Route maintenance parameters and determining factors of AODV routing stability in AD-HOC networks

Abstract

The ad-hoc network is an infrastructure-less, self-configuring, self-motivated, arbitrary, rapidly changing, and multi-hop network that is composed of bandwidth constrained wireless links without an aid of the centrally controlled routers or servers. Owing to these properties, it has gained potential applications in various fields like tactical and civilian environments, emergency operations etc. The possible applications of the ad-hoc network in many fields as well as its contemporary relevance and future promise in the next generation of wireless communications, which has indeed motivated the thesis’s author to select this area as his research work. An ad-hoc network is vulnerable due to its fundamental characteristics such as open medium, dynamic topology, distributed operation and constrained capability etc. Hence, stable routing is one of the challenging tasks in this network. However, routing in adhoc network is highly complex because of mobility, limited transmission range and security issues. Therefore, the author has been motivated to analyze and observe the routing protocol performance (especially ad-hoc on-demand distance vector (AODV)) in the wide range of scenarios. The AODV routing algorithm is actually motivated due to its limited bandwidth that is available in the media and used for communications, especially in the wireless medium. The idea of getting routes purely on-demand makes AODV routing a very useful and preferred algorithm for the ad-hoc network environment. As AODV routing protocol comes under the on-demand routing. Hence, in this routing, the route maintenance parameters are one of the key concepts for the route discovery and maintenance process, which basically deals with the topology changes and provides a stable path in the network. Moreover, some available studies concerning to the ad-hoc networks have been reviewed in this thesis. From there, one can easily observe that the present thesis marks a shift from the previous research work in which it considers the various route maintenance parameters for the study. Therefore, the main variables of this thesis are the number of route maintenance parameters (such as active route timeout (ART) & delete period constant (DPC)) and various determining factors (like mobility, transmission range, network load density (NLD) and active nodes/ source destination pair (ANs/SD pairs)) that may significantly affect the AODV network stability. The prime purpose of the present thesis is to provide more stable routing in AODV network. In order to do this, it is absolutely essential to take into consideration the timeout value for cached routes and timeout value after which the expired cached routes are completely deleted from the routing table. These two timeout values are called as ART and Delete Period (DP) respectively, and they minimize the route failure issues by providing alternative routes for the data packets. In this thesis, ART and DPC (where DPC denotes the multiple for DP, i.e. DP = DPC × Max {ART or Hello_Interval}) are termed as route maintenance parameters and are being used to improve the AODV routing performance. Actually, in AODV routing, it is suggested that the value of ART and DPC should be a constant. In other words, their value is generalized for all kinds of applications or traffic generators. However, it has been observed from the thesis work that the choice of their values according to network behavior and traffic generators may significantly increase the network performance or provide the stable routing in the network. In order to observe and analyze the AODV routing performance under the wide range of created scenarios and situations, the present thesis is organized into following seven chapters. In chapter 1, the theme of the current research work is reported in detail. The historical development of the wireless technology and ad-hoc network has been presented in a chronological manner. Further, chapter also explains the fundamental concepts of the ad-hoc network with its possible present and future applications along with its major salient features. Moreover, some available studies related to ad-hoc networks are reviewed, and the close analysis of the detailed literature review establishes the research gaps. Finally, this chapter concludes with the motivation, research objectives, and organization of thesis. In chapter 2, the AODV routing functioning has been explained in depth. Moreover, this chapter also lists the various route maintenance parameters with their proper explanation, and clearly describes the concept of ART & DPC. Finally, this chapter comes up with the various determining factors, and discusses about how these different determining factors are affecting the AODV performance. In chapter 3, the AODV performance is analyzed by variation in the route maintenance parameters from their default value by using two different simulation tools (NS-3 & QualNet). As, in most of the reported work, the ad-hoc networks were simulated as a function of mobility, number of nodes, and network size but not as a function of the route maintenance parameters. Therefore, this chapter is showing its interest to analyze how the AODV performance is influenced when ART & DPC are varied from their default value. This chapter conducts two different simulation studies under different simulation tools. In the first study, the AODV performance has been observed for a constant scenario by using NS-3. It is tested under the random waypoint topology for a fixed number of nodes, fixed CBR SD pair, and is subjected to the IEEE 802.11 MAC protocol. However, if the result is taken for network performance only on net throughput or PDR, the maximum performance of AODV is found for ‘ART’ & ‘DPC’ combination of 2.5 & 5.0, respectively. Although the AODV protocol suggested ART=3. However, the best performance is observed at ART=2.5. Therefore, as per this analysis ART value is 0.5 second less than the original default value, which results in less memory overheads. In other words, especially, in this type of scenario to get more stable routing, ART value should be less than its default. In the 2nd study, the AODV performance is tested under random waypoint topology for a constant scenario, but for different SD pairs and by using QualNet. Here, two simulation experiments are conducted for different SD pairs (5 & 7). ART & DPC variations impact is analyzed on throughput, delay, jitter, and percentage of loss packet. The impact of these variations phenomena has deeply been analyzed on AODV routing. This study concludes that ART=3 acquires the best performance for all QoS metrics except in case of delay for 5 SD pair, whereas at different DPC. For all QoS, DPC value is less than its original default value excluding jitter, which results in less memory overheads. It is finally concluded that the network performance is constrained of QoS, and choice of the ART & DPC default value may depend on the user’s QoS requirement. Chapter 4 analyzes and compares the AODV performance for CBR & VBR traffic under the influence of route maintenance parameters. As, in most of the reported work, several researchers have carried out comparative studies between various routing protocols. However, they were limited to CBR & TCP models, and very few included the route maintenance parameters but only for CBR. Therefore, this chapter presents the comparative study between CBR & VBR under the influence of route maintenance parameters. Since in real time applications, mostly traffic is variable. Hence, it is absolutely necessary and interesting to carry out this comparative study between CBR & VBR to observe the performance differences. Chapter 4 mainly conducts two different simulation studies to analyze and compare the AODV performance for both traffics: one is for variable ART at fixed ‘NLD & SD pair’ and the second one is for variable SD pair at fixed ‘NLD & ART’. The first study concludes that the curve ART=1 at DPC=6 gives the best performance for VBR traffic, whereas for CBR traffic best performance is observed by the curve ART=3 at DPC=3. Therefore, it can be said that whenever traffic is variable, there is no need to hold route state information for a longer time because it creates more memory overheads. Moreover, this chapter also concludes that almost for all QoS, the overall CBR performance is better than VBR. The VBR poor performance may due to the variable nature of traffic where synchronization is needed each time between the particular source & destination nodes. The 2nd study concludes that the network performance increases with an increase in the SD pair in terms of throughput and delay for both traffics. However, it is expected that after a particular increase in the SD pair, the performance will start to decrease because of a large number of control overhead packets in the network. Finally, this chapter concludes that the choice of default value of ART & DPC also depends on traffic generators. Moreover, the selection of their proper default values according to the applications may greatly increase the network performance, and reduces the memory overheads. In chapter 5, presented work tries to analyze an optimal relation between the route maintenance parameters and various determining factors in which AODV network provides a stable routing. As, in most of the reported work, several researchers have attempted to observe the AODV performance by considering various factors like network size, number of nodes, mobility etc. However, they were restricted to only these factors and a few of them separately considered the route maintenance parameters only, but the combine research work between the route maintenance parameters and determining factors has not been conducted yet. Therefore, an attempt is also made to analyze how AODV performance is influenced, when various determining factors are taken into account along with the route maintenance parameters. Here, provided traffic is CBR, and the performance observation is based on throughput, delay, & jitter. In order to analyze the best relation between the various determining factors and route maintenance parameters, this chapter creates two different scenarios, and each scenario presents two separate cases. The case-1 of scenario-1 analyzes the impact of variations of ART on different QoS for various node’s mobility values. This case concludes that at higher node’s mobility, the performance is not good with an increase in ART value. It may be due to the frequent topology changes that are more common at higher node’s mobility. Hence, it is suggested that the route state information should not be kept for a longer time in a highly mobile environment. If mobility <0.5 mps, the performance becomes good and it is nearly constant for other higher ART values. This result is obvious because the node’s mobility is almost stationary and hence, changes in ART value do not affect the performance. Here, the best throughput is observed at ART=1 whereas for delay & jitter, the best one is at ART=0.5. Moreover, its 2nd case concludes that the overall throughput increases as the NLD increase. However, after a certain increase in NLD value, throughput starts decreasing because of the constant area and there could be congestion in the network. Further, it concludes that at ART=3 for all NLD, throughput is decreased because of the slow reaction to the rapid changes in the network topology, particularly at higher node’s mobility. Moreover, at ART=1, it has a significant throughput outcome at higher node's mobility. Its 2nd scenario is mainly concerned to analyze the effect of various transmission ranges on the throughput at default QualNet transmission power and calculated transmission powers which have been calculated by equation-1. Here, 1st case concludes that the throughput value increases up to 300 meters of the transmission range. After that, it is decreased because 15dbm transmission power is not adequate to transmit the packets beyond 300 meters. It is also concluded that at ART=1, maximum throughput is acquired almost for all transmission ranges, and it is almost constant for other higher ART values. It may due to constant node’s mobility throughout the simulation. Furthermore, the same scenario is observed for the calculated transmission powers. It is concluded that the throughput increases up to 350 meters. Later on, it becomes almost constant for higher values of transmission range. Its 2nd case concludes that the throughput value is higher at lower node's mobility for all transmission ranges. Moreover, it is almost constant at high node’s mobility that is seen in higher transmission ranges, particularly. In chapter 6, the comparative performance analysis for various routing protocols has been carried out under varying pause time environment. Many research studies have already been conducted to analyze and compare the characteristics of various routing protocols through different QoS for various situations and applications. However, these studies have considered the different set of routing protocols for their analysis as chosen in this thesis work. Hence, the prime concern of this chapter is subjected to comparative performance analysis between five routing protocols, out of which four are reactive (AODV, DSR, DYMO, & IERP) in nature and the remaining one is IARP that is proactive. The performance observation and comparison are based on throughput, delay, jitter, & PDR. This chapter concludes that the AODV gives better throughput and PDR than the other remaining protocols throughout the variation of pause time. Although, it also performs satisfactorily in terms of delay and jitter, after IARP. The reason for that may be the lesser number of control overhead packets because of on-demand nature. Also, another reason may be the quick response to the route breakage and topology changes in the network promptly. Further, chapter concludes that the best delay and jitter is observed by IARP because of some features of the proactive maintenance that makes it more suitable towards the topology changes. Moreover, the multiple-hop paths concept within routing zone may also be the reason for that. In chapter 7, the work embodied in the present thesis has been summarized, and the significant conclusions have been drawn from the major findings. Furthermore, this chapter also includes the future prospective of the current research work.