Comparative Study of MANET Routing Protocol AODV with TCP Variants (RENO and NEW RENO) by varying mobile nodes using OPNET Parulpreet Singh, Ekta Barkhodia, Gurleen Kaur Walia
[email protected] Abstract - The transport control protocol (TCP) is responsible for providing reliable data transport in the Internet. As a matter of fact, it is extensively tuned to providing high-quality performance in the conventional wired network. However, it cannot offer reliable service while using e-mail, internet search, several application file transmission in a mobile Ad-hoc network. Several studies reveal that TCP does not perform as well in a mobile environment as it does in other networks. There are several factors that affect the TCP performance in MANETs, such as dynamic topology, shared medium, signal fading and high bit errors. , there are other factors like network size, network load, bandwidth and signal strength that affect the performance of the MANET routing protocols. Therefore, a detailed analysis is required in order to gain an insight of these factors that determine the performance of the routing protocol. Keywords: - MANETs, OPNET, Routing Protocol, TCP. I. Introduction A Mobile Ad-hoc Network (MANET) is a collection of mobile devices dynamically forming a communication network without any centralized control and pre-existing network infrastructure. Due to the presence of mobility in the MANET, the interconnections between stations are likely to change on a continual basis, resulting in frequent changes of network topology. Consequently, routing becomes a vital factor and a major challenge in such a network. The four routing protocols that are considered in the analysis are Optimized Link State Routing (OLSR), Adhoc On-demand Distance Vector (AODV) etc. In addition, from a transport layer’s perspective, it is necessary to consider Transmission Control Protocol (TCP) as well for MANETs because of its wide application, which enjoys the advantage of reliable data transmission in the Internet. However, the factors such as scalability and mobility cause TCP to suffer from a number of severe performance problems in an ad-hoc environment. Hence, it is of utmost importance to identify the most suitable and efficient TCP variants that can robustly perform under these specific conditions.
Different routing protocols:1) OLSR 2) AODV 3) TORA 4) DSR etc OLSR: - The Optimized Link State Routing (OLSR) is operated as a proactive (table-driven) routing protocol i.e. frequently exchanges topology information with other nodes of the network. This protocol is basically an optimization of traditional link state protocol developed for mobile ad-hoc network. The responsibilities of OLSR protocol are to minimize the required number of control packets transmission and also to shorten the size of control packets. AODV: - The Ad-hoc On Demand Distance Vector (AODV) is considered an efficient MANET routing protocol and supports both uni cast and multicast routing mechanisms. The AODV routing protocol utilizes an on-demand technique in order to discover the routes. This means that the route between two endpoints (nodes) is formed as per requirement for the source node and maintained as long as the routes are needed. Moreover, the protocol uses a destination sequence number to recognize the most recent path and to guarantee the freshness of the routes. Reactive protocols like AODV shrinks the control traffic overhead at the cost of higher latency in discovering new routes Transport Control Protocol: - TCP is known as a full duplex protocol meaning each TCP connection provides a pair of byte streams in both directions. TCP implements the congestion control mechanism with each of these byte streams so that the receiver can limit the sender from transmitting more data in the network. Several studies reveal that TCP does not perform as well in a mobile environment as it does in other networks. There are several factors that affect the TCP performance in MANETs, such as dynamic topology, shared medium, signal fading and high bit errors. For dynamic topology, the packet losses
Thus for the multiple packet loss, the New-Reno becomes able to retransmit all the packets.
occur due to the broken routes between the nodes whereas TCP assumes that the losses are due to the network congestion. Therefore, the network experiences the counterproductive invocation of congestion control mechanisms employed by the TCP. TCP VARIANTS: 1) TAHOE 2) RENO 3) NEW RENO 4) SACK etc. These are the different TCP variants (versions) which are used in MANETs. In this paper performance under RENO and NEW RENO is going to be discussed. TCP Tahoe:-The earlier versions of TCP offered a goback-n model which used to implement the cumulative positive acknowledgment. For this purpose, retransmit timer expiration was needed in order to re-transmit the lost data. However, these former versions were unable to reduce the network congestion. Hence, for modification to earlier TCP implementations, the TCP Tahoe variant was implemented with slow-start, congestion avoidance, and fast retransmits algorithms. TCP Reno:- Along with the implementation of the basic principles of Tahoe, the TCP Reno version adds more mechanisms so as to detect the lost packets in shorter time and also prevent the pipeline from being empty every time a packet is lost. The packet segment is assumed to be lost as soon as the duplicate acknowledgements are reached to its threshold level. Then the TCP enters the Fast Re-transmit phase through which the lost segment is retransmitted. When the Fast Retransmit phase is completed, TCP Reno employs the Fast Recovery algorithm which does not let the pipeline to be empty and also provides extra incoming duplicate ACKs to clock subsequent outgoing packets TCP New Reno:- In case of multiple packet loss, the TCP New-Reno does not wait for the retransmission timer to be expired and hence this variant provides a dominating performance over the Reno version. In New Reno, the performance concerns about the behavior of the partial ACKs, which do not take TCP out of Fast Recovery phase while it takes TCP out from the Fast Recovery phase in Reno version. Moreover, in New-Reno, receiving partial ACKs often indicates the loss of the packets which instantly follows the acknowledged packet in the sequence space.
II. Related Work There are large number of papers try to simulate different number of protocols of MANETs under different conditions of different mobile nodes and variable mobility. The basic purpose for this paper is to find out that what will be the performance of routing protocol AODV with different versions of TCP before applying it into real world. It is very complicated to consider all relevant events that can happen in mobile wireless networks and their affect on the performance of the protocols. Simulative comparisons often use the standard models of NS-2 or OPNET. We use OPNET to simulate the network under different conditions.
III. Tools used OPNET modeler: - NS-2 and OPNET are the two very well-known simulators. NS-2 is open source software while OPNET is a commercial simulator and the kernel source code of OPNET modeler is not open for all. However, OPNET has a comprehensive built-in development environment to design and simulate network models .This research is conducted using discrete event simulation software known as OPNET Modeler, which is just one of several tools provided from the OPNET Technologies suite. In order to undertake the experimental evaluation, the most recently available version, namely the OPNET Modeler 14.5 has been adopted in our study. The OPNET is one of the most extensively used commercial simulators based on Microsoft Windows platform, which incorporates most of the MANET routing parameters compared to other commercial simulators available. Aside from this, the modeler incorporates a number of features to support an increase in stability and mobility in the mobile ad-hoc network. In OPNET simulator, a number of parameters are present for MANET environment in order to study the overall network performance. These parameters are known as performance metrics. Specific application and transport layer protocols demand on an own set of performance metrics to evaluate the network efficiency. For instance, with the introduction of a variety of network parameters, end-to-end delay and average throughput are substantially affected by the routing algorithm; hence, such parameters play an important role in the selection of an efficient routing protocol in any communication network.
IV.
Performance Metrics:-
Throughput:The average rate at which the data packet is delivered successfully from one node to another over a communication network is known as throughput. The throughput is usually measured in bits per second (bits/sec). A throughput with a higher value is more often an absolute choice in every network. Mathematically, throughput can be defined by the following formula. Throughput= (number of delivered packet * packet size)/total duration of simulation End-to-End Delay:The end-to-end delay is the time needed to traverse from the source node to the destination node in a network. Endto-end delay assesses the ability of the routing protocols in terms of use- efficiency of the network resources.
V. Simulation Setup The different types of scenarios configured with different network objects has been taken, for a node speed of 10 m/s, a pause time of 200 sec, a network size of 20,50 nodes and the file size of 50,000 bytes (for FTP) in a (1000×1000) square meter area. The server module is configured to support and control the application services (i.e. FTP ) based on the user profile. This is basically a WLAN server through which a particular routing protocol and a TCP variant can be selected. The nodes are defined as workstations with client server applications running over TCP/IP, support the underlying WLAN connection at 1 Mbps, 2 Mbps, 5.5 Mbps and 11 Mbps. The connection speed is set at 11 Mbps in this paper. This paper is going to represent scenarios based on TCP variants (RENO and NEW RENO) using AODV routing of 25 and 20 mobile nodes, showing graphically their delay and throughput. The simulation time is 150 seconds for all cases.
Fig-1. 20 node network with TCP variant (NEW RENO)
Fig-2. 20 node network with TCP variant (RENO)
Fig-4. Comparison of the throughputs of 20 nodes RENO and NEW RENO
Fig-3. Comparison of the delays of 20 nodes RENO and NEW RENO Fig-5. 45 node network with TCP variant (NEW RENO)
Fig-6. 45 node network with TCP variant (RENO)
Fig-8. Comparison of the throughputs of 20 nodes RENO and NEW RENO
Table-1. Resultant values of delays and throughputs of 20 and 45 nodes campus networks with RENO and NEW RENO.
Number of nodes
TCP Variant
Routing protocol
Delay (sec)
Through put(bits/ sec)
20
RENO
AODV
0.019
3.5x10^6
20
NEW RENO
AODV
0.022
4.5x10^6
45
RENO
AODV
0.058
6.1x10^6
45
NEW RENO
AODV
0.059
5.9x10^6
VI.
Fig-7. Comparison of the delays of 45 nodes RENO and NEW RENO
Conclusion
In this paper the performance analysis of routing protocols AODV protocol in MANET have been investigated. The investigation consider the impact of scalability, mobility, network Ftp traffic load on together with the Reno and New-Reno TCP variants on the network performance. In the performance assessment delay and throughput are adopted for the whole scenarios considered.
The simulation using OPNET consider different scenarios that attempt to cover all of the aspects on network evaluation required. The following outcomes can be extracted from the Simulation results:Other routing protocols like DSR, TORA, DSDV, ZRP etc .with TCP VARIANTS can be used, and aside from this which routing protocol is most suitable for by using these variants can be find out. From this paper it can be concluded that when there are more number of nodes more will be the delay and throughput. References [1] V. Talooki and K. Ziarati, “Performance Comparison of Routing Protocols For Mobile Ad Hoc Networks” Asia-Pacific Conference on Communications, APCC, 2006. [2] L. Subedi, M. Najiminaini, and L. Trajkovic “Performance Evaluation of TCP Tahoe, Reno, Reno with SACK, and NewReno Using OPNET Modeler NewReno Using OPNET”. [3] A.Klein performance evaluation and comparison of AODV,OLSR and SBR in MANET in third international symposium on modeling , analysis and simulation of computer and telecomm. System 2007, MASCOTS’07, nov.2008
[4] MANET routing protocol performance evaluation with TCP TAHO,RENO and NEW RENO by prof siddeq.Y.Ameen and Ibrahim.A.Ibrahimi in international journal of e- and u- service, science and technology, MARCH 2011 vol-4. [5] R. Misra and C.R. Mandal, “Performance comparison of AODV/DSR on-demand routing protocols for ad hoc networks in constrained situation” ICPWC International Conference, IEEE, 2005, pp. 86 – 89 [6]Amr M. Hassan, Mohamed I. Youssef and Mohamed M. Zahra, “Evaluation of Ad Hoc Routing Protocols in Real Simulation Environments” The 2006 International Conference on Computer Engineering and Systems, 2006, pp. 288 – 293
BIOGRAPHY
Parulpreet Singh, Final year student of M-TECH (ECE) at Lovely Professional University (LPU). Pursuing dissertation in mobile ad-hoc networks.
