Dynamic Traffic Congestion Control Scheme in MANET with Cooperative Communication

Transcription

1 Dynamic Traffic Congestion Control Scheme in MANET with Cooperative Communication P.Deepa 1, Dr. P.S.K Patra 2 1 Department of Computer Science and Engineering, 2 Head of the Department, Agni College of Technology, Anna University, Chennai ,Tamil Nadu,India deepa.mecse@act.edu.in, csehod@act.edu.in Abstract Cooperative communications enables efficient utilization of system communication resources, thereby allowing nodes or terminals participating in a communication network to collaborate with each other during information transmission. Cooperative communication schemes have advantageous in improving node capacity and diversity Cooperative Communication (CC) is a technology that allows multiple nodes to simultaneously transmit the same data. It is also used in improving enhancing power, network connectivity, improving communication reliability and spectrum efficiency. The three basic cooperation protocols in cooperative relaying are decodeand-forward (DF) and compress -and-forward (CF) and amplify-and-forward (AF ). But these existing works on cooperative communications are mainly focused on the link level physical layer issues. In this paper a brief analyze is made on Capacity-Optimized Cooperative (COCO) a novel topology scheme to improve the network capacity in MANET by considering both the upper layer network capacity and the physical layer cooperative communications. Keywords: cooperative communication, COCO, MANET, network capacity, relay nodes. 1. Introduction A MANET is a type of ad-hoc network that can change locations and self configuring. These networks can exist without a fixed infrastructure i.e., infrastructure less and they can work in an autonomous manner. As nodes are mobile, the connection link between two devices can break depending on the spatial orientation of the mobile nodes. The mobile wireless devices which are out of communication range can use the other devices within their communication range to forward the packets. Thus reliability can be ensured in MANET. Cooperative communication has derived an interest for wireless network. Cooperative communications refer to a type of communication system or technique that allows users to transmit each other's messages to the intended destination. Cooperative communication typically refers to a system where the users can share and coordinate their resources to improve the information transmission quality. Recently, cooperative wireless communication has received tremendous interests as an untapped means for improving the performance of information transmission operating over the ever challenging wireless medium. The demand for the speed in the wireless networks is also rapidly increasing. Most existing works on cooperative communications are focused on link level physical layer issues. Direct transmissions and multi hop transmissions can be regarded as special types of the cooperative transmissions. A direct transmission uses no relays, while a multi hop transmissions does not combine signals at the destination. Most existing works on cooperative communications are focused on link-level physical layer issues. Consequently, the impacts of cooperative communications on networklevel upper layer issues, such as topology control, routing and network capacity, are largely ignored. In this article, we propose a Capacity-Optimized Cooperative (COCO) topology control scheme to improve the network capacity in MANETs by jointly considering both upper layer network capacity and physical layer cooperative communications. Through simulations, we show that physical layer cooperative communications have significant impacts on the network capacity, and the Page 534

2 proposed topology control scheme can substantially improve the network capacity in MANETs with cooperative communications. The relay nodes play a vital role in cooperative communication. The possible ways of realizing cooperation are using extra relay nodes (RNs) to assist the communications between sources and their corresponding destinations. And the next is to allow the communication nodes in a network to help each other in order to communicate with their corresponding destinations. The Systems which using the first way of cooperation is often referred to as relay systems. The systems using the second way of cooperation are often referred to as cooperative systems. The cooperative communication mainly deals with the physical layer issues. Cooperative communication typically refers to a system allows users to share and coordinate their resources in order to enhance the data transmission quality. This is a generalization of the relay communication, where multiple sources also serve as relays for each other. Due to the lack of centralized control, MANETs nodes cooperate with each other to achieve a common goal. The major activities involved in self-organization are neighbor discovery, topology organization, and topology reorganization. Network topology describes the connectivity information of the entire network, including the nodes in the network and the connections between them. Topology control is very important for the overall performance of a MANET. The impacts of cooperative communications on network-level upper layer issues, such as the topology control, routing and network capacity are largely avoided. In Figure 1, S indicates source node, R is the relay node and D is the destination node. This diagram represents the simplest cooperative scenario having only three nodes. In time slot1, the source transmits data to the destination node. At the same time the relay overhears the transmission from source to destination. Then in the time slot2, the relay forwards the message to the destination which uses both of the received messages and jointly decodes the data through maximal ratio. 2. The topology control problem in MANET Topology control have been widely studied and applied in wireless ad hoc networks. It is considered as one of the key energy saving technique. In order to extend lifetime and save energy of networks topology control, each wireless node to select certain subset of neighbors or adjust its transmission power at the same time to maintain network connectivity. Chen and Huang first made analyzes in strongly connected topology control problem, which aims to find a connected topology such that the total energy consumption is minimized. NP complete is proved to be one of such problem. Several following works have focused on finding the minimum power assignment so that the induced communication graph has some good properties such as connectivity or fault-tolerance and disjoint paths. On the other hand, several geometrical structures have been proposed which are used as underlying network topologies. These geometrical structures are usually kept as few links as possible from the original communication graph and can be easily constructed using location information. Recently, a new class of communication techniques, cooperative communication (CC) has been introduced to allow single antenna devices to take the advantage of the multiple-input-multiple-output (MIMO) systems which is shown in fig 2. This cooperative communication explores the broadcast nature of the wireless medium and allows nodes that have received the transmitted signal to cooperatively help relaying data for other nodes. Recent study has shown significant performance gain of cooperative communication in various wireless network applications: energy efficient routing and connectivity improvement. Fig 1: simple cooperative communication Fig 2: multiuser MIMO system Page 535

3 The cooperative communication techniques can also be used in topology control. The topology control problem under cooperative model, aims to obtain a stronglyconnected topology with minimum total energy consumption. NP-complete is the first problem and then proposed two algorithms that start from a connected topology assumed to be the output of a traditional (without using CC) topology control algorithm and reduce the energy consumption using CC model. The first algorithm (DTCC) uses 2-hop neighborhood information of each node to reduce the overall energy consumption within its 2-hop neighborhood without hurting the connectivity under CC model. The second algorithm (ITCC) starts from a minimum transmission power, and it repeatedly increases its power until all nodes within its 1-hop neighborhood are connected under CC model. Finding that the CC technique can also extend the transmission range and thus link disconnected components. In mobile ad hoc wireless communication, each node of the network has a potential of varying the topology through the adjustment of its power transmission in relation to other nodes in the neighborhood. In contrast, wired networks have fixed established pre-configured infrastructure with centralized network management system structure in place. Therefore, the fundamental reason for the topology control scheme in MANET is to provide a control mechanism that maintains the network connectivity and performance optimization by prolonging network lifetime and maximizing network throughput. As topology control is to determine the existence of wireless links subject to network connectivity, the general topology control problem can be expressed as G* = arg max f(g), (1) s.t. network connectivity. The problem Eq. 1 uses the original network topology G, which contains mobile nodes and link connections, as the input. According to the objective function, a better topology G*(V, E*) will be constructed as the output of the algorithm. G* should contain all mobile nodes in G, and the link connections E* should preserve network connectivity without partitioning the network. The topology structure after getting the result is strongly related to the objective function optimization, which is f(g) in Eq. 1. It is difficult to collect the entire network information in MANETs. Therefore, it is desirable to design a distributed algorithm, which generally requires only local knowledge, and the algorithm is run at every node independently. Consequently, each node in the network is responsible for managing the links to all its neighbors only. If all the neighbor connections are preserved, the end-to-end connectivity is then guaranteed. Given a neighborhood graph GN (VN, EN)with N neighboring nodes, we can define a distributed topology control problem as G*N = arg max f(gn), s.t. connectivity to all the neighbors. The objective function f(g) in Eq. 1 is critical to topology control problems. Network capacity is an important objective function. Our previous work shows that topology control can affect network capacity significantly. In the following section, we present a topology control scheme with the objective of optimizing network capacity in MANETs with cooperative communications. 2.1 Topology sensing Fig 3: Topology Control The topology sensing is to make the nodes get to know the topology information of the network, which includes link sensing, neighbour detection and topology discovery. This part gets benefit from OLSR to minimize the flooding of broadcast packets in the network by reducing duplicate retransmissions in the same region. An adaptation of OLSR for the multipath routing is that in the TC message, the protocol not only include the links between local node and MPR, but the links to all the neighbours so that each node can have better information about the networks topology to construct disjoint multipath. In the simulation, it works well at low data rate. But in the scenario of high data rate, it will cause more congestion because this method will increase the size of TC message. 2.2 Routes Computation Contrary to classical OLSR, routes are not renewed each time a node receives a new routing message, but in an ondemand scheme, in order to avoid the loud computation of several routes for every possible destination. When a given source must send packets, the route computation procedure uses the algorithm. 2.3 Route Recovery In OLSR, the hop-by-hop routing is used, which means when a packet reaches an intermediate node, the protocol will check the routing table of the local node and then forward the packet to the next hop. It will help the source Page 536

4 node keep good control of the packets which will be forwarded in the multipath. However, in the mean time, the pure source routing might cause two problems: Firstly, the information in the source node might be not new enough because it needs time to flood the topology control messages to the whole network. It means when computing the routes, the source node might use the links that does not exist anymore. Secondly, even when the information in the source node is updated, the topology might change during the forwarding of the packet. Both of them will cause the failure of the packets forwarding. To solve these problems, the route recovery is used: before a medium node trying to forward the packet, the node first check if the next hop in the source route of the packet is one of its neighbours. If yes, the packet is forwarded as it should be. If no, then it s possible that the next hop has moved out of the transmission range of the node. Then it is necessary to recompute the route and forward the packet through the new route. basic cooperation protocols are amplify-and-forward (AF), decode-and-forward (DF), and compress -and-forward (CF). For the amplify-and-forward protocol, each receiver node simply scales its received signal according to its transmit power constraint and forwards the scaled signal in the next transmission slot. The decode-and-forward protocol, each RN decodes the source message from its received signal, re-encodes it into a new codeword, and transmits it in the next transmission slot. For the compress- maps its received and-forward protocol, each RN first signal into another signal in a reduced signal space, then encodes and forwards the compressed signal as a new codeword by taking the signal received at the destination as side information. AF and CF based cooperation schemes can be viewed as analog cooperation schemes. DF based cooperation schemes can be viewed as digital cooperation schemes. Depending on the network topology and the quality of the backhaul link between the source and the RN, one protocol may outperform the other in terms of system capacity or diversity. 3.2 Phases of Cooperative Transmissions Fig 4: Architecture design of Topology Control with cooperative communication In figure 4, topology control scheme such as COCO topology control is used for achieving the quality of service(qos) and balances the traffic loads. The cooperative communications with MANET achieved by two strategies, one is Amplify-and-forwardd and other one is Decode-and-forward. These two methods are used to reduce the interference and increase the connectivity of wireless networks. 3. Background 3.1 Relay nodes In a relay system, sources first transmit their data to the receiver nodes. Each receiver node then processes and forwards its received data information to the destination nodes following some cooperation protocols. With the received signal from the Receiver nodes, the destinations decode the data from their corresponding sources. Some Most cooperative communication schemes involve two transmission phases [1] Phase 1 refers to a coordination phase. This is the phase where users exchange their own source data and control messages with each other and/or the destination. Phase 2 refers to a cooperation phase. This is the phase where the users cooperatively retransmit their messages to the destination. In Phase I, the source user broadcasts its data to both the relay and the destination In Phase II, the relay forwards the source s data either by itself or by cooperating with the source to enhance reception at the destination 4. COCO: Capacity Optimized COoperative topology The COCO (Capacity -Optimized Cooperative topology control scheme) extends the physical layer cooperative communications from the link-level perspective to the network-level perspective in MANETs. The Energy Efficient path is chosen between the source, the relay and the destination nodes. Then the path is checked for the less number of interference on the relay nodes using the COCO Topology scheme. The main work of the COCO topology scheme is to establish a path via relay nodes with less number of interferences. The extensive research has been done on cooperative communications; most existing works are focused on Page 537

5 physical layer issues, such as decreasing outage probability and increasing outage capacity, which are only link-wide metrics. However, from the network s point of view, it may not be sufficient for the overall network performance, such as the whole network capacity. Therefore, many upper layer aspects of cooperative communications merit further research, e.g., the impacts on network structure and topology control, especially in mobile ad hoc networks (MANETs). Indeed, most current studies on MANETs attempt to create, adapt, and manage a complex network based on traditional simple point-topoint non-cooperative wireless links. Considering upper layer network capacity and physical layer relay selection, this paper proposes a Capacity- Optimized Cooperative (COCO) topology control scheme for MANETs with cooperative communications. Most existing topology control schemes assume that the wireless channel is wellknown and organized. But, in practice, it is difficult to have the good knowledge of a dynamic channel. A good topology is the one that can optimize the global network performance while preserve some global graph property (i.e., connectivity). This paper presents a novel Capacity-Optimized CO-operative (COCO) topology control scheme for MANETs with cooperative communications. Most existing topology control schemes assume a perfect known wireless channel. However, in practice, it is difficult to have the perfect knowledge of a dynamic channel. With only channel estimate available in COCO, the topology control problem in MANETs is formulated as a discrete stochastic optimization problem, which can be solved using a stochastic approximation approach. It considers both upper layer network capacity and physical layer relay selection. One of the main advantages of this iterative approach is that it can track the changing mobile environment to reconfigure the network topology dynamically. To the best of our knowledge, COCO is the first topology control scheme for MANETs with cooperative communications and noisy channel estimates. With this scheme, relay selection is extended to a network-wide behavior taking network capacity into account. The simulation results show that network capacity can be improved substantially in the proposed scheme. It is generally agreed in the literature that the reduced graph should be sparse to mitigate collisions and packet retransmissions, which leads to reduced power consumption and extended network lifetime. However, if too many edges are removed from the topology, data packets may traverse along an unacceptably long path. A fundamental issue has been pointed out in that the performance of multi-hop wireless networks will degrade sharply as the number of hops traversed increases. Another fact is that reducing interference can increase network capacity. However, interference is not sufficient. Therein the network capacity expression is used as its objective function, which takes link capacity and interference into consideration. The network connectivity and path length are regarded as constraints for the optimization problem. A discrete stochastic approximation approach is then adopted to solve the problem. This approach can be improved to track time-varying changes. Therefore, network topology is reconfigured in time. Cooperative communication is the principle of relay communication, where several sources act as a relays to one another. The COCO Topology scheme is proposed to improve the topology control problem in the Cooperative communication by considering link level issues in physical layers and upper layer issues such as network capacity. In cooperative communication, there are three types of transmission manners in its physical layer of MANETS, such as, direct transmission, multi hop transmission and cooperative transmission. There are two must conditions are taken into account in COCO scheme. First is network connectivity, which is basic ingredient in topology control and end-to-end network connectivity is mandatory via hop-by-hop model in objective function The solid lines denote traditional direct transmissions and multi hop transmissions. The dashed lines denote the links involved in cooperative communications. The network performance and the relay performance of the capacity optimized cooperative topology control scheme is depicted in fig 3 and fig 4 for simulation results. Fig 3: network performance graph Page 538

6 [5] K. Woradit et al., Outage Behavior of Selective Relaying Schemes, IEEE Trans. Wireless Commun., vol. 8,no. 8, 2009, pp [6] Y. Wei, F. R. Yu, and M. Song, Distributed Optimal Relay Selection in Wireless Cooperative Networks with Finite-State Markov Channels, IEEE Trans. Vehic. Tech., vol. 59, June 2010, pp [7] Q. Guan et al., Capacity-Optimized Topology Control for MANETs with Cooperative Communications, IEEE Trans. Wireless Commun., vol. 10, July 2011, pp Conclusions Fig 4: relay performance graph A dynamic traffic congestion method with co-operative control scheme called COCO considers both upper layer network capacity and physical layer relay selection in cooperative communications. Simulation results have shown that physical layer cooperative communications techniques have significant impacts on the network capacity, and the proposed topology control scheme can substantially improve the network capacity in MANETs with cooperative communications. By using this method we can reduce the delay of data delivery, reduces the endto-end delay and the number of route discovery requests, and balances the traffic Load. And that our proposed technique attains high delivery ratio and throughput with reduced delay when compared with the existing technique. To the best of our knowledge, COCO is the first topology control scheme for MANETs with cooperative communications and noisy channel estimates. [8] P. Santi, Topology Control in Wireless Ad Hoc and Sensor Networks, ACM Computing Surveys, vol. 37, no. 2, 2005, pp [9] T. Cover and A. E. Gamal, Capacity Theorems for the Relay Channel, IEEE Trans. Info. Theory, vol. 25, Sept. 1979, pp [10] Q. Guan et al., Impact of Topology Control on Capacity of Wireless Ad Hoc Networks, Proc. IEEE ICCS, Guangzhou, P. R. China, Nov [11] P. Gupta and P. Kumar, The Capacity of Wireless Networks IEEE Trans. Info. Theory, vol. 46, no. 2, 2000, pp References [1] Wei Lin et al., "Performance Analysis of Cooperative Networks with Random Decode-and-Forward Relaying," Proc. 10th IEEE Int. Conf. High Performance Computing and Communication., Dalian, Sept. 2008, pp [2]Quansheng guan, technology F. Richard yu,, Ottawa Shengming jiang, Victor c. M. Leung, inc. IEEE 2012 Transactions on Wireless Communications,Volume: 19 [3]J. Laneman, D. Tse, and G. Wornell, Cooperative Diversityin Wireless Networks: Efficient protocols and OutageBehavior, IEEE Trans. Info. Theory, vol. 50, no. 12, 2004, pp [4] P. H. J. Chong et al., Technologies in Multihop CellularNetwork, IEEE Commun. Mag., vol. 45, Sept. 2007,pp Page 539