the location-dependent contention on the channel. This induced traffic affects the
throughput achieved by the transport layer protocol.
• Induced throughput unfairness: This refers to the throughput unfairness at the transport layer
due to the throughput/delay unfairness existing at the lower layers such as the network and MAC
layers. For example, an ad hoc wireless network that uses IEEE 802.11 DCF as the MAC protocol
may experience throughput unfairness at the transport layer as well. A transport layer protocol
should consider these in order to provide a fair share of throughput across contending flows.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Separation of congestion control, reliability, and flow control: A transport layer protocol can
provide better performance if end-to-end reliability, flow control, and congestion control are
handled separately. Reliability and flow control are end-to-end activities, whereas congestion can
at times be a local activity. The transport layer flow can experience congestion with just one
intermediate link under congestion. Hence, in networks such as ad hoc wireless networks, the
performance of the transport layer may be improved if these are separately handled. While
separating these, the most important objective to be considered is the minimization of the
additional control overhead generated by them.
• Power and bandwidth constraints: Nodes in ad hoc wireless networks face resource constraints
including the two most important resources: (i) power source and (ii) bandwidth. The
performance of a transport layer protocol is significantly affected by these constraints.
• Misinterpretation of congestion: Traditional mechanisms of detecting congestion in networks,
such as packet loss and retransmission timeout, are not suitable for detecting the network
congestion in ad hoc wireless networks. This is because the high error rates of wireless channel,
location-dependent contention, hidden terminal problem, packet collisions in the network, path
breaks due to the mobility of nodes, and node failure due to a drained battery can also lead to
packet loss in ad hoc wireless networks. Hence, interpretation of network congestion as used in
traditional networks is not appropriate in ad hoc wireless networks.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Completely decoupled transport layer: Another challenge faced by a transport layer protocol is
the interaction with the lower layers. Wired network transport layer protocols are almost
completely decoupled from the lower layers. In ad hoc wireless networks, the cross-layer
interaction between the transport layer and lower layers such as the network layer and the MAC
layer is important for the transport layer to adapt to the changing network environment.
• Dynamic topology: Some of the deployment scenarios of ad hoc wireless networks experience
rapidly changing network topology due to the mobility of nodes. This can lead to frequent path
breaks, partitioning and remerging of networks, and high delay in reestablishment of paths.
Hence, the performance of a transport layer protocol is significantly affected by the rapid changes
in the network topology.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Design goals of a transport layer protocol for
ad hoc wireless networks
• The protocol should maximize the throughput per connection.
• It should provide throughput fairness across contending flows.
• The protocol should incur minimum connection setup and connection maintenance overheads. It
should minimize the resource requirements for setting up and maintaining the connection in
order to make the protocol scalable in large networks.
• The transport layer protocol should have mechanisms for congestion control and flow control in
the network.
• It should be able to provide both reliable and unreliable connections as per the requirements of
the application layer.
• The protocol should be able to adapt to the dynamics of the network such as the rapid change in
topology and changes in the nature of wireless links from uni-directional to bidirectional or vice
versa.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• One of the important resources, the available bandwidth, must be used efficiently.
• The protocol should be aware of resource constraints such as battery power and buffer sizes and
make efficient use of them.
• The transport layer protocol should make use of information from the lower layers in the protocol
stack for improving the network throughput.
• It should have a well-defined cross-layer interaction framework for effective, scalable, and
protocol-independent interaction with lower layers.
• The protocol should maintain end-to-end semantics.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Classification of Transport Layer Issues

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Feedback-based TCP
• Feedback-based TCP [also referred to as TCP feedback (TCP-F)] proposes modifications to the
traditional TCP for improving performance in ad hoc wireless networks. It uses a feedback-based
approach. TCP-F requires the support of a reliable link layer and a routing protocol that can
provide feedback to the TCP sender about the path breaks. The routing protocol is expected to
repair the broken path within a reasonable time period.
• TCP-F aims to minimize the throughput degradation resulting from the frequent path breaks that
occur in ad hoc wireless networks. During a TCP session, there could be several path breaks
resulting in considerable packet loss and path reestablishment delay. Upon detection of packet
loss, the sender in a TCP session invokes the congestion control algorithm leading to the
exponential back-off of retransmission timers and a decrease in congestion window size.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• In TCP-F, an intermediate node, upon detection of a path break, originates a route failure
notification (RFN) packet. This RFN packet is routed toward the sender of the TCP session. The
TCP sender's information is expected to be obtained from the TCP packets being forwarded by the
node. The intermediate node that originates the RFN packet is called the failure point (FP).
• The FP maintains information about all the RFNs it has originated so far. Every intermediate node
that forwards the RFN packet understands the route failure, updates its routing table accordingly,
and avoids forwarding any more packets on that route. If any of the intermediate nodes that
receive RFN has an alternate route to the same destination, then it discards the RFN packet and
uses the alternate path for forwarding further data packets, thus reducing the control overhead
involved in the route reconfiguration process.
• Otherwise, it forwards the RFN toward the source node. When a TCP sender receives an RFN
packet, it goes into a state called snooze. In the snooze state, a sender stops sending any more
packets to the destination, cancels all the timers, freezes its congestion window, freezes the
retransmission timer, and sets up a route failure timer. This route failure timer is dependent on
the routing protocol, network size, and the network dynamics and is to be taken as the worst-case
route reconfiguration time. When the route failure timer expires, the TCP sender changes from
the snooze state to the connected state.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Figure 12 shows the operation of the TCP-F protocol. In the figure, a TCP session is set up
between node A and node D over the path A-B-C-D [refer to Figure 12(a)]. When the intermediate
link between node C and node D fails, node C originates an RFN packet and forwards it on the
reverse path to the source node [see Figure 12 (b)]. The sender's TCP state is changed to the
snooze state upon receipt of an RFN packet. If the link CD rejoins, or if any of the intermediate
nodes obtains a path to destination node D, a route reestablishment notification (RRN) packet is
sent to node A and the TCP state is updated back to the connected state [Figure 12 (c)].

Fig. 12

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• As soon as a node receives an RRN packet, it transmits all the packets in its buffer, assuming that
the network is back to its original state. This can also take care of all the packets that were not
acknowledged or lost during transit due to the path break. In fact, such a step avoids going
through the slow-start process that would otherwise have occurred immediately after a period of
congestion.
• The route failure timer set after receiving the RFN packet ensures that the sender does not
remain in the snooze state indefinitely. Once the route failure timer expires, the sender goes back
to the connected state in which it reactivates the frozen timers and starts sending the buffered
and unacknowledged packets.
• This can also take care of the loss of the RRN packet due to any possible subsequent congestion.
TCP-F permits the TCP congestion control algorithm to be in effect when the sender is not in the
snooze state, thus making it sensitive to congestion in the network.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Advantages and Disadvantages
• TCP-F provides a simple feedback-based solution to minimize the problems arising out of frequent
path breaks in ad hoc wireless networks. At the same time, it also permits the TCP congestion
control mechanism to respond to congestion in the network. TCP-F depends on the intermediate
nodes' ability to detect route failures and the routing protocols' capability to re-establish a broken
path within a reasonably short duration. Also, the FP should be able to obtain the correct path
(the path which the packet traversed) to the TCP-F sender for sending the RFN packet. This is
simple with a routing protocol that uses source routing [i.e., dynamic source routing (DSR)]. If a
route to the sender is not available at the FP, then additional control packets may need to be
generated for routing the RFN packet.
• TCP-F has an additional state compared to the traditional TCP state machine, and hence its
implementation requires modifications to the existing TCP libraries. Another disadvantage of TCP-
F is that the congestion window used after a new route is obtained may not reflect the achievable
transmission rate acceptable to the network and the TCP-F receiver.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

TCP with Explicit Link Failure Notification
• Holland and Vaidya proposed the use of TCP with explicit link failure notification (TCP-ELFN) [8]
for improving TCP performance in ad hoc wireless networks. This is similar to TCP-F, except for the
handling of explicit link failure notification (ELFN) and the use of TCP probe packets for detecting
the route re-establishment. The ELFN is originated by the node detecting a path break upon
detection of a link failure to the TCP sender. This can be implemented in two ways: (i) by sending
an ICMP destination unreachable (DUR) message to the sender, or (ii) by piggy-backing this
information on the RouteError message that is sent to the sender.
ICMP - Internet control message protocol (IETF RFC 792) is used for defining control messages for
aiding routing in the Internet.
RouteError - Certain routing protocols for ad hoc wireless networks have explicit RouteError
messages to inform the sender about path breaks so that the sender can recompute a fresh route to
the destination. This is especially used in on-demand routing protocols such as DSR.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Once the TCP sender receives the ELFN packet, it disables its retransmission timers and enters a
standby state. In this state, it periodically originates probe packets to see if a new route is re-
established. Upon reception of an ACK by the TCP receiver for the probe packets, it leaves the
standby state, restores the retransmission timers, and continues to function as normal.
 Advantages and Disadvantages
• TCP-ELFN improves the TCP performance by decoupling the path break information from the
congestion information by the use of ELFN. It is less dependent on the routing protocol and
requires only link failure notification about the path break.
• The disadvantages of TCP-ELFN include the following:
(i) when the network is temporarily partitioned, the path failure may last longer and this can lead
to the origination of periodic probe packets consuming bandwidth and power and
(ii) the congestion window used after a new route is obtained may not reflect the achievable
transmission rate acceptable to the network and the TCP receiver.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

TCP Bus
• TCP with buffering capability and sequence information (TCP-BuS) is similar to the TCP-F and TCP-ELFN in its
use of feedback information from an intermediate node on detection of a path break. But TCP-BuS is more
dependent on the routing protocol compared to TCP-F and TCP-ELFN. TCP-BuS was proposed, with
associativity-based routing (ABR) protocol as the routing scheme. Hence, it makes use of some of the special
messages such as localized query (LQ) and REPLY, defined as part of ABR for finding a partial path. These
messages are modified to carry TCP connection and segment information. Upon detection of a path break,
an upstream intermediate node [called pivot node (PN)] originates an explicit route dis-connection
notification (ERDN) message.
• This ERDN packet is propagated to the TCP-BuS sender and, upon reception of it, the TCP-BuS sender stops
transmission and freezes all timers and windows as in TCP-F. The packets in transit at the intermediate nodes
from the TCP-BuS sender to the PN are buffered until a new partial path from the PN to the TCP-BuS receiver
is obtained by the PN.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• In order to avoid unnecessary retransmissions, the timers for the buffered packets at the TCP-BuS
sender and at the intermediate nodes up to PN use timeout values proportional to the roundtrip
time (RTT). The intermediate nodes between the TCP-BuS sender and the PN can request the TCP-
BuS sender to selectively retransmit any of the lost packets.
• Upon detection of a path break, the downstream node originates a route notification (RN) packet
to the TCP-BuS receiver, which is forwarded by all the downstream nodes in the path. An
intermediate node that receives an RN packet discards all packets belonging to that flow. The
ERDN packet is propagated to the TCP-BuS sender in a reliable way by using an implicit
acknowledgment and retransmission mechanism. The PN includes the sequence number of the
TCP segment belonging to the flow that is currently at the head of its queue in the ERDN packet.
The PN also attempts to find a new partial route to the TCP-BuS receiver, and the availability of
such a partial path to destination is intimated to the TCP-BuS sender through an explicit route
successful notification (ERSN) packet.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• TCP-BuS utilizes the route reconfiguration mechanism of ABR to obtain the partial route to the
destination. Due to this, other routing protocols may require changes to support TCP-BuS. The LQ
and REPLY messages are modified to carry TCP segment information, including the last
successfully received segment at the destination. The LQ packet carries the sequence number of
the segment at the head of the queue buffered at the PN and the REPLY carries the sequence
number of the last successful segment the TCP-BuS receiver received. This enables the TCP-BuS
receiver to understand the packets lost in transition and those buffered at the intermediate
nodes. This is used to avoid fast retransmission requests usually generated by the TCP-BuS
receiver when it notices an out-of-order packet delivery.
• Upon a successful LQREPLY process to obtain a new route to the TCP-BuS receiver, PN informs the
TCP-BuS sender of the new partial path using the ERSN packet. When the TCP-BuS sender
receives an ERSN packet, it resumes the data transmission.
• Since there is a chance for ERSN packet loss due to congestion in the network, it needs to be sent
reliably. The TCP-BuS sender also periodically originates probe packets to check the availability of
a path to the destination. Figure 13 shows an illustration of the propagation of ERDN and RN
messages when a link between nodes 4 and 12 fails.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Fig. 13

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• When a TCP-BuS sender receives the ERSN message, it understands, from the sequence number
of the last successfully received packet at the destination and the sequence number of the packet
at the head of the queue at PN, the packets lost in transition. The TCP-BuS receiver understands
that the lost packets will be delayed further and hence uses a selective acknowledgment strategy
instead of fast retransmission. These lost packets are retransmitted by the TCP-BuS sender. During
the retransmission of these lost packets, the network congestion between the TCP-BuS sender
and PN is handled in a way similar to that in traditional TCP.
 Advantages and Disadvantages
• The advantages of TCP-BuS include performance improvement and avoidance of fast
retransmission due to the use of buffering, sequence numbering, and selective acknowledgment.
TCP-BuS also takes advantage of the underlying routing protocols, especially the on-demand
routing protocols such as ABR. The disadvantages of TCP-BuS include the increased dependency
on the routing protocol and the buffering at the intermediate nodes. The failure of intermediate
nodes that buffer the packets may lead to loss of packets and performance degradation. The
dependency of TCP-BuS on the routing protocol may degrade its performance with other routing
protocols that do not have similar control messages as in ABR.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Ad Hoc TCP
• Similar to TCP-F and TCP-ELFN, ad hoc TCP (ATCP) also uses a network layer feedback mechanism
to make the TCP sender aware of the status of the network path over which the TCP packets are
propagated. Based on the feedback information received from the intermediate nodes, the TCP
sender changes its state to the persist state, congestion control state, or the retransmit state.
When an intermediate node finds that the network is partitioned, then the TCP sender state is
changed to the persist state where it avoids unnecessary retransmissions.
• When ATCP puts TCP in the persist state, it sets TCP's congestion window size to one in order to
ensure that TCP does not continue using the old congestion window value. This forces TCP to
probe the correct value of the congestion window to be used for the new route. If an
intermediate node loses a packet due to error, then the ATCP at the TCP sender immediately
retransmits it without invoking the congestion control algorithm. In order to be compatible with
widely deployed TCP-based networks, ATCP provides this feature without modifying the
traditional TCP. ATCP is implemented as a thin layer residing between the IP and TCP protocols.
The ATCP layer essentially makes use of the explicit congestion notification (ECN) for maintenance
of the states.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Figure 14 (a) shows the thin layer implementation of ATCP between the traditional TCP layer and
the IP layer. This does not require changes in the existing TCP protocol. This layer is active only at
the TCP sender. The major function of the ATCP layer is to monitor the packets sent and received
by the TCP sender, the state of the TCP sender, and the state of the network. Figure 14 (b) shows
the state transition diagram for the ATCP at the TCP sender. The four states in the ATCP are (i)
NORMAL, (ii) CONGESTED, (iii) LOSS, and (iv) DISCONN. When a TCP connection is established, the
ATCP sender state is in NORMAL. In this state, ATCP does not interfere with the operation of TCP
and it remains invisible.

Fig. 14

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• When packets are lost or arrive out-of-order at the destination, it generates duplicate ACKs. In
traditional TCP, upon reception of duplicate ACKs, the TCP sender retransmits the segment under
consideration and shrinks the contention window. But the ATCP sender counts the number of
duplicate ACKs received and if it reaches three, instead of forwarding the duplicate ACKs to TCP, it
puts TCP in the persist state and ATCP in the LOSS state. Hence, the TCP sender avoids invoking
congestion control. In the LOSS state, ATCP retransmits the unacknowledged segments from the
TCP buffer. When a new ACK comes from the TCP receiver, it is forwarded to TCP and the TCP
sender is removed from the persist state and then the ATCP sender changes to the NORMAL
state.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• When the ATCP sender is in the LOSS state, the receipt of an ECN message or an ICMP source
quench message changes it to the CONGESTED state. Along with this state transition, the ATCP
sender removes the TCP sender from the persist state.
• When the network gets congested, the ECN4 flag is set in the data and the ACK packets. When
the ATCP sender receives this ECN message in the normal state, it changes to the CONGESTED
state and just remains invisible, permitting TCP to invoke normal congestion control mechanisms.
• When a route failure or a transient network partition occurs in the network, ATCP expects the
network layer to detect these and inform the ATCP sender through an ICMP destination
unreachable (DUR) message. Upon reception of the DUR message, ATCP puts the TCPsender into
the persist state and enters into the DISCONN state. It remains in the DISCONN state until it is
connected and receives any data or duplicate ACKs. On the occurrence of any of these events,
ATCP changes to the NORMAL state. The connected status of the path can be detected by the
acknowledgments for the periodic probe packets generated by the TCP sender.
• The receipt of an ICMPDUR message in the LOSS state or the CONGESTED state causes a
transition to the DISCONN state. When ATCP puts TCP into the persist state, it sets the congestion
window to one segment in order to make TCP probe for the new congestion window when the
new route is available. In summary, ATCP tries to perform the activities listed in Table 2.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Table - 2

 Advantages and Disadvantages

• Two major advantages of ATCP are (i) it maintains the end-to-end semantics of TCP and (ii) it is
compatible with traditional TCP. These advantages permit ATCP to work seamlessly with the
Internet. In addition, ATCP provides a feasible and efficient solution to improve throughput of TCP
in ad hoc wireless networks.
• The disadvantages of ATCP include (i) the dependency on the network layer protocol to detect the
route changes and partitions, which not all routing protocols may implement and (ii) the addition
of a thin ATCP layer to the TCP/IP protocol stack that requires changes in the interface functions
currently being used.
Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi
Split TCP
• One of the major issues that affects the performance of TCP over ad hoc wireless networks is the
degradation of throughput with increasing path length, as discussed early in this chapter. The
short (i.e., in terms of path length) connections generally obtain much higher throughput than
long connections.
• This can also lead to unfairness among TCP sessions, where one session may obtain much higher
throughput than other sessions. This unfairness problem is further worsened by the use of MAC
protocols such as IEEE 802.11, which are found to give a higher throughput for certain link-level
sessions, leading to an effect known as channel capture effect. This effect leads to certain flows
capturing the channel for longer time durations, thereby reducing throughput for other flows.
The channel capture effect can also lead to low overall system throughput.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• Split-TCP provides a unique solution to this problem by splitting the transport layer objectives into
congestion control and end-to-end reliability. The congestion control is mostly a local
phenomenon due to the result of high contention and high traffic load in a local region. In the ad
hoc wireless network environment, this demands local solutions. At the same time, reliability is
an end-to-end requirement and needs end-to-end acknowledgments.
• In addition to splitting the congestion control and reliability objectives, split-TCP splits a long TCP
connection into a set of short concatenated TCP connections (called segments or zones) with a
number of selected intermediate nodes (known as proxy nodes) as terminating points of these
short connections. Figure 15 illustrates the operation of split-TCP where a three segment split-TCP
connection exists between source node 1 and destination node 15. A proxy node receives the TCP
packets, reads its contents, stores it in its local buffer, and sends an acknowledgment to the
source (or the previous proxy). This acknowledgment called local acknowledgment (LACK) does
not guarantee end-to-end delivery.
• The responsibility of further delivery of packets is assigned to the proxy node. A proxy node
clears a buffered packet once it receives LACK from the immediate successor proxy node for that
packet. Split-TCP maintains the end-to-end acknowledgment mechanism intact, irrespective of
the addition of zone-wise LACKs. The source node clears the buffered packets only after receiving
the end-to-end acknowledgment for those packets.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Fig. 15

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

• In Figure 15, node 1 initiates a TCP session to node 15. Node 4 and node 13 are chosen as proxy
nodes. The number of proxy nodes in a TCP session is determined by the length of the path between
source and destination nodes. Based on a distributed algorithm, the intermediate nodes that receive
TCP packets determine whether to act as a proxy node or just as a simple forwarding node.
• The most simple algorithm makes the decision for acting as proxy node if the packet has already
traversed more than a predetermined number of hops from the last proxy node or the sender of the
TCP session. In Figure 15, the path between node 1 and node 4 is the first zone (segment), the path
between nodes 4 and 13 is the second zone (segment), and the last zone is between node 13 and 15.
• The proxy node 4, upon receipt of each TCP packet from source node 1, acknowledges it with a LACK
packet, and buffers the received packets. This buffered packet is forwarded to the next proxy node
(in this case, node 13) at a transmission rate proportional to the arrival of LACKs from the next proxy
node or destination. The transmission control window at the TCP sender is also split into two
windows, that is, the congestion window and the end-to-end window. The congestion window
changes according to the rate of arrival of LACKs from the next proxy node and the end-to-end
window is updated based on the arrival of end-to-end ACKs.
• Both these windows are updated as per traditional TCP except that the congestion window should
stay within the end-to-end window. In addition to these transmission windows at the TCP sender,
every proxy node maintains a congestion window that governs the segment level transmission rate.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

 Advantages and Disadvantages
• Split-TCP has the following advantages: (i) improved throughput, (ii) improved throughput
fairness, and (iii) lessened impact of mobility. Throughput improvement is due to the reduction in
the effective transmission path length (number of hops in a zone or a path segment). TCP
throughput degrades with increasing path length. Split-TCP has shorter concatenated path
segments, each operating at its own transmission rate, and hence the throughput is increased.
This also leads to improved throughput fairness in the system. Since in split-TCP, the path
segment length can be shorter than the end-to-end path length, the effect of mobility on
throughput is lessened.
• The disadvantages of split-TCP can be listed as follows: (i) It requires modifications to TCP
protocol, (ii) the end-to-end connection handling of traditional TCP is violated, and (iii) the failure
of proxy nodes can lead to throughput degradation. The traditional TCP has end-to-end
semantics, where the intermediate nodes do not process TCP packets, whereas in split-TCP, the
intermediate nodes need to process the TCP packets and hence, in addition to the loss of end-to-
end semantics, certain security schemes that require IP payload encryption cannot be used.
During frequent path breaks or during frequent node failures, the performance of split-TCP may
be affected.

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

A Comparison of TCP Solutions for Ad Hoc
Wireless Networks

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi

Pratik Shankar, Dept. of ECE, PDACE, Kalaburagi