Disaster Relief Management Using Reinforcement Learning-Based Routing

Disaster Relief Management Using Reinforcement Learning-Based Routing

Gajanan Madhavrao Walunjkar (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India & AIT, Pune, India), Anne Koteswara Rao (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India) and V. Srinivasa Rao (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India)
DOI: 10.4018/IJBDCN.2021010102
OnDemand PDF Download:
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Effective disaster management is required for the peoples who are trapped in the disaster scenario but unfortunately when disaster situation occurs the infrastructure support is no longer available to the rescue team. Ad hoc networks which are infrastructure-less networks can easily deploy in such situation. In disaster area mobility model, disaster area is divided into different zones such as incident zone, casualty treatment zones, transport areas, hospital zones, etc. Also, in order to tackle high mobility of nodes and frequent failure of links in a network, there is a need of adaptive routing protocol. Reinforcement learning is used to design such adaptive routing protocol which shows good improvement in packet delivery ratio, delay and average energy consumed.
Article Preview
Top

1. Introduction

Effective disaster management plays very vital role in saving lives of the people who are trapped in the disaster scenario. To carry out complete disaster operation successfully the lifetime of the communication network must be enough. But unfortunately when disaster occurs the infrastructure support is no longer available to the rescue team. Ad hoc networks are more preferred in such applications where the laying down the infrastructure network is not possible either due to short period of time or any emergency situations such as battlefield, military and disaster scenario. The various advantages gained from mobile ad hoc networks over wireless communication are cost factor, resource sharing, good QoS, security and reliability.

The various characteristic of mobile ad hoc network are:

  • 1.

    Multi-Hopping: Packets moves through intermediate nodes to reach to the destination.

  • 2.

    Mobility: Nodes can move randomly which frequently changes the topology.

  • 3.

    Self-Organization: Nodes are intelligent and configures without the support of external entity.

  • 4.

    Limited Energy: Nodes are operated through battery which has limited energy. Energy of the node should be conserved for increasing the overall lifetime of network.

  • 5.

    Bandwidth Constrained: Frequent breakages of communication link changes the topology and again reestablishment of communication link is required. This increases the bandwidth consumption.

Various existing routing protocols are divided into two classes – proactive and reactive (Nossenson & Schwartz, 2013). Proactive protocols always maintain paths and thus consistent information is always available in routing database. Destination-Sequenced Distance Vector (DSDV) comes under this category. Reactive routing protocols (Recent Trends in Networks and Communications, 2010) which are on-demand protocols where the actual communication path is obtained only when never it is required. Dynamic source routing (DSR), ad hoc on-demand distance vector (AODV), ad hoc on demand multipath (AOMDV) comes under this category.

In DSDV, routes are updated periodically by triggering to their neighbors. But it is observed, that routing and caching overhead is high and throughput is low in DSDV. AODV is more suitable and adaptable to large highly dynamic topologies. DSR reduces route discovery overhead but creates more delay to the packets reaching to the destinations. It is also observed, in high mobility, on-demand routing protocols gives good performance. The DSDV and AOMDV generate high number of control packets and thus consume more bandwidth.

There are several mobility models present. Random waypoint model includes select random direction, speed and variable pause time. The node moves towards random destination with a velocity chosen randomly from [0, Vmax]. In Random walk mobility model, node moves to a new location by randomly choosing a direction and speed. This is similar to RWP but nodes change their speed/direction every time slot, New direction θ is chosen randomly between (0, 2π], New speed chosen from uniform (or Gaussian) distribution (Handbook of Mobile Ad Hoc Networks for Mobility Models, 2011). When node reaches boundary it bounces back with (π-θ). In Manhattan Grid Model nodes move only on predefined paths. In disaster mobility model, disaster area is divided into different zones such as incident zone, casualties treatment zone, transport area and hospital zone (Walunjkar & Rao, 2019a) etc. Figure 1 shows the division of zones in disaster area scenario.

Figure 1.

Disaster area scenario

IJBDCN.2021010102.f01

This paper is organized as follows: Section 2 introduces the problem and the issue about non adaptive routing and various methods of designs of adaptive routing. Section 2 also describes the need of further optimization required on adaptive routing protocols in case of higher mobility and higher data rate. The research methodology used in proposed method in detail is given in section 3. Results obtained by performance comparison with existing routing protocol are given in section 4.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 18: 2 Issues (2022): Forthcoming, Available for Pre-Order
Volume 17: 2 Issues (2021)
Volume 16: 2 Issues (2020)
Volume 15: 2 Issues (2019)
Volume 14: 2 Issues (2018)
Volume 13: 2 Issues (2017)
Volume 12: 2 Issues (2016)
Volume 11: 2 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing