Design and Development of ECA based Routing Protocols for Ubiquitous Network

Abstract

The ubiquitous network is a heterogeneous wireless network comprising of computing devices such as laptops, smartphones, tablets, PDAs, etc., are connected to any device at any time, anywhere that enabling mobile users to access and exchange information through network access technologies such as Wi-Fi, Wi-Max, 5G, Bluetooth, UMTS and so on. The Ubiquitous Network (UN) fragmented into subnetworks in which mobile nodes of one subnetwork can exchange information with other mobile nodes in another subnetwork via network access technologies. The optimal route establishment is vital to exchange packets starting from the originating node to the desired destination.

It is important to design and build routing protocols that play a crucial role in diverting the packets to different routes through an optimal path from the originating node to the ultimate destination. In essence, reactive and proactive routing protocols are designed and developed to determine the optimal routes for the network-wide propagation of control and data packets. The UN mobile nodes move in an arbitrary direction and frequently join or depart the UN at any point in time. Many routing-related issues have emerged due to node mobility, such as frequent network topology change, link failure, node disappearance, high latency, packet loss, etc. To provide the solution to the preceding issues, we design and develop routing protocols for the ubiquitous network using the Event Condition Action (ECA) scheme.

The objectives of the thesis is to design solutions to some of the ubiquitous network routing issues. The ECA scheme is employed to make dynamic routing decisions at runtime under critical network conditions. In this thesis, we designed and developed reactive and proactive routing protocols to determine the optimal routes. The thesis consists of four major ECA-based research works viz; i) AODV route establishment and maintenance; ii) Improved AODV routing protocol; iii) Multicast group creation and selection of group leader; iv) Designing a link-state routing protocol. The ECA scheme structure is composed of three modules, like the event, condition, and action. The function of the event module is to accumulate a stream of routing events within the UN environment. The condition module’s job is to apply the logic rules that correspond to the event stream and execute the rule quickly post the conditions are met. Eventually, the action module’s task is to make dynamic routing decisions at runtime within a stipulated time.

The design and development of routing protocols are present as follows. First of all, we designed AODV Route Establishment and Maintenance (AREM) to implement subroutines necessary for finding an optimal route from the originating node to the intended destination like route discovery, route setup, and route maintenance. The events like link failure or node disappearance from the network occur, then the source node nullifies the route and provokes the route investigation process once again. However, we have modified the control packet format in our proposed ECA scheme by adding metrics such as bandwidth, processor speed, buffer capacity, and battery power of the node to avoid reinitiation of the route discovery process and select an optimal path from where a route is broken. The proposed scheme discovers the shortest route based on Resourceful Mobile Nodes (RMN) that is an added advantage for rerouting the packets in a critical routing condition and making an intelligent routing decision at runtime. We estimated the time and space complexity of the developed protocol by selecting various routing functionalities. We developed a performance modeling using the M/M/1 queuing theory concept to investigate that dynamic routing decision in the UN. Dynamic routing decisions essentially depends on the expected time required for processing the ECA rules in the local memory aspect of each mobile node and the time required to process an ECA rule in the UN aspect.

Secondly, we designed an Improved AODV Routing Protocol (IARP). In the IARP, we designed and implemented two improved routing protocols specifically, ECA based Improved AODV Position and Speed Aware Routing (ECA-IAODV-PSAR) and Improved AODV Edge Connectivity (ECA-IAODV-EC). In specific, these two routing protocols use the speed and position information of the node received from the Global Positioning System (GPS) to deal with routing control packet overhead during elevated mobility. The primary routing protocol ECA-IAODV-PSAR is an alteration of the PSAR routing protocol. The ECA-IAODV-PSAR protocol set limits flooding of the route query packet to the tiny zone of the UN to select the optimal route towards the ultimate destination. We calculated the time and space complexity of the developed routing protocols and built the performance modeling. Distinct performance parameters are evaluated comprised of routing control packet overhead, the number of route query packets, packet delivery ratio, normalized routing load, and latency.

Thirdly, we have designed the Multicast Group Creation and Selection of Group Leader (MGCSGL), in which we have designed and implemented the ECA based Multicast Ad-hoc On-demand Distance Vector (ECA-MAODV) routing protocol. The ECA-MAODV protocol plays a predominant role in selecting an efficient route by creating an efficient multicast tree and selecting a unique group leader using a Computational Intelligence (CI) approach. The vague set is a CI that is an enhanced fragment of a fuzzy set. Each element in the Fuzzy set is mapped to [0,1], reflecting its membership grade. We estimated ECA-MAODV’s time and spatial complexity. Besides, we assessed ECA-MAODV performance and compared it with conventional routing protocols such as MAODV and AODV.

Finally, we designed a Link State Routing Protocol (LSRP) using the novel agent technology. LSRP includes two types of agents like OLSR (acts as a static agent) and MPR (acts as a mobile agent). The OLSR agent accumulates information about neighboring nodes to establish a QoS route to transmit reliable data from the source node to the intended destination. Thus, OLSR creates and dispatches MPR agents across the network. MPR agents migrate and visit every mobile node in the network hosted by an OLSR agent to accumulate and supply mobile node and network status information to the OLSR agent. The OLSR agent collates the information and provides it to the ECA scheme for selecting an intelligent ECA-QoS route. The ECA scheme employs suitable constraints on the detailed information supplied by the OLSR agent to generate the desired QoS path. Eventually, we evaluated the performance and compared ECA-QoS-OLSR with conventional OLSR routing protocols.

In summary, we have designed and developed reactive and proactive routing protocols for the ubiquitous network to make runtime routing decisions using the ECA scheme. • AODV route establishment and maintenance essentially selects an optimal route in the event of the link failure and divert the packets towards the active route. • Improved AODV routing protocol emphasizes the limitation of route discovery area to the UN tiny zone to select an optimal route to the desired destination node. • Design and implementation of the ECA-MAODV routing protocol that plays a predominant role in selecting an efficient route and unique group leader for the creation of an efficient multicast tree and multicast group to communicate and share adequate information among group members more efficiently and effectively. • Designing of a link-state routing protocol that uses novel agent technology to select an intelligent QoS route starting from the originating node to the target destination for reliable transmission of data packets. • Finally, we simulated the aforementioned proposed scheme in the various scenarios by considering varying performance metrics. The results obtained in both simulation and analysis shows the significance of the approaches and the proposed scheme’s effectiveness.