AN EFFICIENT CLUSTER BASED ROUTING PROTOCOL WITH POWER BALANCING TECHNIQUE FOR POWER HETEROGENEOUS MANET Neenu Sam P.G Student Electronics and Communication Department K.S.R College of Engineering [email protected] Abstract—Power heterogeneous networks are common in MANETs. High power nodes network can improve network scalability, connectivity, broadcasting robustness of the MANETs. Due to the presence of high power nodes unidirectional links are formed which may reduce network throughput. So in order to reduce this issue a routing protocol which avoids packet forwarding through high power nodes is developed. Here a routingaware optimal clustering mechanism that balances power consumption at different cluster heads. For power balancing here we are maximizing the coverage time of the network. The coverage time of the network is defined as the time until one of the cluster head runs out of battery, resulting in an incomplete coverage area. Via simulations the effectiveness of this cluster based routing protocol on improving the performance of power heterogeneous MANETs is demonstrated. Index Terms—Cluster, mobile ad hoc networks (MANETs), power heterogeneous, routing. I. INTRODUCTION IN RECENT years, there has been growing research and analysis in heterogeneous mobile ad hoc networks Vijaykumar B Assistant Professor Electronics and Communication Department K.S.R College of Engineering [email protected] (MANETs). These network consists of devices with heterogeneous characteristics in terms of transmission power [1], [2], energy [3], capacity [4]. An example of such network is the vehicular ad- hoc networks (VANETs). These networks consist of heterogeneous wireless equipment carried by human and machines. In such heterogeneous networks, different devices are likely to have different capacities and thus transmit data with different power levels. In 802.11-based power heterogeneous MANETs, mobile nodes have different transmission power. On one hand, the benefits of high-power nodes are the increase in network coverage area and the reduction in the transmission delay. High power nodes also generally have advantages in power, storage, computation capability, and data transmission rate. As a result, research efforts have been carried out to explore these advantages, such as backbone construction [5] and topology control [6]. On the other hand, the increased transmission range of high power nodes leads to increased interference, which will lead to the reduction in the spatial utilization of network channel resources [7], [8]. Because of different transmission power and other factors like interference, barrier, and noise asymmetric or unidirectional links will exist in MANETs. Many research results show that the performance of routing protocol over unidirectional links in multi-hop network is poor [9]. However, the existing routing protocols in power heterogeneous MANETs will only detect the unidirectional links and avoid the transmissions based on asymmetric links. These protocols will not consider the benefits from high power nodes. Hence, the problem is how to improve the performance of routing in power heterogeneous MANETs, which is the focus of this paper. Expectedly, the clustering paradigm increases the burden on the CHs, forcing them to deplete their batteries much faster than in a non-clustered network [10]. The additional energy consumption is attributed to the aggregation of intra-cluster traffic into a single stream that is transmitted by the CH and to the relaying of inter-cluster traffic from other CHs. Such relaying is sometimes desirable because of its powerconsumption advantage over direct (CHto-sink) communication. Given the high density of nodes in common deployment scenarios the traffic volume coming from a CH can be orders of magnitude greater than the traffic volume of an individual node. Even though the CH may be equipped with a more durable battery than the individual nodes it serves, the large difference in power consumption between the two can lead to shorter lifetime for the CH. Once the CH dies, no communications can take place between the nodes in that cluster and the rest of the network. In this paper an efficient cluster based routing protocol that balances power consumption at cluster head is developed. This routing protocol explicates the benefits of high power nodes by avoiding packet forwarding through high power nodes. Hence the unidirectional links in the networks are avoided and archives maximum throughput. Routing aware optimal clustering mechanism is used for balancing the power consumption at the cluster head. Simulation results show that this cluster based routing protocol achieves much better performance than other existing protocols. The rest of this paper is organized as follows. In Section II, review the related work is added. In Section III, the proposed cluster based routing technique using power balancing technique is discussed. In Section IV, the performance of this cluster based routing via extensive simulations is evaluated and in section V we conclude our discussion. II. RELATED WORK Several routing protocols have been developed in the networking community to address various scenarios. To study the taxonomy of ad hoc routing protocols and to survey the representative protocols [11][13] in different categories many research works has been done in this field. For example, Boukerche et al. [12] gave a comprehensive summary of the routing protocols for MANETs. Unfortunately, most of the existing protocols are limited to homogenous networks and will not perform effectively in power heterogeneous networks. There are some routing protocols for heterogeneous MANETs. Loose virtual cluster based routing protocol for heterogeneous MANET (LRPH) [14] is one of the most recent routing protocols which eliminates unidirectional links and avoids packet forwarding through high power nodes. By using LRPH it is possible to achieve the benefits from high power nodes. Multiclass (MC) [15] is another routing protocol for heterogeneous network. This is a position-aided routing protocol. In MC the entire routing area is divided into cells and a high power node in each cell as the backbone node (B-node). Then, a new medium access control (MAC) protocol called hybrid MAC (HMAC) is designed to cooperate with the routing layer. Based on the cell structure and HMAC, MC achieves better performance. But the fixed cell makes MC to work well only in a network with high density of high-power nodes. In [16], a cross-layer approach is presented that simultaneously extends the MAC and network layers to reduce the problems caused by link asymmetry and exploits the advantages of heterogeneous MANETs. Different from the existing routing on power heterogeneous MANETs, our proposed approach does not rely on geographic information and can be deployed on general 802.11-based mobile devices. It uses a routing aware optimal clustering mechanism for the power balancing in the cluster heads. III. CLUSTER BASED ROUTING WITH POWER BALANCING TECHNIQUE To improve the network performance we propose cluster based routing protocol with power balancing technique for power heterogeneous MANET. This protocol mainly consists of two functions. First is the creation of a cluster which will reduce the network lifetime and increases the efficiency of the network. Second one is the route discovery procedure. This will create route from source to destination by avoiding high power nodes in the network hence eliminate the unidirectional links in the network. In the following, we list the network model and definitions. A. System Model and Assumptions There are two types of nodes in the networks: B-nodes and general nodes (Gnodes). B-nodes refer to the nodes with high power and a large transmission range. G-nodes refer to the nodes with low power and a small transmission range. The numbers of B-nodes and G-nodes are denoted as NB and NG, respectively. There are three type of G-node in the network. Definition 1–Gisolated: They are defined as G-nodes that are not covered by any B-node. Definition 2–Gmember: They are defined as G-nodes whose bidirectional neighbours (BNs) are covered by its cluster head. Definition 3–Ggateway: They are defined as G-nodes whose BNs are not covered by its cluster head. B. Network Model We consider a circular area A of radius R. The sink is located at the centre, as shown in Figure 1. The circular geometry, albeit too idealistic, serves as a basis for understanding the intrinsic tradeoffs involved in a joint clustering and routing framework. The nodes are uniformly distributed across A with density ρ. Data captured and generated by all sensors need to be delivered to the sink. Due to energy considerations, only those nodes within the area {(x, y)/ x2 + y2 ≤ r02}, where r0 < R, can communicate directly with the sink; all other nodes are organized into clusters and they communicate their data through their respective CHs. Without loss of generality, we assume that each CH is located at the centre of its cluster. The procedure for cluster formation consists of two steps: the deployment of CHs and the assignment of nodes to CHs. Because of the symmetric nature of the area A and the uniform distribution of nodes, the formation of clusters is also symmetric, i.e., any two clusters with the same distance from their centres to the sink should have the same coverage. Such clusters are said to be of the same type. Suppose there are K types of clusters in the network. We consider the following clustering approach: sensors whose distances to the sink fall in (ri−1, ri] are organized into clusters of the ith type, where 1 ≤ i ≤ K and r0 < r1 < . . . < rK = R. As a result, the clusters of the ith type cover the ith ring, defined by the area{ (x, y)/ r2 i−1 < x2 + y2 ≤ r2 i }. This clustering approach can be easily realized in practice, e.g., by using pilot signals that are broadcasted by the sink. Accordingly, the CHs of the ith ring are placed evenly along the circle {(x, y)/ x2 + y2 = d2i} with equal space between consecutive CHs, where {di = ri−1+ri2 }. We assume that the initial battery energy at all CH is same. A node located in the ith ring is assigned to the nearest CH in the same ring. In the analysis, we assume that a sufficiently large number of CHs are placed in each ring such that the area covered by each CH can be approximated by a small circle, as shown in Figure 1. In the simulations section, we show that this assumption has a negligible impact on network performance. Remark: Although our model assumes a circular sensing area and a two-tier network structure, the analysis adequately captures the intrinsic interaction between inter and intra-cluster Figure 1: Network topology with 3 rings (k=3) traffic. The analysis can be extended to handle a non-circular region by covering it with a series of circles, similar to the approach used in cellular networks (in cellular networks, the region is approximately covered by hexagons). A multi-layered organization of sensors, such as the “spine” hierarchy can also be accommodated in our analytical framework. C. Traffic Model Each node generates data at a rate λ (in bits/second). The data are transmitted from the source node to its CH, which then forwards the data to the sink, directly or through other CHs. We assume that each sensor has sufficient power to communicate directly with its CH. Furthermore, we assume that the CH depletes its energy at a much faster rate than the nodes it serves. This assumption is justified by the low data rate and duty cycle of commonly used nodes. Accordingly, we focus our attention on energy depletion at CHs. From a strategic point of view, a CH is more critical to the coverage of the network than individual sensors. D. Shortest-Distance Relay In this model, packets are relayed through the closest CH of the adjacent ring. More specifically, a CH in the ith ring receives traffic originating from its own cluster as well as traffic relayed from CHs in the (i + 1)th ring, and forwards the combined traffic to the closest CH in the (i − 1)th ring. Traffic relaying continues hopby-hop until the sink is reached. For this scenario, we consider a routing-aware clustering mechanism that balances power consumption at different CHs. Clearly, the radius profile of the clusters, {1/2 (r1 − r0), ..., 1/2 (rK − rK−1), is critical to power consumption at different CHs. For example, reducing 1/2 (ri − ri−1) results in smaller clusters in the ith ring, which leads to less local traffic from these clusters, shorter transmission distances to subsequent CHs in the (i − 1)th ring, and a higher number of CHs in the ith ring. Because of the symmetry in the topology and traffic load, the traffic from the CHs in the (i + 1)th ring will be evenly shared among an increased number of CHs in the ith ring, so the volume of the relayed traffic carried by individual CHs in the ith ring will decrease. All of these factors contribute to reduced power consumption at the CHs in the ith ring. On the other hand, the reduction in the area of the ith ring must be compensated for by other clusters (e.g., the clusters in the jth ring), because of the fixed number of rings in the system. In an analogous manner, power consumption at CHs in ring j will increase. Therefore, by deliberately adjusting the cluster size in different rings, a more balanced power consumption at different CHs, and hence an increase in the coverage time, is achieved. E. Route Discovery Procedure When a source node S wants to send a data packet to destination node D, S will first searches whether there is any route to D exists in its route cache. If there is a route, S will directly send the data packet. Otherwise, S initiates the route discovery procedure to find a route to D. When a node obtains a complete source route to D, it responds with a RREP packet to S directly and tells S about the discovered route. Here bidirectional links will be used because the RREP packets are sent by unicast scheme. However, there exists unidirectional links between B-node and G-node in the discovered path. Consequently, those unidirectional links must be repaired. In addition, by analysing the fact that transmitting through B-nodes can dramatically decreases the throughput of a network, this scheme will try to avoid B-nodes in the path by replacing B-nodes with multi-hop G-nodes. The disadvantage is that this scheme may increase route hops and hence the increases delay of the route discovery procedure but network throughput can be ultimately improved. Fig. 2 shows an example to illustrate how to process the RREP packet. Assume that the route to destination D is S → · · ·→a → B1 → c → d · · ·→D, where B1 → c is a unidirectional link. Therefore, when c receives the RREP packet, it replaces the route a → B1 → c in the RREP packet with multi-hop G-nodes a → m → n → c. Thus S receives a bidirectional route to D with less number of B-nodes as possible. A timer can also be set to initiate a new round of route discovery until the timer expires, and there 350 Throughput 300 250 200 150 LRPH 100 ECRPH 50 0 0 Maximun node speed (m/s) Figure 2: Route Discovery Procedure IV. SIMULATION RESULTS The set of simulations are used to evaluate the performance of existing system (LRPH) and the proposed cluster based routing with power balancing techniques (ECRPH) under different mobility by varying the node’s speed from 0 to 20 m/s. The simulations run on a network with 35 G-nodes and 5 B-nodes in an area of 1500 × 1500 m2. RG and RB are 200 and 600 m, respectively. All sources transmit packets of 512 B at the rate of four packets per second. Figure 3: shows the throughput of two schemes under different node mobility. Several observations are done. First, the throughput of two schemes decreases as the node speed increases. This can be explained as follows: A higher mobility causes more broken links and transmission failures. Second, the throughput of proposed technique is higher than that of LRPH because life time of the network is increased due to the power balanced scheme. Figure 3: Throughput Vs Maximum Node Speed (m/s) Energy Consumpyion Rate is no response from the RREP packets. If the route discovery fails for number of times, data transmission will be cancelled. 5 10 15 20 0.02 0.015 0.01 LRPH 0.005 ECRPH 0 0 5 10 15 20 Maximum node speed (m/s) Figure 4: Energy Consumption Rate Vs Maximum node speed (m/s) Figure 4: shows the ECRP of LRPH and the proposed cluster based routing with power balanced technique (ECRPH) under different node mobility. There are few observations. First, The ECRP of two schemes increases as the node speed grows because more energy should be used for retransmission and rerouting. The proposed system shows reduced energy consumption rete then energy consumption rate than LRPH. Packet Delivery Ratio(%) REFERENCES 96 95 94 93 92 91 90 89 88 87 [1] Y. Huang, X. Yang, S. Yang, W. Yu, and X. Fu, “A cross-layer approach handling link asymmetry for wireless mesh acces networks,” IEEE Trans. Veh. Technol., vol. 60, no. 3, pp. 1045–1058, Mar. 2011. [2] V. Shah, E. Gelal, and P. Krishnamurthy, “ Handling asymmetry in power heterogeneous ad hoc networks,” J. Comput. Netw.—Int. J. Comput. Telecommun. Netw., vol. 51, no. 10, pp. 2594–2615, Jul. 2007. LRPH ECRPH 0 5 10 15 20 Maximum Node Speed (m/s) Figure 5: Packet Delivery Ratio(%)Vs Maximum Node Speed (m/s) Figure 5: shows the packet delivery ratio for the LRPH and proposed cluster based routing with power balanced (ECRPH) scheme. Normally the packet delivery ratio increases as node speed increases due more link breaks. Here for the proposed scheme packet delivery ratio is higher than the existing LRPH due to increased throughput of our proposed power balanced technique. V. CONCLUSION In this paper a cluster based routing protocol with power balancing technique is developed. This protocol will reduces the power consumption at cluster head and also increases the coverage time of the network hence making the cluster head active for more time. The clustering scheme used here will eliminate unidirectional links and hence provide benefits from high-power nodes such as transmission range, processing capability and bandwidth. The routing schemes will forward data packet by avoiding high power nodes. Through extensive set of simulations, the effectiveness of proposed scheme over power heterogeneous MANETs is demonstrated. [3] W. Liu, C. Zhang, G. Yao, and Y. Fang, “Delar: A device–energy–load aware relaying framework for heterogeneous mobile ad hoc networks eous networks,” IEEE J. Sel. Areas Commun., vol. 29, no. 8, pp. 1572–1584, Sep. 2011. [4] P. Leonardi, E. Garetto, and M. Giaccone, “Capacity scaling in ad hoc networks with heterogeneous mobile nodes: The super-critical regime,” IEEE/ACM Trans. Netw., vol. 17, no. 5, pp. 1522–1535, Oct. 2009. [5] J. Wu and F. Dai, “Virtual backbone construction in MANETs using adjustable transmission ranges,”IEEE Trans. Mobile Comput., vol. 5,no. 9, pp. 1188– 1200, Sep. 2006. [6] A. A. Jeng and R.-H. Jan, “Adaptive topology control for mobile ad hoc networks,” IEEE Trans. Parallel [7] Distrib. Syst., vol. 22, no. 12, pp. 1953–1960, Dec. 2011. survey,” IEEE Veh. Technol. Mag., vol. 2, no. 2, pp. 12–22, Jun. 2007. X. Zhang, Q. Gao, J. Zhang, and G. Wang, “Impact of transmit power on throughput performance in wireless ad hoc networks with variable rate control,” Comput. Commun., vol. 31, no. 15, pp. 3638–3642, Sep. 2008. [14] Peng Zhao, Xinyu Yang, Wei Yu, and Xinwen Fu, “A Loose-VirtualClustering-Based Routing for Power Heterogeneous manets” vol. 62, No. 5, June 2013 [8] J. Liu, X. Jiang, H. Nishiyama, and N.Kato, “Exact throughput capacity under power control in mobile ad hoc networks,” in Proc. 31th IEEE INFOCOM, Mar. 2012, pp. 1–9. [9] M. K. Marina and S. R. Das, “Routing performance in the presence of unidirectional links in multihop wireless networks,” in Proc. ACM MobiHoc, Lausanne, Switzerland, Jun. 2002, pp. 12–23. [10] Tao Shu, Marwan Krunz, Sarma Vrudhula, “Power Balanced Coverage-Time Optimization for Clustered Wireless Sensor Networks” May 25–27, 2005 [11] A. Boukerche, Algorithms and Protocols for Wireless, Mobile Ad Hoc Networks. Hoboken, NJ: Wiley, Nov. 10, 2008. [12] A. Boukerche, “Performance evaluation of routing protocols for ad hoc wireless networks,” Mobile Netw. Appl., vol. 9, no. 4, pp. 333– 342, Aug. 2004. [13] F. Li and Y. Wang, “Routing in vehicular ad hoc networks: A [15] X. Du, D. Wu, W. Liu, and Y. Fang, “Multiclass routing and medium access control for heterogeneous mobile ad hoc networks,” IEEE Trans. Veh. Technol., vol. 55, no. 1, pp. 270–277, Jan. 2006. [16] S. Yang, X. Yang, and H. Yang, “A cross-layer framework for positionbased routing and medium access control in heterogeneous mobile ad hoc networks,” Telecommun. Syst., vol. 42, no. 1/2, pp. 29–46, Oct. 2009.
© Copyright 2026 Paperzz