Efficient and Secure Dynamic Source Routing For Reliable Data Transmission
(Paper ID: 24ET3003201506)
Abhinav Pareek Rahul Agarwal Vinay Baghela
Department of CSE and IT, Department of CSE and IT, Department of CSE and IT, Govt. Engineering College Bikaner Govt. Engineering College Bikaner Govt. Engineering College Bikaner
[email protected] [email protected] [email protected]
Abstract: Mobile ad hoc networks have various kinds of security difficulties, caused by their nature of collaborative and open systems and by limited availability of resources. In our work we look at DSR in detail, study and analyses various attacks that can be possible on it. My proposed work is an extension to DSR of the secure DSR protocol extension, which includes tuning strategies aimed at improving its performance. In MANET’s, the routing protocols find routes based on the shortest hop count. Due to There is a possibility that the selected route may be unreliable in terms of stability, blocking and high interference, if the network is loaded with heavy traffic. Mobility may also high in such random selected routes. We have proposed a routing protocol that uses the link quality facts that is calculated using the received signal’s noise value to choose the routes during route discovery process. The routing protocol is based on a reactive route discovery process related to Dynamic source routing protocol. Our proposed method is efficient and secure Dynamic source (ES-DSR) protocol.
The noise of a received signal calculated at the physical layer is used during the route discovery process to either select or reject a link to take part in the route discovery. Because it is, the possibility that a route will break due to the low quality of an intermediate link which decreases the network performance in terms of network throughput.
I am using the end-to-end delay, first packet received and routing throughput as the metrics for performance evaluation for our proposed routing solution. During the simulations it has been seen that the selection of high quality links based on the noise greatly improves overall network performance when it is measured with different mobility and load networks throughout the simulation process. Different routing protocols were studied, and their effects were elaborated by stating how these routing interrupt the performance of MANET. Simulation is done in exata cyber 2.0 verson.
Keywords: Noise, SINR, MANET, Routing Protocols, ad-Hoc networks, Mobility.
I. INTRODUCTION
Mobile Ad-hoc Networks are wireless networks scenario which are deployed without a predefine Structure, which are Usually collected on a temporary basis to serve a specific deployment purposes like in emergencies such as natural hazards rescue or battlefield communication[1] .It provides a consistent communication in situations where the
deployments of infrastructure based system is not cost effective or Unfeasible.
IN MANET’s, the routing protocols find routes based on the shortest hop count metric. That means the route with the lowest length in terms of number of hops is selected for routing[2].Due to this the possibility of the selected route may be unreliable in terms of stability, blocked & high interference if the network is loaded with heavy traffic[8].
Mobility may also be high in such random selected routes.
Therefore, the route discovery process of MANETs should be changed in such a way that it can discover routes in terms of other metrics that are more reliable and suitable other than the lowest hop count during the route discovery and selection process[3][4].
1) Reactive Protocols: Reactive protocols identified as on demand driven reactive protocols. The statement they are recognized as reactive protocols is, it do not start route discovery by them, when they are requested, when a source node requests to find a route. These protocols arrange routes when demanded [3]. When a node desires to communicate with extra node in the network, and the source node does not have a path to the node it require to communicate with, reactive routing protocols will create a path for the source to destination node.
2) Hybrid protocols: take advantage of both the proactive and the reactive approach. Distributed dynamic routing algorithm and zone routing protocol are most well known proactive protocols [8].
3) Proactive Protocols: each node maintains information enabling it to decide how to route messages towards any other node in the network. Such information is usually stored in a certain number of tables (updated) providing each node with a view of the network topology. Destination- Sequenced Distance Vector [3] and Optimized Link State Routing [4] are the most well-known proactive protocols.
II. OUR CONTRIBUTIONS
The main contribution of this work is to find routes between a given source-destination pair that are reliable and stable enough to efficiently transmit data traffic in moderate mobility and congestion scenarios. In this paper, we propose
and implement a noise value based DSR routing protocol (ES_DSR), that discovers routes with higher stability during the route discovery process. To make sure that the routes discovered during the route discovery process are of good quality and secure [1] [3]. We make sure the discovered routes contain radio links that are of high quality from the existing radio channels in the network.
Fig. 1. Design of the Proposed Efficiency-Based DSR Approach
III. PROPOSED ES-DSRAPPROACH
Figure 1 shows the overall design of the proposed Efficiency- based DSR Approach [2] [5].In the existing routing methodology the routes are selected based upon the shortest route between the source and destination which is not an efficient way to select the routes because of the possibility that the selected route may be unreliable in terms of stability, blocked &high interference if the network is loaded with heavy traffic. Mobility may also high in such random selected routes [3].
Therefore, to keep this in mind we have developed a new routing protocol that uses the Noise value information of a link to decide whether to include or not the current link in the route discovery process. The results obtained from various simulations shows the effectiveness of our proposed approach.
The functionality of our proposed design is divided in two main phases
1) Route Request propagation Phase: In this, we try to discover route for a given destination in such a way that the exposed route is consists of the links that has high efficiency (lower Noise) values so that the probability of their breakage is lower during the long communication process duration. To achieve this we use an ES-based route discovery process as described in the algorithm and the flowchart [1] [2]. In this phase, a route from the destination to the source is created upon which the route reply message will travel from destination to source in unicast form.
2) Route Reply propagation phase: The above first phase is completed once the RREQ message is received by the destination node. When destination nodes receives several RREQ messages for the same communication flow. It uses one link to reply with the RREP message [3]. That one link is again chosen as described in the algorithm and the flowchart.
We make sure that the RREP message sent by the destination node will flow the reverse route that is created during the RREQ propagation phase. When the RREP is received by the source node we have an ES-based good quality forward route from the source node to the destination node.
In the proposed ES-DSR protocol, each node selects a radio link with the highest efficiency value from the available set of neighbor links it has during the route discovery process between a source-destination pair. An info field called N_info is added in the RREQ (Route Request) message of DSR routing protocol to pass the Noise value from the physical layer of the received RREQ message to the network layer using a cross-layer design technique [5]. The value of the N of a received message is calculated using its received SINR (Signal to Noise Ratio) which is the signal-to noise and interference ratio of a signal. The SINR is calculated as a ratio of received signal power to the interference power of interfering signals in addition with the noise received. The information about the efficiency value of the received RREQ message is then used by the network layer during the route discovery process to select the route that consists with the links that has the highest efficiency as compared to the other routes available between the source destination pair for which the route discovery phase is initiated.
Two additional data structures that are used in the implementation process of our proposed ES-DSR protocol.
The first data structure is a modification in the DSR protocols RREQ control message. In this, we add an INFO field that is added by the physical layer whenever it receives a RREQ message [6]. In the INFO field the physical layer adds the N information of the received RREQ message which is then extracted from it when this RREQ is reached at the network layer. The second data structure is the RREQ_BUFFER_TABLE that is created at each node during the route discovery process. It has following fields namely:
1) Source address 2) Flooding_Id
3) RREQ_msg
The data structures used to create the RREQ_BUFFER table and fields added in the RREQ message of DSR protocol are discussed above. The INFO field is added in the RREQ messages info field because the RREQ message fields cannot be accessed at the physical layer as they are created at the network layer [4]. Therefore, the info field is used to store the value of the calculated N of the RREQ in its info field.
The Algorithm 1 is elaborated implemented in the of Exata simulator in order to see the effectiveness and correctness of our proposed routing method using the simulation results generated during the simulation processes. Let’s assume that the S is the source and D is the destination node in the network for which a route has to be discovered using our proposed ES-DSR routing protocol. When a source node S got a data packet for transmission to a destination node D, it checks its routing table for an active route for the destination [5].
If an active route is available in the routing table of node S the data packet is transmitted towards the node D using the next hop address given in its routing table. On the other hand, if the routing table of node S does not have a valid entry for the node D.it starts our proposed route discovery process using the ES-DSR routing protocol. The node S initiates the RREQ message and broadcast it in the underlying network. When an intermediate node which is also the neighbor node of the node S in this case receives the RREQ message at the physical layer it calculates the Noise value of the received RREQ message. To compute the Noise as mentioned that SINR is used at the physical layer in which the intermediate node will use the ratio of the received signal power and the noise and interference in the received signal.
Then noise value is converted into efficiency value [7].
The calculated efficiency value of the received RREQ message is then placed in the INFO field of the RREQ message and that RREQ message is then send to the MAC layer from where it will reach to our ES-DSR protocol which is at the network layer. Upon reception of the RREQ message the node will buffer the RREQ message in the buffer implemented by us at each node in the network. With the insertion of the RREQ message in the buffer the node also starts a timer of 10 micro seconds.
Until this timer expires, this node will buffer all the duplicate RREQ messages it receives from its neighbor nodes. When the timer associated with RREQ storing buffer the node will sort the RREQ messages in the buffer with the RREQ message that has the highest efficiency value is placed in top of the RREQ buffer. Once sorted the node will extract the top RREQ message and process it. The node discards all the remaining RREQ messages in the RREQ buffer. This process in done on all the intermediate nodes until the RREQ message reaches to its destination node [5].
When a destination node receives the RREQ message it also stores it in the RREQ message buffer and wait for the timer to expire. When the timer expires, the destination node extract the RREQ message from the RREQ buffer with the lowest N.
The node from which the selected RREQ message is received will be considered as the next hop node towards the source of the received RREQ message [2][4]. This next hop node Variable used in the Algorithm:
‘S: Source node, ‘D’: Destination node
‘Int_node’: Intermediate node
‘R_buf’: RREQ message buffer
‘N_rreq’: N of received RREQ message
‘RT ’: Routing table of a node
‘Tx’: Timer of R_buf at node x if S got data packet for D then
if S not have route for D in its RT then
S starts the ES-Dsr protocols route discovery process;
else
S send packet to next-hop towards destination node D;
if Receive a fresh RREQ or duplicate message then Physical layer of Int_node calculates the sinr and max hope value and add it in the INFO field and hash field with RREQ message;
Int_node store the RREQ message in its R_buf and set the timer if it’s fresh RREQ;
When Ti expire node Int_node extract the RREQ with the highest noise;
Node Int_node rebroadcast the extracted RREQ and discard the R_buf;
end
if D receives the RREQ message then
Int_node store the RREQ message in its R_buf and set the timer if it’s fresh RREQ;
When Ti expire node Int_node extract the RREQ with the highest SNR ;
D gets the pervious hop address of the extracted RREQ message;
D creates a RREP message and sends it towards S using the previous hop selected in last step;
End
if S receives the RREP message then S updates its RT and sends the buffered data packet to D;
end end end
ES-DSR Routing Protocol is then used by the destination node to forward the RREP message towards the source node.
In this way our proposed route discovery algorithm always selects the best quality radio link at each intermediate node during the route discovery process for the destination node and the destination will use the same route for the RREP
message forwarding. Therefore the highest quality route is discovered and used by the source node for data communication process [6].
IV. EXPERIMENTAL RESULTS
We present the performance analysis and impact analysis of our implemented ES-DSR protocol on different scenarios over MANETs. To perform all the simulations, we created the scenarios using well knows network simulator called EXata.
The results obtained on various scenarios when DSR routing protocol and our proposed efficiency based DSR routing (ES- DSR) protocol is compared with each other on three different network layer metrics.
A. Simulation Model
In our simulation process we uses the mobility model used in infrastructure-less mobile networks known as Random way point mobility model. In this model, the speed of a node is randomly chosen before moving to the target point is between 0 m/s to 25 m/s. The pause time is set to 10 seconds. We use a terrain with dimensions 1200m x 1200m to randomly deploy 50 nodes in it. The nodes in the network are configured with 802.11a/g MAC specification and their transmission power is 250 meters which is calculated using a nodes transmitting power. All the source-destination pairs are selected randomly in from the network [7]. To model the source nodes as a data generating nodes we configure each source node in the network using the constant bit rate (CBR) application. The CBR generates data according the following given parameters:
i) Inter-packet time: 33 milliseconds.
ii) Packet size: 512 bytes.
iii)Intervals: the starting and stopping time of the CBR sessions.
Each node stores the received data packets into its output buffers during its wait for a route for destination node. All packets (either data packets or control messages) sent by the routing layer are stored in the packet queue which is implemented as a buffer until they are extracted from the buffer by MAC layer to transmit them to the physical layer.
Routing packets are given higher priority than data packets in the buffer. All the simulations performed in this paper run for a time period equal to 400 simulated seconds. Each data point shown in the graphs and tables are represent an average of three runs with similar traffic models, but different randomly generated mobility scenarios by using different seed values.
B. Performance metrics
The following metrics are used in varying scenarios to evaluate the three different protocols:
1) Throughput it is measured by the total amount of packets which is received by a destination node. It is measured by byte/sec or bit/sec. High throughput is always expected for any routing protocol [8].
2) Average end-to-end delay of data packets: This metric is calculated by the destination node whenever it receives a data packet. The destination node will calculate the delay of each received data packet by using its send timestamp and its received timestamp at the destination. At the end of the simulation the total time of the data packets received at the destination is divided to the total number of received data packets. We calculate average end-to-end delay for packets received by each destination node as follows:
3) First Packet Received:-Network load denotes the total load in bit/sec submitted to wireless LAN layers by all higher layers in all WLAN nodes of the network [9].
C. Simulation Results
In order to compare and evaluate performances of the DSR and ES-DSR protocols in different network conditions, two parameters are varied in the simulations: Maximum mobility of the nodes and Number of data sessions.
At first, simulations are carried out by keeping the number of sources constant in our case we fix it to 6 source destination pairs and varying the mobility in the network. 6 sources are modeled respectively to study the effect of varying mobility in network. In the other scenario, the number of sources is varied from 3 to 15. When varying the number of sources, node’s mobility is kept random which falls between the range of 0 to 10 m/s and pause time is set to 30 seconds. The effects of change in network mobility are analyzed on traditional DSR protocols and our proposed efficiency based DSR routing (ES-DSR) protocol [10]. The mobility of the nodes is changed by increasing the range by 5 m/s in each simulation.
In figure 2 we had plotted Average End-to-End Delay with increase in network mobility, the end-to-end delay of the data sessions are fluctuating at some mobility points and the general trade shows that the end-to-end delay increasing with the increases in the network mobility because as the mobility.
This is because as the mobility increases the number of routes that are broken during the communication process also increases which increase the network mobility.
Network Mobility (data rate) Fig. 2: End To End Delay N/W Mobility
In figure 2 i had plotted Average End-to-End Delay (EED) with increment in network mobility, the end-to-end delay of the data sessions are fluctuating at some mobility points and the general trade shows that the end-to-end delay increasing with the increment in the network mobility because as the mobility. This is because as the mobility increases, the number of routes that are broken during the communication process also increases, which increase the network mobility.
This is Fig. 2: Average End-to-End Delay (EED) with increase in network mobility due to the fact that increment in the mobility means that the intermediate nodes on an active route can move from the routes which cause the route breaks.
As it can be observed that from the figure 2 that the end-to- end delay of my proposed protocol is lower at low and moderate mobility networks because of the selection of routes that are consists with the links that has high lifetime this decreases the number of route breaks during the communication, which decreases the eed[11].
Fig. 3: Average Throughput With Increase In Network Mobility
In figure 3 i had plotted Average Throughput with increment in network mobility, the Throughput of the network decreases with the increment in the network mobility. This is due to the increment in the route breaks during the communication process the number of packets that are on an active route, which is broken and lost. As it can be saw from the figure 3 that the throughput of my proposed ES-DSR is much higher as compared to the traditional DSR routing protocol because of the lower number of re-routing processes caused due to the selection of high quality route selection process that i have implemented in suggested ES-DSR routing protocol
Fig. 4: First Packet Received With Network Mobility
In figure 4 I have plotted first packet received with increment in network mobility which highlights the effect on routing time for DSR and ES-DSR protocols with the increase in network mobility. As it can be observed from the figure 4 that the first packet received increases with the increment in the network mobility. As it can also be observed from the figure 4 that the first packet received of my proposed ES-DSR is lower than the DSR protocol because of the same reason i have given for the throughput increment that the ES_DSR has lower number of re-routings due to its link-quality based route selection process.
Fig. 5: avgas end to end delay with n/w load
In Figure 5 i have plotted the Average End-to-End Delay with increment in network load, the end-to-end delay increases with increment in network load for both the comparing protocols. This is because as the network load increases the congestion in the network increases with these two things happens [12] Firstly, it increases the number of route failures in a data communication session. Secondly, the time taken to transmit a data packet to next hop could increment due to increase in the number of re-transmissions required at the MAC layer due to the congestion or contention on the transmitting link. Although, the end-to-end delay of the proposed ES-DSR is not as high as the DSR because it selects routes which has higher received signal power which makes the selected link more reliable as compared to the links that are selected if DSR routing protocol is used for route discovery.
Fig. 6: average throughput with increase in network load
In Figure 6 i had plotted the throughput with increment in network load, the throughput of the network decreases with the increment in network load because the increment in network load increases the contention and congestion in the selected routes[13]. Due to this, the active communication routes breaks during the data communication and the data packets on these routes are dropped during the route discovery process.
Fig. 6 throughput with increment in network load in network traffic because the newly added flow might use a common link for routing. Although, the throughput of my proposed ES-DSR is high even in high load networks because it avoids the use of highly loaded links during its route discovery process by not selecting the links with high interference [14].
Fig. 7: First Packet received With Network Load
The first packet received of the network with the increment in network load for DSR and ES-DSR protocols is shown in Figure 7. It can be observed from the Figure that the routing overhead of our proposed ES-DSR is much better than the FPR of the traditional DSR protocol. On the other hand, the first packet received in the DSR routing protocol increases rapidly with the increment in network load because of the same comments and reasons that are provided above to
support the increment in network EED with the increased network load [15].
V. IV.CONCLUSION
In this paper, we have proposed an efficient and secure route discovery process that uses the link quality under consideration at each step during its reactive route discovery process. Our proposed ES-DSR protocol shows that it is more effective for data transmission in moderate mobility and congested network than the traditional routing protocols in MANETs. Our proposed efficiency based DSR routing protocol uses the received signal power, interfering signal power and noise over a link to identify whether it is a stable radio link or not during the route discovery process.
We have analyzed our proposed work with the help of simulation results that are generated using the network simulator called Exata. The results are generated on large number of scenarios with various parameter values to show its effectiveness in all kinds of situations. Therefore at the end the path which is selected for the data transmission is the one which consists of radio links that has lower interference and low noise in conjunction to high received signal strength.
REFERENCES
[1] H. vijayvergia; A. chaudhary; G. Singh; R. S. Shekhawat;”bit-error rate based adaptive Routing Protocol for Reliable Datatranmission on manet”IEEE international conference on signal propagation and computer and technology, (IEEE_ICSPT 2014).
[2] Hamad, S.; Noureddine, H.; Al-Raweshidy, H., "LSEA: Link Stability and Energy Aware for efficient routing in Mobile Ad Hoc Network,"14th International Symposium on Wireless Personal Multimedia Communications 2011(WPMC’11), pp.1-5, 3-7 Oct.
2011.
[3] Charles E. Perkins and PravinBhagwat. Highly dynamic destinationsequenceddistance-vector routing (dsdv) for mobile computers. SIGCOMMComput.Commun. Rev., Vol. 24, PP. 234 - 244, OctoN 1994.
[4] P. Jacquet, P. Mhlethaler, T. Clausen, A. Laouiti, A. Qayyum, and L.Viennot.Optimized link state routing protocol for ad hoc networks.PP.3626, 2001.
[5] Al-Akaidi, M.; Alchaita, M., "Link stability and mobility in ad hocwireless networks," IET Communications, vol.1, no.2, pp.173-178, April 2007.
[6] D. Johnson, Y. Hu, and D. Maltz, âA˘ IJThe Dynamic Source Routing(DSR) protocol for mobile ad hoc networks for IPv4,â˘A˙IRFC 4728,Feb.2007.
[7] Peng Yang, "QoS Routing Protocol Based on Link Stability with Dynamic Delay Prediction in MANET," Pacific-Asia Workshop on Computational Intelligence and Industrial Application, 2008. , vol.1, pp.515-518, 19-20 Dec. 2008.
[8] Chakrabarti, S., Mishra, A.: QoS issues in ad hoc wireless networks.Communications Magazine, IEEE Vol. 39, PP. 142â˘A
¸S148, 2001.
[9] Wen-Fong Wang; Po-Hun Shih, "Study on an Enhanced Link Stabilitybased Routing Scheme for Mobile Ad hoc Networks," 3rd AnnualIEEE Communications Society on Sensor and Ad Hoc Communicationsand Networks, 2006(SECON’06. 2006), vol.3, pp.797-802, 28-28 Sept.2006.
[10] Hui Zhang; Yu-ning Dong, "Mobility Prediction Model Based LinkStability Metric for Wireless Ad Hoc Networks," International Conferenceon Wireless Communications, Networking and Mobile Computing, 2006(WiCOM’06). pp.1,4, 22-24 Sept. 2006.
[11] Yafeng Zhou; Sang-Hwa Chung; Lihua Yang; "A Link- Quality Aware Routing Metric for Multi-hop Wireless Network," International Conference on Communication Software and Networks, 2009(ICCSN’09). pp.390-394, 27-28 Feb. 2009.
[12] Jenn-Hwan Tarng; Bing-Wen Chuang; Fang-Jing Wu, "A Radio- LinkStability-based Routing Protocol for Mobile Ad Hoc Networks,"
IEEE International Conference on Systems, Man and Cybernetics,2006(SMC’06). , vol.5, pp.3697-3701, 8-11 Oct. 2006.
[13] Effatparvar, M.R.; Yazdani, N.; Lahooti, F.; EffatParvar, M., "Link Stability Approach and Scalability Method on ODMRP in Ad Hoc Networks," Communication Networks and Services Research Conference, 2009(NSR ’09). pp.416-421, 11-13 May 2009.
[14] LiminMeng; Wanxia Wu, "Dynamic Source Routing Protocol Basedon Link Stability Arithmetic,"International Symposium on Information Science and Engineering, 2008(ISISE ’08), vol.2, pp.730, 733, 20-22 Dec. 2008.
[15] Du, H.; Hassanein, H.; ChihsiangYeh, "Zone-based routing protocolfor high-mobility MANET," Canadian Conference on Electrical andComputer Engineering, 2003(IEEE CCECE 2003), vol.2, pp.1055-10584-7 May 2003.
[16] Wu, C.; K.; Kato,"AMANET protocol considering link stability and bandwidth efficiency", International Conference on Ultra Modern Telecommunications & Workshops, 2009(ICUMT ’09), pp.18, 12-14 Oct. 2009.
