Term paper on End-to-End delay And Route Discovery in Ad-hoc Networks

Name harjinder kaur, Roll no. A35,Reg. No.11005705,Section E1E09 Abstract: ad hoc wireless networks are
new communication networks that can be dynamically formed and deformed on-thefly anytime and anywhere. User data is routed with the help of an ad hoc mobile routing protocol.before the deployment of ad hoc mobile services, communication performance of such networks has to be evaluated to demonstrate the practicality limits based on hardware and innovative communication software. This paper describes the realization of ad hoc networks, packet size, beconing interval, route discovery and end to-end delay.

INTRODUCTION
A communication network must be able to transport user traffic towards its targeted destination. The communication performance of a network can effect da satisfaction level of user. The ability to send imformation at high speed demands low end to end delay .User would like to transmit a variety of information such as data,audio and video. An ad hoc network is not useful if it can not offer acceptable communication services.As technology advances, wireless and portable computersand devices are becoming more powerful and capable.These advances are marked by an increase in CPU speed,memory size, disk space, and a decrease in size and power consumption. The need for these devices to continuously

communicate with each other and with wired networks is becoming increasingly essential. MANET is a collection of mobile devices which form a communication network with no preexisting wiring or infrastructure. MANETs allow the applications running on these wireless devices to share data of different types and characteristics. There are many applications of MANETs, each with different characteristics of network size, node mobility, rate of topological change, communication requirements, and data characteristics. Such applications are conferences, classroom, campus, military and disaster recovery. Each node is directly connected to all nodes within its own effective transmission range. Nodes in the network are allowed to move in and out of range of each other. Communication between nodes that are not within range of each other is accomplished by establishing multihop routes that involve other nodes which act as routers. New nodes can join the network at any time and existing nodes can leave the network as well. Due to the dynamic nature of MANETs, designing communications and networking protocols for these networks is a challenging process. One of the most important aspects of the communications process is the design of the routing protocols used to establish and maintain multihop routes to allow the communication of data between nodes. A considerable amount of research has been done in this area, and multihop routing protocols have been developed. Most of these protocols such as the Dynamic Source Routing protocol (DSR), Ad Hoc on Demand Distance Vector protocol(AODV), Temporally Ordered Routing Protocol (TORA) and others establish and maintain routes on a besteffort basis. While this might be sufficient

for a certain class of MANET applications, it is not adequate for the support of more demanding applications such as multimedia, audio and video. Such applications require the network to provide guaranteeson the Quality of Service (QoS).Most QoS routing algorithms represent an extension of existing classic best effort routing algorithms. Many routing protocols have been developed which support establishing and maintaining multihop routes between nodes in MANETsThese algorithms can be classified into two differen categories: ondemand (reactive) such as DSR, AODV, and TORA, and table-driven (proactive) such as Destination Sequenced Distance Vector protocol (DSDV). In the ondemand protocols, routes are discovered between a source and a destination only when the need arises to send data. This provides a reduced overhead of communication and scalability. In the table-driven protocols, routing tables which contain routing information between all nodes are generated and maintained continuously regardless of the need of any given node to communicate at that time. With this approach,the latency for route acquisition is relatively small, which might be necessary for certain applications, but the cost of communications overhead incurred in the continued update of information for routes which might not be used for a long time if at all is too high. Furthermore, this approach requires more memory due to significant increase in the size of the routing table. These,requirements put limits on the size and density of the network. A third hybrid approach, the Zone Routing Protocol (ZRP), has also been proposed and attempts to reap the benefits of both methods. In ZRP, the network is divided

into zones. A proactive table driven strategy is used for establishment and maintenance of routes between nodes of the same zone, and a reactive on-demand strategy is used for communication between nodes of different zones. This approach can be effective in larger networks with applications that exhibit a relatively high degree of locality of communication, where communication between nodes with close proximity to one another is much more frequent than that between nodes which are farther apart.

Performance parameters:
In order to evaluate the communication performance of the ad hoc wireless network several experimental field trails were performed in an outdoor environment. 1) Route discovery time:- the time needed for the source node to discover a route to the destination. 2) End to end delay:- the delay experienced by a packet from the time it was sent by the source till the time it was received at the destination. 3) Communication throughput:- the useful data rate in bits per second. 4) Packet loss:- the percentage of packet loss during100 ping of data. 5) Route reconstruction time:the time taken upon triggering route reconstruction or repair till completion. end to end delay:- it generally accounts for all delays along the path frm the source to the destination . this includes the transmission delay ,propagation dealy,processing delay and queing delay experienced at every node in the route. The end-to-end delay of a path is the summation of the node delay at each node plus the link delay at each link on the path. Node delay includes the protocol processing time and the queuing delay at node.in wireless link

propagation delays are very less and almost equal for each hop on the path the queing delay and MAC delay are consider as two main factor that accumulated the node delay . the end-to-end delay of a path can be estimated by adding all the node delays and link delays in path. Measure of end-to-end delay for the QoS requirement 250 ms at different node mobility. The end-to-end delay increases as the node speed increases. Higher mobility causes more link broken and frequent re-routing and thus causes larger end to end d The end-to-end delay in EDTORA is within the limit (250ms) and gives up to 60 % improvement. But TORA exceeds the QoS requirementelay.

(b) Effect of number of nodes on end-to-end delay (a) mobility 10 m/s (b) mobility 20 m/s Fig. 7 shows end-to-end delay for number of nodes from 10 to 50 at two different mobile speeds. It increases as the,number of nodes increases because of more number of links. EDTORA shows better performance than TORA for higher number of nodes in both cases. Calculation of end to end delay:dend-end= N[ dtrans+dprop+dproc]

Effect of mobility on end-to-end delay for pause time of 10 s

where dend-end= end-to-end delay dtrans= transmission delay dprop= propagation delay dproc= processing delay dqueue= Queuing delay N= number of links (Number of routers + 1) to measure the end to end delay packets sre sent from source to destination using ping utility over different route lengths and beconing intervals. The EED is taken as on half the RTT. The impact of varying route length and beconing interval on EED discussed as:-

(a)

EED vs. Packet Size

To obtain a comprehensive understanding of the network EED, different ICMP packet sizes were transmitted from the source. In addition, we dis-cuss here the EED at one second beaconing interval. For one-hop links, the packet size was varied from 64 to 48,856 bytes. For twoand three- hop links, packet sizes up to 1448 bytes were used. EED varies with packet size for one-hop routes. It is observed that EED has a linear relationship with increasing packet size over different beaconing intervals. EED is consid-ered to be the sum of transmission, propagation, queueing, and processing time. Therefore, as trans-mission time increases with ICMP packet size, EED increases too. For one-hop routes, EED is 3.25ms when the packet size is 64 bytes and it increases to 340ms at the maximum packet size (48,856 bytes). In the two-hop case (Figure 5), EED displays a sim-ilar monotonic increase as before, with 1.74ms in-crease per 100 bytes. Delay values for the three-hop case are presented . The linear relation-ship of packet size and EED is still true for three-hop routes. Initial delay is at 3.2ms when the packet size is 64 bytes and this increases linearly to 38.3ms at the maximum packet size. In summary, both route length and packet size increases the EED.

packet size) are kept constant. Exceptions to this rule are the extremely high frequency cases of 10ms and 50ms intervals, where EED increases due to increased contention in the channel. beaconing has practically no influence on EED performance for a one-hop route at low beaconing frequencies. However, at high beconing frequencies, EED increases . At two- and three-hops, the impact of high beaconing beaconing on EED is even more significant, Hence, we gather that beacon-ing does not impact EED greatly at low beaconing frequencies.

EED vs. Route Length

EED was previously examined as a result of vary-ing packet sizes or beaconing intervals. Ad hoc net-works are usually multihop, so it is important to evaluate EED with di_erent route lengths. As expected, when route length increases, EED would also increase. EED increases for every additional hop when the same PING packet size is used. We present EED experienced by a sample of packet sizes and at the di_erent route lengths. EED values for the 64- and 1000-byte PING packets are shown in Table 1. From Table 1 we see that a packet propagating through the network experiences approximately the same amount of processing and transmission delay at every node . additional delay is equal to end to end delay experienced by a single hop link for the same packet size. This is true for all packet sizes since EED and packet size have a linear rela-tionship. The results seem to imply that one can support ad hoc routes greater than 3 hops

EED vs. Beaconing Interval

In another investigation, the effect of beaconing fre-quency on EED was examined. Experimental re-sults show that beaconing did not have a significant influence on EED if other network conditions (such as route hop count and

without too much delay. For example, at 10 hops, the EED is predicted to be around 30ms and 100ms for 64 and 1000 bytes respectively.

quency increases, RD time increases slightly. The RD time variation is less than 4ms for all the routes concerned, over the complete range of beaconing in-tervals used in our experiments.

RD time vs. Route Length Route Discovery (RD) Time

Many related work overlooked the need to evaluate route discovery time. This is an important param-eter for ad hoc wireless

networks. When transmis-sion is initiated with an on-demand routing pro-tocol,

the source has to _rst discover the route to the destination node (if there is no previously cached routes) and then initiate the data transmission. Therefore, the time taken to discover a route is critical. In ABR, the route discovery time is measured from the moment when the route search packet (BQ) is sent till the moment when the route reply packet is returned back to the source. Note that since route discovery involves sending control pack-ets only, it is therefore not a_ected by the size of data packets. In a similar fashion, we shall discuss how RD time is a_ected by beaconing and route length.

Unlike beaconing intervals, route length has a significant impact on RD time. Similar to EED, RD time increases as route length increases. As the route search packet transit from one node to an-other, the control packet is intercepted, examined, and processed prior to retransmission to the next node downstream. As route length increases, the route search packet size increases due to the enclosure of intermediate nodes' addresses,associativity and QoS information. Hence, transmission time will gradually increase too. In ABR, if an intermediate mobile node has two neighbors, it will append its ID2 and the corre-sponding associativity metrics for each neighbor be-fore it rebroadcasts the BQ packet. The REPLY packet will contain a chain of IP addresses related to nodes in the selected route. In summary, each hop will introduce a delay comprising of: (a) the fixed transmission and propagation delay of the basic control packet, (b) the variable delay because of the additional information, and (c) the processing time at each node. These components contribute to the RD time.

RD Time vs. Beaconing Interval

Figure 11 presents the experimental results reveal-ing the relationship between RD time and beacon-ing intervals for di_erent route length. The results show that varying beaconing intervals has a weak inuence on RD time. Speci_cally, as beaconing fre-

Conclusion:In this paper, we present the fndings and obser-vations from our experimental research on an ad hoc wireless networking testbed. We have exam-ined how di_erent network conditions (beaconing interval, route length, and packet size) a_ect the communication performance (throughput, end-to-end delay, and route discovery time) of the network. The results obtained indicate that that varying the beaconing interval has very little impact on throughput, delay, packet loss, and route discov-ery time with the

exception of very small intervals. This suggests that beaconing at 1 second may al-low the host to respond quickly in times of mo-bility without jeopardizing communication perfor-mance. Increasing the route length introduces arelatively fixed increase in delay, with slight variation when the packet size is large due to increased packet transmission delay. Although transmitting larger packets inceases communication throughput, the latter saturates due to the constraint of a sin-gle channel and a node has to successfully receive a packet and await for channel access to retransmit the packet downstream. Our experiments, therefore, gave us insights on the communication performance of a current-fo_-the-shelf (COST) ad hoc wireless network implementation, which will assist us in investigating the feasibility of future ad hoc mobile applications.

Refrences
1)C-K. Toh, Guillermo Guichal, and Santithorn Bunchua. `ABAM: On-Demand Associativity-Based Multicast Routing for Mobile Ad Hoc Net-works'. In Proceedings of IEEE International Conference on Vehicular Technology, September 2000 (2)E. Royer and C-K. Toh. A review of current rout-ing protocols for ad-hoc mobile wireless networks. IEEE Network, April 1999 (3) Z.J. Haas. 'A New Routing Protocol for the Re-con_gurable Wireless Networks'. In Proceedings of ICUPC'97, San Diego, CA, October 1997 (4)D. B. Johnson and D. A. Maltz. `Dynamic Source Routing in Ad-Hoc Wireless Networks'. Mobile Computing, ed. T. Imielinski and H. Korth, 1996. (5)S Murthy and J.J. Garcia-Luna-Aceves. `An Ef-_cient Routing Protocol for Wireless Networks'. ACM Mobile Networks and ApplicationsJ., Special Issue on Routing in Mobile Communication Net-works, pages 183{197, October 1996