EFFICIENT P2P FILE SHARING IN MOBILE AD HOC NETWORKS(MANET) USING FILE REPLICATION

Transcription

1 EFFICIENT P2P FILE SHARING IN MOBILE AD HOC NETWORKS(MANET) USING FILE REPLICATION Mrs. Sri Devi.S 1, S.Narasimman 2, R.Ashwinmuthiya 3, C.GopalaKrishnana 4, H.Prabhakaran 5 1 Assistant professor, Department Of Computer Science and Engineering, School of Engineering, Vels University,Chennai,India 2,3,4,5 UG Student, Department Of Computer Science and Engineering,School of Engineering, Vels University,Chennai,India Abstract - File sharing applications are recently widely used in MANET. Because of limited communication range and node mobility, the file availability for sharing becomes tedious task. Peer To Peer over MANET is the most commonly used method for file sharing in MANET. In P2P file sharing system, File replication technology is mostly used to improve the file query efficiency. Most of the current file replication methods replicate files in all nodes or two end points on a client-server query path[1].thus, these methods either have low effectiveness or have highly over headed cost. File replication in server side enhances replica hit rate, hence, look up efficiency but produces overloaded nodes and cannot significantly reduce query path length[2].an intuitive method to alleviate this problem is to create file replicas in the network. Actually, a node that has a higher meeting frequency with others provides higher availability to its files. This becomes even more evident in sparsely distributed MANETs, in which nodes meet disruptively [3]. In this paper, we introduce a new concept of resource for file replication, which considers both node storage and meeting frequency. And also this paper presents an efficient file replication algorithm that achieves high query efficiency and high replica utilization at a significantly low cost. Simulation results demonstrate the efficiency and effectiveness of distributed replication protocol in comparison with other approaches in both static and dynamic environments. It dramatically reduces the overhead of file replication and yields significant improvements on the efficiency and effectiveness of file replication in terms of query efficiency, replica hit rate, and overloaded nodes reduction. Keywords: P2p,File Sharing,File Replication. I. INTRODUCTION Mobile ad hoc network (MANETs) is a mobile, dynamically and self-organizing wireless networks in absence of a fixed infrastructure, which is usually used in emergency environment such as disaster recovery, military battlefields etc. Recent years, the personal mobile devices such as smart phone, PDA, ipad are increasing rapidly. And with the rapid development in the field of wireless communication technologies, e.g., WLAN, WiMax, 3G, 3.5G and emerging 4G, mobile users now want to have access to applications and data whenever and wherever through mobile handheld devices. The mobile user would like to generate personal content, store useful information, search content from Internet or other mobile devices and share content with their friends all in handheld mobile platforms However, the limited communication capacity of the base station is not easy to satisfy the big requirement. MANETs consisted by the mobile devices which are sufficient with wireless technologies such as WIFI or Bluetooth, are effective and important supplement to the application between mobile users. The mobile devices bring new application scenario to All Rights Reserved 424

2 Normally peer-to-peer technology used for the distribution and sharing of digital documents and computer files in a network. This process is called peer-to-peer file sharing.p2p file sharing allows users to access media files such as texts, music, movies, and games using a specialized P2P software program that searches for other connected computers on a P2P network and locates the desired content. The nodes or peers of such networks are end-user computer systems that are interconnected via the Internet. In centralized system the scalability is hard to achieve but in decentralized system there is no such bottleneck. P2P over MANETs is the most popular used pattern of file sharing within MANETs. P2P system is initiated in the middle of 1990s. P2P system is deployed on the Internet which is formed by creating an overlay network. P2P overlay consists of upper-layer connections among nodes, or peers, which are independent of the underlay or substrate network, abstract peers view of the connections that make up the network. Here the interaction among peers generally appears at network and application level. P2P file sharing network is very widely used in recent Internet. And within MANETs, P2P file sharing approach is also widely adopted. File replication replicate a hot file to some other nodes, distribute the file query load among a number of nodes and avoid hot spots to improve file query efficiency. File sharing applications in mobile ad hoc networks (MANETs) have attracted more and more attention in recent years. With the increasing popularity of mobile devices, e.g., smart phones and laptops, we envision the future of MANETs consisted of these mobile devices. By MANETs, we refer to both normal MANETs and disconnected MANETs, also known as delay tolerant networks (DTNs). The former has a relatively dense node distribution in an area while the latter has sparsely distributed nodes that meet each other opportunistically. On the other side, the emerging of mobile file sharing applications (e.g., Qik [1] and Flixwagon [2]) motivates the investigation on the peer-to-peer (P2P) file sharing over such MANETs. The local P2P file sharing model provides three advantages. First, it enables file sharing when no base stations are available (e.g., in rural areas). Second, with the P2P architecture, the bottleneck on overloaded servers in current client server based file sharing systems can be avoided. Third, it exploits otherwise wasted peer to peer communication opportunities among mobile nodes. As a result, nodes can freely and unobtrusively access and share files in the distributed MANET environment, which can possibly support interesting applications. For example, mobile nodes can share files based on users proximity [3] in the same building or in a local community. Tourists can share their travel experiences or emergency information with other tourists through digital devices directly even when no base station is available in remote areas. Drivers can share road information through the vehicle-to-vehicle communication. Fig1.shows the system architecture However, the distinctive properties of MANETs, i.e., node mobility, limited communication range and resource, have rendered many difficulties in realizing such a P2Pfile sharing system. For example, file searching turns out to be difficult since nodes in MANETs move around freely and can exchange information only when they are within the communication range. Broadcasting can quickly discover files, but it leads to the broadcast storm problem [4] with high energy consumption. Probabilistic routing and file discovery protocols [5], [6], [7] avoid broadcasting by forwarding a query to a node with higher probability of meeting the destination. But the opportunistic encountering of nodes in MANETs makes file searching and retrieval non-deterministic. File replication is an effective way to enhance file availability and reduce file querying delay. It creates replicas for a file to improve its probability of being encountered by requests. Unfortunately, it is impractical and inefficient to enable every node to hold the replicas of all files in the system considering limited node All Rights Reserved 425

3 Fig.1 System Architecture Also, file querying delay is always a main concern in a file sharing system. Users often desire to receive their requested files quickly no matter whether the files are popular or not. Thus, a critical issue is raised for further investigation: how to allocate the limited resource in the network to different files for replication so that the overall average file querying delay is minimized? In spite of efforts, current file replication protocols lack a rule to allocate limited resources to files for replica creation in order to achieve the minimum average querying delay, i.e., global search efficiency optimization under limited resources. They simply consider storage as the resource for replicas, but neglect that a node s frequency to meet other nodes (meeting ability in short) also influences the availability of its files. Files in a node with a higher meeting ability have higher availability. II RELATED WORK 1) File Sharing in Normal MANETs The topic of file replication for efficient file sharing applications in MANETs has been studied recently. An individual node or a group of nodes decide the list of files to replicate according to file querying frequency. Hara and Madria proposed three file replication protocols: static access frequency (SAF), dynamic access frequency and neighborhood (DAFN), and dynamic connectivity based grouping (DCG). In SAF, each node replicates its frequently queried files until its available storage is used up. SAF may lead to many duplicate replicas among neighboring nodes when they have the same interested files. DAFN eliminates duplicate replicas among neighbors. DCG further reduces duplicate replicas in a group of nodes with frequent connections by creating replicas for files in the descending order of their group based querying frequencies. Though DAFN and DCG enable replicas to be shared among neighbors, neighboring nodes may separate from each other due to node mobility. Also, they incur high traffic load in identifying duplicates or managing groups. 2) File Sharing in Disconnected MANETs/DTNs Huang et al. discussed how to cache files in servers to realize the optimal file availability to mobile users in WiFi based wireless networks based on node mobility pattern, AP topology, and file popularity. However, the file servers in this paper are fixed nodes connecting to APs, while we consider a more general P2P scenario, in which all mobile nodes are both file servers and clients. Pitkanen and Ott [17] proposed the DTN storage module to leverage the DTN store-carry-and-forward paradigm and make DTN nodes keep a copy of a message for a longer period of time required by forwarding. Gao et al. [14] proposed a cooperative caching method in DTNs by copying each file to the node in All Rights Reserved 426

4 network central location, which is frequently visited by other nodes. When the central node is full, less popular replicas are moved to its neighbor nodes. However, central nodes may be frequently changed, leading to frequent file transfers and high overhead. QCR [15] leverages caching for multimedia content dissemination in opportunistic networks. It considers data retrieval delay and the probability that users will require the same content based on previously experiences to decide the caching policy. SEDUM [20] also uses replication to create redundant messages in routing for DTNs, thereby enhancing routing success rate. III PROPOSED SOLUTION Decentralized File Replication Method Instead of creating replicas on all nodes or two ends on a client to server path, Effective Decentralized File Replication(EAD)method chooses query traffic hubs (i.e., query traffic conjunction nodes) as replica nodes to ensure high replica hit rate. It achieves a comparable query efficiency to Path but creates much less replicas. It also produces higher hit rate than the client side, and dramatically reduce the lookup path length and avoids overloading replica nodes in ServerSide. EAD takes the full advantage of file replicas by dynamically choosing replica nodes based on file query rate. First the EAD aims in minimizing the replicas and to achieve high file query efficiency. More replicas leads to higher query efficiency and vice versa. How to reduce replicas without compromise query efficiency. Rather than statically replicating a file along a query path, EAD replicates a file in nodes with high query traffic of file, thus reducing more replicas while it ensures high hit rate and comparable query efficiency. Second, rather than depending on a file owner to determine replica creation and deletion in a centralized manner, the EAD aims to conduct the operations in a decentralized manner without compromising replica utilization. i.e, the EAD enables nodes themselves to determine whether and when they should create replicas or delete replicas based on their actual experienced query traffic. If a node has too high query traffic of a file, it requests to be a replica node of the file or if a replica node receive too few queries of a replica, it removes the replica. Splitting a large file into small pieces can increase the service capacity of a large file rapidly. Replicating file location hint along a query path can also improve file query efficiency V. PERFORMANCE EVALUTION We use ns2 to evaluate the performance of the Efficient and Adaptive decentralized File replication.the simulation area is a grid of 1500 meters by 320 meters. We divide this grid into 10 sub-grids of 150 meters by 320 meters and place a super-node at the center of each sub-grid. The random waypoint movement model is used in which 50 nodes move at a speed uniformly distributed between 0-20m/s. We assume the wireless bandwidth is 2 Mbps and the transmission range is 250m. The run time of the experiment is kept as 500 second. Three metrics are measured: average delay, message overhead and packet delivery ratio (PDR). Average delay is the time elapsed between the query is sent and the data is transmitted back to the requester averaged upon all the queries. Message overhead includes all the query and response messages of locating both data and optimal route. Packet delivery ratio is defined as the percentage of the data queries which receives the requested data. More work and detail plan needs to be done in this respect. There are 100 data items randomly distributed among all 50 nodes. In the P2P model, each peer can play the role of both a client and server. In our client query model, each node sends out queries with the interval of query generate time. The query generate time is uniformly distributed within (0; Tquery], where Tquery is the mean query generate time. Once a All Rights Reserved 427

5 sends out a query, it doesn t need to wait until the data returned back before launching another data query. Fig2: Server and client nodes are created Fig3- server responds to the client and leaf nodes A. Results and Performance Analysis Three key performance metrics are evaluated: Fig4:client get replica file from the server Packet delivery fraction: The ratio of the data packets delivered to the destination to those generated by the All Rights Reserved 428

6 Average end-to-end delay of data packets: This includes all possible delays caused by buffering during route discovery latency, queuing at the interface queue, retransmission delays at the MAC, propagation and transfer times. Throughput: It is the ratio of the total amount of data that reaches a receiver from a sender to the time it takes for the receiver to get the last packet. Throughput= Received packets/end Time Fig 1.Packet Delivery Ratio. Fig2. Drop Ratio All Rights Reserved 429

7 VI. CONCLUSION This paper proposes an EAD file replication algorithm that chooses query traffic hubs and frequent requesters as replica nodes to guarantee high utilization of replicas and high query efficiency. Unlike current methods in which file servers keep track of replicas, EAD creates and deletes file replicas by dynamically adapting to no uniform and time varying file popularity and node interest in a decentralized manner based on the experienced query traffic. It leads to higher scalability and ensures high replica utilization. File replication is an effective method to deal with the problem of server overload by distributing load over replica nodes. It helps to achieve high query efficiency by reducing server response latency and lookup path length (i.e., the number of hops in a lookup path).ead relies on exponential moving average technique to reasonably measure file query rate for replica management. REFERENCES 1. S. Saroiu, P. Gummadi, and S. Gribble, "A Measurement Study of Peer-to-peer File Sharing Systems," Proc. Conf. Multimedia Computing and Networking (MMCN), V. Gopalakrishnan et al., "Adaptive Replication in Peer-to-Peer Systems," Proc. Int'l Conf. Distributed Computing Systems (ICDCS), Gnutella, M. Roussopoulos and M. Baker, "CUP: Controlled Update Propagation in Peer to Peer Networks," Proc. USENIX, L. Yin and G. Cao, "DUP: Dynamic-Tree Based Update Propagation in Peer-to-Peer Networks," Proc. Int'l Conf. Data Eng. (ICDE), A. Ghodsi, L. Alima, and S. Haridi, "Symmetric Replication for Structured Peer-toPeer Systems," Proc. Int'l Workshop Databases, Information Systems and Peer-to-Peer Computing, p. 12, T. Pitoura, N. Ntarmos, and P. Triantafillou, "Replication, Load Balancing and Efficient Range Query Processing in DHTs," Proc. Conf. Extending Database Technology(EDBT), J. Kangasharju, K.W. Ross, and D.A. Turner, "Optimizing File Availability in Peer-to-Peer Content Distribution," Proc. IEEEINFOCOM, J. Kangasharju, K.W. Ross, and D.A. Turner, "Adaptive Content Management in Structured P2P Communities," Proc. Int'l Conf.Scalable Information Systems (Infoscale), S. Tewari and L. Kleinrock, "Proportional Replication in Peer-to-Peer Networks," Proc. IEEEINFOCOM, L. Massoulie and M. Vojnovic, "Coupon ReplicationSystems,"Proc.ACM SIGMETRICS, S. Tewari and L. Kleinrock, "On Fairness, Optimal Download Performance and Proportional Replication in Peerto-Peer Networks," Proc. IFIP Networking Conf., S. Tewari and L. Kleinrock, "Analysis of Search and Replication in Unstructured Peerto-Peer Networks,"Proc.ACMSIGMETRICS, D. Rubenstein and S. Sahu, "Can Unstructured P2P Protocols Survive Flas Crowds?" IEEE/ACMTrans.Networking,vol.3,no.3,pp , H. Shen, "EAD: An Efficient and Adaptive Decentralized File Replication Algorithm in P2P File Sharing Systems," Proc. Int'l Conf. Peer-to-Peer Computing (P2P), S. Ioannidis, L. Massoulie, and A. Chaintreau, Distributed Caching over Heterogeneous Mobile Networks, Proc. ACM Int l Conf.Measurement and Modeling of Computer Systems (SIGMETRICS), M.J. Pitkanen and J. Ott, Redundancy and Distributed Caching in Mobile DTNs, Proc. ACM/IEEE Int l Workshop Mobility in the Evolving Internet Architecture (MobiArch), X. Zhuo, Q. Li, W. Gao, G. Cao, and Y. Dai, Contact Duration Aware Data Replication in Delay Tolerant Networks, Proc. IEEE 19th Int l Conf. Network Protocols (ICNP), X. Zhuo, Q. Li, G. Cao, Y. Dai, B.K. Szymanski, and T.L. Porta, Social-Based Cooperative Caching in DTNs: A Contact Duration Aware Approach, Proc. IEEE Eighth Int l Conf. Mobile Adhoc and Sensor Systems (MASS), Z. Li and H. Shen, SEDUM: Exploiting Social Networks in Utility-Based Distributed Routing for DTNs, IEEE Trans. Computers, vol. 62, no. 1, pp , Jan V. Gianuzzi, Data Replication Effectiveness in Mobile Ad-Hoc Networks, Proc. ACM First Int l Workshop Performance Evaluation of Wireless Ad Hoc, Sensor, and Ubiquitous Networks (PE-WASUN), pp , S. Chessa and P. Maestrini, Dependable and Secure Data Storage and Retrieval in Mobile Wireless Networks, Proc. Int l Conf. Dependable Systems and Networks (DSN), All Rights Reserved 430