Mobile Adhoc Networks are self-organizing and self-configuring multi-hop wireless networks capable of adaptive re-configuration as effected by node mobility. Typically the network is made up of equal nodes that are equipped for wireless communication and with networking capability. Every node in the network is capable of functioning as a mobile router (i.e., maintain routes & forward packets), which makes possible the multi-hop forwarding of packets from a source node to a destination node without reliance on a fixed infrastructure. All nodes share the same random access wireless channel.
Interference and Noise Signal and actual Signal Strength parameters are available and can be measured at physical layer of OSI reference architecture. Signal to Interference noise ratio (SINR) is calculated based on using such measured parameters. SINR parameter is cross-layered to the routing layer in order to compare it with the threshold value. The corresponding threshold value maintained in the routing layer is Signal to Interference Noise Ratio Threshold (SINRT). SINR is compared with SINRT and if it is less than the threshold value, then that packet in AODV is rejected. Otherwise the packet is processed further. This will enable to form the reliable links in AODV as the weaker signals are discarded and hence the stronger links are made to persist.
Keywords- Mobile Ad hoc networks, Cross layer designs, Routing, Noise, Signal to Noise Ratio.
Mobile ad hoc networks are collection of portable wireless nodes that are self organizing and rapidly deployable in which neither wired backbone nor centralized control exists. These multi-hop networks eradicate the need of infrastructure setup and administration, which allow anywhere, anytime network connectivity and make them suitable for different applications ranging from military to commercial opearations. They are very useful in disaster management, rescue operations, defense missions, and emergency meetings. In fourth generations(4G) mobile systems and beyond, it is strongly anticipated that ad hoc networks will be used to economically extend the coverage and capacity of existing GSM, UMTS, and Wi-Fi networks.
Signals in the process of communication always pick up some undesired signals. Indeed, any type of processing performed on signal tends to introduce these undesired disturbances which we call noise. Noise is thus an undesired signal which is not connected with the desired signal in any way. The power supply hum in radio receiver, the oscillations in the feedback system etc are noise signals. They can be predicted and eliminated by proper design .
Computation of interference and noise at each receiver is a critical factor in wireless communication modelling, as this computation becomes the basis of SINR (Signal to interference and noise ratio) or SNR (signal to noise ratio). The power of interference and noise is calculated as the sum of all signals on the channel other than the one being received by the radio plus the thermal (receiver) noise. The resulting power is used as the base of SNR, which determines the probability of successful signal reception for a given packet .
This paper addresses the interference and noise present at the physical layer of the OSI layered architecture and mechanism to select the proper signals, discarding the noisy signals. The paper achieves this by means of cross layer design and modifying the AODV routing protocol with Signal to Interference Noise Ratio (SINR) as the cross layer parameter. The simulation results shows that there is considerable improvement in 'Throughput', 'Packet delivery Ratio', 'overhead' and 'End to End Delay'.
II. Cross Layer Designs
A. SINR AND SINRT
In Cross layer design, the SINR is calculated at the physical layer and is made available to top layers (MAC). By means of a derived data type, we cross layer the SINR parameter from physical layer to MAC layer. SINR parameter thus cross layered is then compared with the threshold value in order to select the signals of interest and discard the unwanted signals which whose signal strength is less than the SINRT (threshold) value.
Fig.1 Information sharing in Cross Layer Design.
B. Comparison of SINR with SINRT
Interference and Noise is inherent part of the communication which can be measured at the physical layer of the OSI layered architecture. Interference noise can be due to electrical disturbance, thermal due to temperature variations, etc. If the interference noise level is higher, it may be possible for the communication devices to falsely include them as the message signal and thereby causing the under performance of the overall higher layer protocols. This project aims at minimizing such effect of interference noise on the network layer protocol performance by using the cross layer design.
Signal to noise interference ratio is calculated in the physical layer of the OSI protocol layer stack. When noise due to interference appears at the physical layer, it may some times be interpreted as the signal, leading to the under performance of the protocol at the higher layer. Higher the signal to interference noise ratio, greater the probability of the signal to be the message signal or the signal desired. Usually this ratio is measured in terms of decibels.
Signal Power = Pt * Gt * Gr * (lambda^2)
(4 * pi * d)^2 * L
Where pt is the transmitted power, Gt and Gr are the antenna power gain of the transmitter and receiver respectively, L is a constant which is a system loss and lambda is the wavelength.
Noise Power = K*B*T
Where K is a Boltz man's constant which is 1.3807*10-23, T is a ambient temperature in degree Kelvin and B is the bandwidth.
SINR = Signal Power/Noise Power
Signal to interference noise ratio (SINR) which is calculated at the physical layer measures how much is the level of signal power with respect to the noise power. It is cross layered to the higher layer (MAC). This cross layered parameter is then used at the routing layer to compare with the predetermined SINRT (threshold) value.
Signal to Interference Noise Ratio (SINR) thus calculated at the physical layer is thus cross layered to the higher layer for making the decision about the routing of the messages. SINR is compared with the SINRT (Signal to interference noise ratio threshold). If the received signal to interference noise ratio is greater than the threshold value, the received signal is considered as the desired signal and AODV protocol is made to process the message. If the received signal to interference noise ratio is less than the threshold value, it is discarded and not processed further.
Ad -hoc on demand distance vector protocol by default processes all the messages which are given by the lower layers (MAC sub layer). Since we introduce the computation of SINR at the physical layer and cross layering the parameter at the network layer, the processing of messages by the AODV protocol is now selective. We compare the parameter(SINR) received from the physical layer by cross layer design with the threshold value (SINRT), AODV is made to process only those messages which are having greater signal to interference noise ratio when compared to the threshold value.
Since the processing of messages is selective and there exists clear separation of real message with the interference and noise, we envisage overall improvement in performance of the protocol with respect to throughput, delay, packet delivery ratio and overhead. This design ensures the separation of noise signal from the message signal, thereby making the selective processing of messages by the protocol at the higher layers.
We present the steps involved in the implementation of the project requirement by making required changes.
At physical layer, calculate signal to interference noise ratio.
Using cross layer design, pass SINR parameter to network layer.
At network layer, compare SINR with SINR Threshold (SINRT)
Is SINR is less than the threshold value, discard the AODV packet.
If SINR is equal or greater than SINRT, then process the AODV packet.
III. Simulation Environment and Result Analysis
We used NS2, a scalable network simulator in our evaluation.
We consider the network to be consisting of several nodes ranging from 50, 75, 100, 125, and 150. For each of this nodes selection, simulation is run and the results are obtained. We set the network channel to wireless channel and the propagation model to Two Ray Ground. Network interface is set to wireless physical. Since we need to choose the proper MAC layer protocol, Mac 802.11 is chosen in our case. Interface queue is considered to be that of Drop tail's priority queue. Considering the Link layer protocol, Omni directional antenna is taken into consideration with X dimension of topography as 1000 and Y dimension of topography as 1000. Base routing protocol is considered to be that of AODV which is modified as Cross Layer AODV (CLAODV). Simulation time is taken as 50 and flow varied from one value to another value with specific range.
We considered AODV as the base protocol from which the further changes are made as per the project requirements. The modified protocol is considered as CLAODV which implements the changes. Requirement implementation is put in CLAODV containing modules which are modified according to the requirement.
Major changes to the AODV protocol is in the routing table entries. Since the SINR parameter is cross layered from physical layer to the MAC layer, routing table entry is modified to contain cross layered parameter. Mac layer related modules are accordingly modified to reflect the cross layering of parameter considered. Calculation of signal power and interference noise power and its ratio will be carried out in wireless physical layer module of the protocol which is placed at the bottom of the protocol stack.
The number of nodes for the protocol input is varied from 50 to 150 and accordingly, the protocol performance is looked at. The constant bit rate source at the rate of 100kb and packet size of 512 is considered for the protocol input.
NS2 is considered as the appropriate platform to execute the project since it has all the in-built functionality which is necessary for the execution of the network projects. Front end is to be implemented using tcl and back end to be in C++ language which is object oriented. Required modules in C++ are modified to implement the project and the number of nodes and new protocol selection is done through the front end tcl.
Fig.2 Number of Nodes v/s Throughput.
Above graph shows that there is an increase in throughput for CLAODV when compared to AODV because of selective processing of signals.
Fig.3 Number of Nodes v/s Delay.
As the above figure indicates, there is substantial decrease in the end to end delay when compared to normal AODV.
Fig.4 Number of Nodes v/s Packet Delivery Ratio.
Packet Delivery ratio is higher in modified AODV protocol when compared to normal protocol due to the elimination of noisy signals.
Fig.5 Number of Nodes v/s overhead.
Over head becomes lesser in CLAODV when compared to normal AODV protocol.
Noise is an inherent part of the communication system. Successful signal processing depends on how well the noise is handled and discarded, separating it from the actual signal. Using cross layer design, we can selectively process the messages by comparing the SINR with the SINRT. SINR is present and measured at the physical layer which is cross layered to the Network layer by means of a cross layer design. By doing so, we envisage improvement in the protocol performance with respect to Throughput, End to End delay, Overhead and Packet Delivery Ratio. In our future work we try to combine the path loss with the SINR comparison to further improve the performance.
Mohammad Ilyas, "The hand book of ad hoc wireless networks", CRC press, 2003
Communication Systems, B P LATHI
Mineo Takai, Jay Martin, Rajiv Bagrodia, UCLA Computer Science Department, Los Angeles, "Effects of Wireless Physical layer Modelling in Mobile Ad hoc Networks"
Charles E Perkins, Sun Microsystems Laboratories, Elizabeth M Royer, Dept of electrical and computer engineering, "Ad hoc on demand Distance Vector Routing"
R Challappa Doss, L. Pan, School of Engineering and Information Technology, Deakin University, Australia, "A case for Cross Layer Design: The impact of Pysical Layer Properties on Routing Protocol: Performance in MANETs"
B Ramachandran and S Shanmugavel, Department of ECE, SRM University Chennai, "Received Signal Strength based Cross Layer Designs for Mobile Ad Hoc Networks"
Zhi Ren, Jing Su, Wei Guo, School of Communication and information engineering, China, "A Cross Layer AODV Routing Protocol"
Sanjay Shakottai, Theodore S Rappaport and Peter C Carlson, "Cross Layer Design for Wireless Netwroks"
Michael Gerharz, Christian D Waal, Peter Martini, Paul James, Institute of computer science, University of Bonn, Germany, "Strategies for finding Stable Paths in Mobile Wireless Ad hoc Networks.
Vineet Srivastava, Institute for Infocomm Research and National University of Singapore, Mehul Motani, National Univeristy of Singapore, "Cross Layer
Design: A survey and the Road Ahead"
Marco Conti, Gaia Maselli, Giovanni, Turi, IIT-CNR, Silvia, Giordano, SUPSI, "Cross Layering in Mobile Adhoc Netwrok Design"
NS Manual and NS Doc