Improvement of Robustness for Ad Hoc Networks Through
Energy-Aware Routing

[attachment=17992]

Introduction
In an ad hoc network the mobile nodes agree to serve as
both routers and hosts. The nodes can dynamically join
and leave the network, frequently without warning, and
possibly disrupting communication amongst other nodes.
Moreover, the limitations on power consumption imposed
by portable wireless radios result in a node transmission
range that is typically small relative to the span of the network.
This limits the propagation range of a mobile node.
In such an environment, it may be necessary for one mobile
host to enlist the aid of others in forwarding a packet
to its destination. One of the main challenges in the design
of ad hoc networks is the development of dynamic routing
protocols that can efficiently find routes between two communicating
nodes.


Background and motivation
The issue of energy1 conservation in ad hoc networks has
stirred up a significant amount of interest amongst many research
groups. Numerous protocols have been proposed to
address this problem. Metrics that have been proposed for
improving the battery life of a mobile node can be broadly
classified into two major categories: power-aware and costaware.
Power-aware metrics aim at minimizing the total
power. Here the important point to note is that the term
power implies power consumed by the node in routing a
packet between a source and its destination. Cost-aware
metrics look at methods that extend the nodes’ battery lifetimes
[2]. We are also aware of the recent development of
Yu and Lee along the same lines of our proposed algorithm
[1]. However there are striking differences between the two
approaches and also in the actual purpose behind proposing
both the algorithm. The differences are discussed in
the subsequent sections.


2.1 Robustness and energy efficiency
The term robustness refers to the ability of a routing protocol
to induce minimum initial set- up delay and maintain
connectivity without a significant degradation in throughput
even when there is movement of nodes that are actively
participating in the communication. A robust protocol must
also find an optimal path to the destination based on several
other factors apart from the traditional ’number of hops’
metric. A route is said to be optimal based on factors like
the least number of hops and the minimum energy consumed.
This will avoid the inclusion of redundant nodes
and hence minimize consumption of energy per end-toend
connection. The routing protocol must also be robust
enough to rediscover the optimal route to the node if there
is a better route that might become available later in time.


2.2 The network partitioning problem
The partitioning problem refers to a scenario where nodes
in an ad hoc network divide into two or more clusters that
eventually lead to loss of connectivity. Such adverse scenarios
can be observed in practical situations while maintaining
connectivity among troops in a battlefield. For
example, during combat operations, groups of nodes are
eventually formed within an ad hoc network. The groups
are a result of assignment to a common task in the battlefield.
Grouping of mobile nodes may result in partitioning
of an ad hoc network into multiple groups. In such situations,
the approach adopted by the routing protocol to
maintain connectivity for as long as possible plays a very
important role. Later, these groups might converge again to
re-form the original ad hoc network.


3.1 Performance Metric
The aim here is to find the most robust protocol amongst
those selected as reasonable candidates for ad hoc networks,
which relates to the ability of the routing protocol
to maintain the connectivity for as long as possible even
during adverse situation. This also implies that the protocol
should avoid any sort of degradation in the throughput
of the network during adverse conditions. Hence, throughput
was chosen as the metric to assess the robustness of
the protocol. The average throughput was calculated separately
for every connection in the network.