VIEWS: 21 PAGES: 6 CATEGORY: Research POSTED ON: 6/17/2010
UBICC, the Ubiquitous Computing and Communication Journal [ISSN 1992-8424], is an international scientific and educational organization dedicated to advancing the arts, sciences, and applications of information technology. With a world-wide membership, UBICC is a leading resource for computing professionals and students working in the various fields of Information Technology, and for interpreting the impact of information technology on society.
Link analysis on Indoor and Outdoor environment for Mobile Ad Hoc Network using Random Way point Mobility Model and Manhattan Mobility Model Ibrahim khider ,Prof.WangFurong.Prof.YinWeiHua Ibrahim khider,Huazhong University of Science and Technology Communication Software and Switch Center,Dept of Electronic and Information Systems,Wuhan 430074,P.RChina E_Mail: Ibrahimchina2008@yahoo.com,WangFurong@mail.hust.edu.cn,yinyinwh@163.com ABSTRACT Mobility and connectivity metrics in Mobile Ad hoc NETwork (MANET) is the most important topic. Each MANET mobility and connectivity metrics has its own strengths and weakness. It turns out the impact of mobility model has become a critical issue in the research of performance comparison. In this paper, we provide an a study of the link analysis in indoor and outdoor environment in which mobile nodes moves according to random waypoint mobility model (RWP) and Manhattan mobility model (MH) respectively. The analysis includes the mobility and connectivity metrics in environment model. We provide in-depth investigation such as the impact of speed, number of nodes and simulation time on link, and how parameters of combined mobility model influence link dynamics. Our study about the link analysis gives us an insight into the network topology dynamics of MANET, and it provides us the basis of further analysis of MANET such as network connectivity. Simulation results verify our analysis of the link of MANET in indoor and outdoor environment. Keyword: mobility model, mobility metrics, connectivity metrics, environment, Mobile Adhoc Networks, Simulation. 1 INTRODUCTION particular environments in the realistic world are A Mobile Ad hoc NETwork (MANET) is considered studied. Accordingly, two environment-aware to be an autonomous system of self-organized mobile mobility models are introduced and simulated. The nodes without relying on any infrastructure. Node Random Waypoint model is used to model the mobility is one of the key characteristics of MANET, movement in buildings in the simulation area. The and it is also one of the critical factors that have Manhattan model can be used to construct streets significant influence on the performance of MANET such as in a city area. The remainder of this paper is protocols, mainly the routing protocols. organized as follows. The two mobility models Conventional mobility models proposed for MANET (Manhattan and RWP) are overviewed in section2 can be classified into two categories: Entity model and 3 respectively. Section4 depicts a combine and Group model. Entity models are used to mobility model in indoor and outdoor environment. represent the movement of an individual mobile node. Section 5 describes mobility and connectivity Among Entity models, the Random Waypoint model metrics Description of the intensive simulations and (RWP) [1] is the most popular model used in this a study of the results are given in section 6 and 7 field. However, the interaction among the mobile respectively and the last section presents the nodes cannot be reflected by Entity models. Group conclusion. models are therefore proposed. A typical mode is the 2-Random Waypoint Reference Point Group Mobility (RPGM) model [2]. Random Way Point mobility model (RWP) [3] [4] A major drawback of conventional models is that [5] [6] is a simple, widely used, model in the many some environment factors such as spatial constraints, simulation studies of ad hoc routing protocols. In this speed limits, etc are ignored. Street traffic system model each node is assigned an initial position could be an example environment. Cars are moving uniformly distributed within a region (rectangular along the roads and choose one way out if a junction region). Then, each node chooses a destination is met. The people follow the routes to building, uniformly inside the region, and selects a speed spend some time there then go out from one of the uniformly from [minspeed, maxspeed] independently room exits. Moreover, in some certain environment of the chosen destination. That means the such as arts exhibition, the destinations of visitors are distributions of nodes’ speeds and locations are not random, but more or less deterministic in that stationary. To avoid the transient period from the they always visit some places more attractive to them. beginning, one solution is to choose the nodes’ initial These mobility scenarios cannot be handled properly locations and speeds according to the stationary by most of existing models. In this paper, two distribution; another one is to discard the initial time Ubiquitous Computing and Communication Journal 1 period of simulation to reduce the effect of such transient period on simulation results. The node then Current State/ moves toward the chosen destination with the MH state RWP state selected speed along a straight line starting from Condition current waypoint. After reaching the destination, the node stops for duration called “pause time”, and then repeats the procedure. All nodes move independently Condition indoor RWP RWP of each other at all times. 3- Manhattan Mobility Model Condition outdoor MH MH The MH model is used to emulate the nodes movement on streets defined by maps [4] [5]. The Fig.1 State Transition table of Random Waypoint map is composed of a number of horizontal and and Manhattan Mobility Model vertical streets. Each street has two lanes, one in each direction (North and South for vertical streets, and East and West for horizontal ones). Each node is only allowed to move along the grid of horizontal and vertical streets. At an intersection of horizontal State Transition and vertical streets, a mobile node can turn left, or Condition Indoor right, or go straight with probabilities 0.25, 0.25, and environment 0.5, respectively. The speed of a mobile node is RWP temporarily dependent on its previous speed If two mobile nodes on the same freeway lane are within the Safety Distance (SD), the velocity of the following node cannot exceed the velocity of Outdoor preceding node Mobility models capture the Indoor geographic restrictions. The speed of a node s (t) is updated according to: s (t + 1) = min (Smax, max (0, s (t) + a (t) * X)) where X Uniform [−1, 1], and a (t) is Acceleration Speed. Outdoor Action environment 4- DESCRPTION OF COMBINED MOBILITY MODEL Transition Manhattan This section introduces a combined model of Manhattan and Random Waypoint, the movements of a node switch from one mobility model (Manhattan or Random Waypoint) to another based on its location in the network. When the nodes are on the street, they move as Manhattan mobility model Fig.2 State Transition Diagram of Random Waypoint movement pattern, when they are located in the and Manhattan Mobility Model building, they will move as Random Waypoint model. The movement nodes are divided in two In this paper, we limit the study to an urban area groups depending on their speed a “pedestrian” modelized by a Manhattan and RWP mobility model. group with a low speed and a “vehicular” group with The area is wrapped around North-South and West- a higher speed. The pedestrian group of users is East and the grid is composed of 3 by 3 buildings. moving with a normal distributed speed with a mean The buildings are 300x300 m and the street has two of 3 km/h and a standard deviation of 0.3 km/h [4]. opposite lane, the distance between lane 1 m and the The vehicular group of users has also a normal width of lane 6 meter. Fig.2 shows the layout of distributed speed but with a mean of 50 km/h and a indoor and outdoor environments in urban area and standard deviation of 2.5 km/h. At each cross-road, the movement of nodes. Fig.3 shows the movement users of both groups have can either continue straight of nodes in simulation area using manhattan and with the probability Pr (straight) = 0.5 or turn Random Waypoint mobility models. left/right with the probability Pr (right) = Pr (right) = 0.25.To represent the movement of mobile nodes in outdoor environment (streets) and indoor environment (buildings) we have used Finite State Machine (FSM) to explain the movement of mobile nodes by using state diagram and state transition table as shown in fig.1 and TABLE.1, . Ubiquitous Computing and Communication Journal 2 Outdoor environment ⎧ T X (i, j , t ) ⎪∑t =1 LC (i. j ) ifLC ≠ 0 Indoor indoor Indoor LD = ⎨ ........... (4) environment environment environment ⎪ ⎩∑t =1 X (i, j , t )otherwise T Outdoor environment 3- Route based Metrics a metrics listed bellow used Indoor Indoor RWP Indoor to measure routing protocol performance: - environment environment environment -Number of link change LC (i, j): Number of link changes for a pairs of nodes i and j is a number of times the link between them down to up. Average number of link changes LC (i, j) averaged over all node pairs is given by: Indoor Indoor Indoor environment environment LC (i, j ) = ∑t =1 c(i, j, t ) ………… (5) environment T 4- Density-based Metrics Fig.3 Layout of Urban environment These metrics are mobility independent, but does give an important feedback about the local 5-Overview of Mobility and Connectivity Metrics environment. Node Density is the number of Many metrics have been used with the aim to adapt neighbors seen by the considered node. Average routing protocols to the possible changes in the number of nodes that stay within a node radio range network topology due to node mobility [4]. In this at a time t is given by the following equation: section we will present a variety metrics have been used in my work, and can help measuring link ∑ ∑ T N N (i, t ) analysis and the performance in environment model. ND = T =1 i =1 ……..(6) Figs.2, 3 show a typical movement scenario in TN indoor and outdoor environment that we will Where, N is number of nodes, T is simulation time, principally use in our simulation tests. and N(i,t)is the number of neighbor nodes for node i 1- Speed-based Metrics: - at time t. - Average Relative Speed between all nodes (RS) 6- Simulation Description is the average magnitude of relative speed of two A variety of matrices have been used for the nodes over all neighborhood pairs and all time. RS is MANET environment. In this section, we study the given by: most popular metrics, relative speed, spatial r r dependence, temporal dependence, number of link RS (i, j , t ) = Vi (t ) − V j (t ) ……… (1) change and node degree in the environments where -Spatial Dependence: the average function value MH and RWP exist. Our evaluations are based on of angle of Relative velocity of two nodes over all the simulation using Network Simulator (NS-2) neighborhood pairs and all times, the following environment with CMU wireless adhoc networking equation shows Degree of spatial dependence: extension [7] and we extract the useful data from r r r r trace file using mobility trace analyzer tool (version Dspatial(i, j, t) = RD vi (t),v j (t))* SR(vi (t),v j (t)).(2) ( 1.0 beta)[8], then the graphs are generated using RD is relative direction and SR is speed ratio Matlab.Simulation environment consists of 40 - Degree of temporal dependence: shows the wireless nodes forming an ad hoc network, moving similarity of node’s velocities at different times [1], over a 1000 X 1000 flat space, DSR routing protocol and it is given by: for 900 seconds of simulated time. Each run of the r r r r Dtemporal(i, t, t′) = RD(vi (t),vi (t′))* SR(vi (t),vi (t′)) (3) simulator accepts as input a scenario file that 2- Link-based Metrics: describes the exact motion of each node and the The idea is based on using link information’s exact sequence of packets originated by each node, experienced by every node. There exists a link together with the exact time at which each change in between two nodes, if they are in each other motion or packet origination is to occur. We have transmission range. The most used for link-based generated different scenario files with varying Metrics are: movement patterns and traffic loads (CBR), the - Link Duration (LD): the duration of one link is traffic consist of cbr type with 40 connections, data calculated as the time that two nodes are within rate 1 packet/sec, packet size 512 byte and the transmission rang of one other .LD is given by: transmission range 250m. And then ran against each of these scenario files. When Nodes in street they are move according to manhattan model otherwise they move as Random Waypoint model .The movement Ubiquitous Computing and Communication Journal 3 scenario files we used for each simulation are characterized by a number of nodes and simulation 4.8 times. Each simulation ran for 900 seconds. We ran 4.6 our simulations with movement patterns generated for 4 different numbers of nodes, 10, 20, 30, and 40, 4.4 with constant speed and 40 nodes with different 4.2 simulation time and different maximum speed. In Relative Spead simulation, we chose to make evaluation according 4 to the following metrics: Relative Speed, Spatial 3.8 Dependence, Temporal Dependence, Link Duration, Link Change and Node Degree. 3.6 7- Simulation Results 3.4 The simulation results bring out some important characteristic differences between mobility and 3.2 connectivity metrics. We conducted our simulations 3 on changing the parameters for mobile nodes' movement scenarios and their connection pattern 2.8 0 10 20 30 40 50 files. We supposed different simulation for Max Speed (m/sec) "movement scenarios" files. Figures below Fig.4, Relative Speed Vs Speed demonstrate the simulation results by applying different maximum speed, simulation time of mobile nodes and different number of nodes. Fig.4 shows 0.25 that the relative speed is high when maximum speed increase with different maximum speed .the values of spatial dependence is low as shown in fig.5 fig.6 0.2 shows that the values of temporal dependence decrease when the maximum speed increase. The Spatial Dependence link duration and node degree are high as shown 0.15 fig.7 and 8 and this lead to increasing the throughput accordingly fig.9 and 10 show the effect of number of node in changing link and node degree, when 0.1 number of node increase the link change and node degree increase.fig.11 shows that the temporal dependence increase exponentially when the time 0.05 increase. We can conclude that with the different maximum speed spatial and temporal dependence are low, relative speed and link duration are high. We 0 observe that by increasing number of nodes the link 5 10 15 20 25 30 35 40 45 50 Max Speed (m/sec) change and node degree almost increasing linearly. The temporal dependence was effect by time Fig.5, Spatial Dependence Vs Speed simulation results agree with expected results based 1.25 on theoretical analysis as much as possible. This metric is also studied in 1.2 various communication and mobile nodes movement scenarios. 1.15 Temporal Dependence 1.1 1.05 1 0.95 0.9 5 10 15 20 25 30 35 40 45 50 Max Speed (m/sec) Fig.6, Temporal Dependence Vs Speed Ubiquitous Computing and Communication Journal 4 50 1.005 1 48 0.995 Temporal correlation 46 0.99 Link Duration (sec) 44 0.985 42 0.98 0.975 40 0.97 38 0.965 100 20 300 400 500 600 700 800 900 36 8- CONCLUSION 0 Simulation time (sec) 5 10 15 20 25 30 35 40 45 50 Max Speed (m/sec) Fig.7, Link duration Vs Speed Fig.10, node degree Vs #mobile nod 2.8 40 2.6 35 2.4 Node Degree 2.2 30 node degree 2 25 1.8 20 1.6 1.4 15 5 10 15 20 25 30 35 40 45 50 10 Max Speed (m/sec) Fig.8, Node Degree Vs Speed 5 10 15 20 2 30 35 40 800 # Mobile5nodes (node) 700 Fig.11, temporal Dependence Vs time # links change (link) 600 8. CONCLUSION 500 The area of ad hoc networking has been receiving 400 increasing attention among researchers in recent years, as the available wireless networking and 300 mobile computing hardware bases are now capable of supporting the promise of this technology We did 200 investigation on MANET in urban area and the effect of mobility in this environment .In this paper we 100 introduce urban environment which include indoor and outdoor environment with combine mobility 0 10 15 20 25 30 35 40 model (RWP and MH), in addition we investigate a # Mobile nodes (node) set of metrics to capture the characteristics of mobility then we implemented a combined mobility Fig.9, #link change Vs #mobile node model in the IMPORTANT framework in NS-2 and we used a framework to analyze the link .Our study Ubiquitous Computing and Communication Journal 5 indicates that the mobility and connectivity metrics are useful to capture and understand the mobility characteristics , we observe that some scenario improves with add velocity, number of nodes and simulation time. Also we observe that the metrics does influence the performance in environment model in urban area. Our study has shown that the simulation results are highly dependent on the movement behaviours of mobile node and simulation environment. ACKNOWLEDGEMENT This work is supported by National Natural Science Foundation of China under Grant NO 572047. 9- References [1]. T. Camp, J. Boleng, and V. Davies, "A Survey of Mobility Models for Ad Hoc Network," Research in Wireless Communication and Mobile Computing (WCMC): Special issue on Mobile Ad Hoc Networking: Research, Trends and Applications, vol.2, no. 5, pp. 483-502, 2002. [2]. X. Hong, M. Gerla, G. Pei, and C. Chiang. "A group mobility model for ad hoc wireless networks," In Proceedings of the ACM International Workshop on Modeling and Simulation of Wireless and Mobile Systems (MSWiM), August 1999. [3] J. Yoon, M. Liu, and B. Noble, “Sound Mobility Models,” in Proc. ACM/IEEE Int’l Conf. Mobile Computing and Networking (MOBICOM ’03), pp. 205-216, September2003. [4] F. Bai, N. Sadagopan, and A. Helmy, “IMPORTANT: A framework to systematically analyze the Impact of Mobility on Performance of RouTing protocols for Adhoc NeTworks,” in IEEE INFOCOM’03, March/April, 2003. [5] F. Bai and A. Helmy, “A Survey of Mobility Modeling and Analysis in Wireless Adhoc Networks”, chapter , University of southern California, USA. [6]W. Navidi and T. Camp, “Stationary Distributions for the Random Waypoint Mobility Model,” IEEE Transactions on Mobile Computing ITMMC-3 (2004), pp. 99-108. Adhoc Networks,” Book Chapter in submission to Kluwer Academic Publishers [7] http://nile.wpi.edu/NS/ [8] http://nile.usc.edu/important/software.htm Ubiquitous Computing and Communication Journal 6