Cooperative Caching Strategies for MANETs and IMANETs Atul Rao, Prashant Kumar and Naveen Chauhan Department of Computer Science and Engineering National Institute of Technology, Hamirpur, India E-mail: raonithmr@gmail.com, prashantkumar32@gmail.com, naveenchauhan.nith@gmail.com Abstract Caching in mobile computing environment is a capable technique that can improve data access performance and reduce the heavy communication between client and server. Cooperative caching allows sharing and coordination of cached data between mobile hosts. In this paper all the caching schemes for wired, Ad hoc and internet mobile Ad hoc network are discussed and compared in terms of cache resolution and cache replacement policy. Keywords: Cooperative caching, Internet-based mobile adhoc network, cache Introduction The term Mobile Adhoc Networks (MANETs) refers to a multi hop packet based wireless network composed of a set of mobile nodes that can communicate and move at the same time, without using any kind of fixed wired infrastructure. MANETs are actually self-organizing and adaptive networks that can be formed and deformed on-the-fly without the need of any centralized administration. The main two characteristics of MANETs are mobility and multi hop communication. MANETs usage areas: Military scenarios Sensor networks Rescue operations Students on campus Free Internet connection sharing Conferences With the recent advent in wireless technologies and mobile devices, wireless networks have become a ubiquitous communication infrastructure. In addition, growing interest in accessing the wired network or Internet has fueled the development of mobile wireless networks, which can be used in many realistic applications. It is envisaged that in the near future, users will be able to access the Internet services and information anytime and anywhere. Internet-based mobile Ad hoc networks (IMANETs) [3, 8] are an emerging technique that combines a wired network (e.g. Internet) and a mobile Ad hoc network (MANETs) for developing a ubiquitous communication infrastructure. Internet-based Mobile Ad hoc Networking is a technology that supports selforganizing, mobile networking infrastructures, and is one which appears well-suited for use in future commercial and military applications. Thus, to put the MANET technology into the context of real life, we consider an Internet-based MANET (IMANETs), which is an evolving communication infrastructure that combines the wired Internet and wireless mobile Ad hoc networks. IMANET is getting more attention and is applied to realistic Internet applications because of its flexible accessibility and information availability. The followings are some of the applicable uses for IMANETs: Case1: During events such as Olympic Games the demand from users to access the Internet and communicate among themselves are very high. While a fixed infrastructure may be in place, it is challenging to accommodate all the users due to limited wireless bandwidth. With an IMANET, users can either access the required information directly or indirectly (through relays). Case2: In a battle field or emergency site, one MT may be connected to the Internet by a satellite and serve as a proxy for other MTs. The accessed information and services can be shared by the other MTs via local Ad hoc communication. Figure 1: System Model for IMANET. Caching is an important technique to improve the performance of wireless or wired network. Aim of various caching schemes is to reduce the web traffic, congestion control, bandwidth management and data accessibility in internet and in IMANETs. The rest of paper is organized as follows. Section II describes the different cooperative schemes in wired and adhoc networks. Section III describes the comparison of various caching scheme and section IV gives [Page No. 89]
5 th IEEE International Conference on Advanced Computing & Communication Technologies [ICACCT-2011] ISBN 81-87885-03-3 overview of IMANETs and available caching schemes. Section V concludes the paper. Cooperative Caching A number of data replication schemes and caching schemes have been proposed in order to facilitate data access in mobile Ad hoc networks (MANETs). Data replication studies the issue of allocating replicas of data items to meet access demands. These techniques normally require a priori knowledge of the network topology. Caching schemes however do not facilitate data access based on the knowledge of distributed data items. In Simple Cache the requested data item has always been cached by the requester node. The node uses the cached copy in order to serve subsequent requests when they arrive. The requester node has to get the data from the data center in case of cache miss. However increasing the hop distance between the requester node and caching node will increase the response time for the request. In the research area of mobile Ad hoc networks have been developed a number of caching protocols. Why Cooperative Caching? Caching of frequently accessed data in multi-hop adhoc environment is a potential technique that can improve the data access performance and availability. Cooperative caching allows the sharing and co-ordination of cached data among different clients and groups. Due to mobility and constraints on resources like bandwidth, computational resources and limited battery power in mobile adhoc networks, some cooperative cache management schemes need to be designed. Cooperative Caching in Wired Network: Squirrel Cache Squirrel is a decentralized peer-to-peer web cache [15]. It is scalable, self-organizing and resilient to peer failures. Without the need for additional hardware or administration it is able to achieve the functionality and the performance of a traditional centralized web cache. It is proposed to run in a corporate LAN type environment, located e.g. in a building, a single geographical region. Squirrel is build up on Pastry, an object location and routing protocol for large scale peer-to-peer systems, which provides the mentioned features. The goals and the motivation for web caching are decline of load on external web servers, corporate routers and of course external traffic (traffic between the corporate LAN and the internet) which is expensive, especially for large organizations. Squirrel is a possibility to achieve these goals without the use of a centralized web cache or even clusters of web caches. Squirrel uses Pastry as a location and routing protocol. When a client1 requests an object it first sends a request to the Squirrel proxy running on the client s machine. If the object is unreachable then the proxy forwards the request directly to the origin Web server. Otherwise it checks the local cache. If a fresh copy of the object is not found in this cache, then Squirrel tries to locate one on some other node. To do so, it uses the distributed hash-table and the routing functionalities provided by Pastry. First, the URL of the object is hashed to give a 128-bit Object Id from a circular list. Then the routing procedure of Pastry forwards the request to the node with the Node Id (assigned randomly by Pastry to a participating node) numerically closest to the Object Id. This node then becomes the home node for this object. Squirrel then proposes two schemes named: Home-store Directory schemes The home-store scheme is the one to be used in reality and approaches the performance of a centralized web cache with infinite storage while using e.g. 100 MB cache size per node. Disadvantages of Squirrel Cache: In decentralized caching churn arises from continued and rapid arrival and failure (or departure) of a large number of participants in a peer-to-peer system. This will increase host loads and block a large fraction of normal insert and lookup operations in the system. There is no cooperation between the mobile hosts or peers. Summary Cache In the summary cache scheme [5], each proxy stores a summary of its directory of cached document in every other proxy. When a user request misses in the local cache, the local proxy checks the stored summaries to see if the requested document might be stored in other proxies. If it appears so, the proxy sends out requests to the relevant proxies to fetch the document. Otherwise, the proxy sends the request directly to the Web server. The key to the scalability of the scheme is that summaries do not have to be up-to-date or accurate. A summary does not have to be updated every time the cache directory is changed; rather, the update can occur upon regular time intervals or when a certain percentage of the cached documents are not reflected in the summary. The sharing of caches among Web proxies is an important technique to reduce Web traffic and alleviate network bottlenecks A summary only needs to be inclusive (that is, depicting a superset of the documents stored in the cache) to avoid affecting the total cache hit ratio. That is, two kinds of errors are that may occur: False misses: The document requested is cached at some other proxy but its summary does not reflect the fact. In this case, a remote cache hit is not taken advantage of, and the total hit ratio within the collection of caches is reduced. False hits: The document requested is not cached at some other proxy but its summary indicates that it is. The proxy will send a query message to the other proxy, only to be notified that the document is not cached there. In this case,a query message is wasted. The errors affect the total cache hit ratio or the inter proxy traffic, but do not affect the correctness of the caching scheme. For example, a false hit does not result in the wrong document being served. In general we strive for low false misses, because false misses increase traffic to the Internet and the goal of cache sharing is to reduce traffic to the Internet. Two factors limit the scalability of summary cache: the network overhead (the interproxy traffic), and the memory required storing the summaries (for performance reasons, the summaries should be stored in DRAM, not on disk). The network overhead is determined by the frequency of summary [Page No. 90]
Cooperative Caching Strategies for MANETs and IMANETs updates and by the number of false hits and remote hits. The memory requirement is determined by the size of individual summaries and the number of cooperating proxies. Since the memory grows linearly with the number of proxies, it is important to keep the individual summaries small. In summary cache, cache sharing under finite cache sizes, a number of schemes are explored for the evaluation of summery cache. Different scheme are as under: No Cache Sharing. Simple Cache Sharing Single-Copy Cache Sharing Global Cache Two questions are answered whether simple cache sharing significantly reduces traffic to Web servers, and whether the more tightly coordinating schemes lead to a significantly higher hit ratio. Here the hit ratio includes both local hits and remote hits. Local hits are those requested documents found in the proxy s cache; remote hits are those documents found in the neighboring proxies cache. Both kinds of hit avoid traffic to web servers. Impact of update delays is also studied in the summary cache. Cooperative Proxy Caching: Mainly a cache is defined as a fast, temporary store for commonly used items. High speed memory units on microprocessor chips cache data from main memory; main memory units cache sections of disk files; and local disks cache documents from the network file server. Cooperating proxy caches [4] are groups of HTTP proxy servers that share cached objects. Client caching on the Web succeeds at the browser level (L1) and the local proxy level (L2) for the same reasons that memory and disk caches succeed; specifically Local disk and network transfers are faster than remote network transfers from the Web server Web documents are often re-requested by the same user or by other users on the same local network. At the next level of Web caching (L3), proxy servers cooperate to share their cached documents. If a cache miss occurs at the local proxy server, that proxy may forward the request to a remote proxy instead of to the origin Web server. However, the viability of cooperative proxy caching is unclear because, unlike browser and local proxy caches, a remote proxy is not inherently faster than the origin Web server. Retrieving objects from either a remote cache or a remote server involves wide-area network transfers, and there is no guarantee that the remote cache will have faster server hardware, higher bandwidth, or better routes to the client. When cache and server have similar resources and communication costs, cooperative proxy caching actually increases response time due to cache overhead. In cache hierarchies, L3 proxy caches are separate physical sites such as national caches. In cache meshes, the same site serves as a local L2 proxy cache to its clients and as a remote L3 proxy cache to its cooperative peers. Cooperating caches have three separate functions: Discovery Dissemination Delivery of cache objects Figure 3: Cooperating proxy caches organized as a mesh a hierarchy Discovery refers to how a proxy locates cached objects. Dissemination is the process of selecting and storing objects in the caches. Delivery defines how objects make their way from the Web server or a remote proxy cache to the requesting proxy. Cooperative Caching in MANETS Push and Pull Based Approach These two are the basic cache sharing techniques in MANET [1, 4]. With push-based cache sharing, when a node acquires and caches a new data, it actively advertises the caching event to the nodes in the neighborhood. Mobile nodes in the vicinity will record the caching information upon receiving such an advertisement and use it to direct subsequent requests for the same item. This scheme enhances the usefulness of the cached contents. The cost we have to pay is the communication overhead for the advertisement; an advertisement is useless if no demands for the cached item arise in the neighborhood. In the push-based scheme, the caching information known to a node may become obsolete due to node mobility or cache replacement. The pull-based approach may overcome this problem. Withpull based cache sharing, when a mobile node wants toaccess a data item that is not cached locally it will broadcast a request to the nodes in its vicinity. A nearby node that has cached the data will send a copy of the data to the request originator (a pull operation) unlike pushing; pulling allows the node to utilize the latest cache contents. Zone Based Cooperative Caching Scheme In Zone Cooperative caching scheme[1, 12], one hop neighbor of a mobile client form a cooperative cache zone since the cost for communication with them is low both in terms of energy consumption and message exchange. Zone cooperative caching scheme is for data retrieval in mobile Ad hoc networks. The design rationale of ZC caching is that it is considered advantageous for a client to share cache with its neighbors lying in the zone (i.e., mobile hosts that are accessible in one hop) Mobile hosts belonging to the zone of a given host then form a cooperative cache system for this host. In ZC caching Cache discovery is a problem. During Cache Discovery Process When a dataa request is initiated in an MH, it first looks for the data item in its own cache. If there is a local cache miss, the MH checks if the data item is cached in other MHs within its home zone. When an MH receives the [Page No. 91]
5 th IEEE International Conference on Advanced Computing & Communication Technologies [ICACCT-2011] ISBN 81-87885-03-3 request and has the data item in its local cache (i.e., a zone cache hit), it will send a reply to the requester to acknowledge that it has the data item. In case of a zone cache miss, the request is forwarded to the neighbor along the routing path. If the data item is not found on the zones along the routing path (i.e., a remote cache miss), the request finally reaches the data source and the data source sends back the requested data. that the cache space can be used to accommodate more distinct data items. Authors categories cached data copies based on whether they are already available in the neighborhood or not. A data copy is primary if it is not available within the neighborhood. Otherwise the data copy is secondary. Cache misses the main factor for differentiating between primary and secondary data.cache miss cost is directly proportional to the travel distance of a data request. The inter-category and intra-category rules are used to decide caching priorities of primary and secondary data. Figure 2: Cache Discovery Process in ZC caching. COOP-A Cooperative Caching service in MANET It is a novel cooperative caching scheme for on-demand data access applications in MANETs. The objective is to improve data availability and access efficiency by getting together local resources of mobile nodes. The cooperation of caching nodes is twofold. First, a caching node can answer the data requests from other nodes. Second, a caching node stores the data not only on behalf of its own needs, but also based on other nodes needs. COOP addresses two basic problems for cooperative caching in MANETs: Cache resolution Cache Resolution addresses how to restore a data request with minimal cost of time, energy and bandwidth. COOP S cache resolution is a cocktail scheme, which consists of three basic schemes: Adaptive Flooding: It calculates proper flooding range based on the cost to fetch the requested data. Limited flooding is used for cache resolution, not only because it has potential to discover the closest cache around the requested, but also because flooding can serve as an announcement in the neighbourhood and effectively segment the whole network into the cluster, with in which they can share and manage cached contents. Profile Based Resolution: It maintains a historical profile of previously received data request and determines a closer data sources for user s request based on the profile. Roadside Resolution: If a data request cannot get resolved using these two schemes the data. Request is forwarded to the original data source. The Roadside Resolution is used to resolve the data request along the forwarding path. Cache management In COOP cache management scheme we have to decide which data item to keep in a node s local cache.the aim is to increase the cache hit ratio, which largely depends on the capacity of the cache. To enhance the capacity of cooperative caches, COOP tries to reduce duplicated caching within short distance neighborhood, such Limitations of Coop To improve data availability and access performance, COOP addresses two basic problems of cooperative caching. For cache resolution, COOP uses the cocktail approach which consists of two basic schemes: hop-by-hop resolution and zone-based resolution. By using this approach, COOP discovers data sources which have less communication cost. For cache management, COOP uses the inter- and intra- duplications between the category rules to minimize caching nodes within a same cooperation zone and this improves the overall capacity of cooperated caches. The disadvantage of the scheme is that flooding incurs high discovery overhead and it does not consider factors such as size and consistency during replacement. Semantic Caching: Semantic caching [6] is used to manage the location dependent data in mobile computing environment. Location dependent data (LDD) is the data whose value is determined by the location to which it is related. Examples include local yellow pages, traffic reports, weather information, and maps and so on. A location dependent query is a query that is processed on location dependent data, and whose result depends on the location criteria explicitly or implicitly specified. The idea of semantic caching is that the mobile client maintains both the semantic descriptions and associated answers of previous queries in the cache. If a new query is totally answerable from the cache, no communication with the server is necessary; if it can only be partially answered, the original query is trimmed and the trimmed part is sent to the server to be processed. Semantic caching is by nature an ideal cache scheme for location dependent applications due to the following reasons: Semantic caching is built on the semantic locality among queries, which just fits the LDD applications where much semantic rather than temporal or spatial locality is exhibited. Continuous LDD queries can be incrementally processed by semantic caching. With each successive request, a much smaller trimmed LDD query is processed at the server side and only the differences are transmitted over thewireless link. Semantic caching makes cache management more flexible. The cache can be managed based on temporal or location information. Semantic caching also facilitates disconnections even though the data at current location cannot be obtained, the mobile user might still be able to learn the information for other neighbour locations from the local cache. Semantic caching has been widely used in centralized systems, client-server environment, OLAF systems mobile [Page No. 92]
Cooperative Caching Strategies for MANETs and IMANETs computing and heterogeneous systems. The cache is composed of a set of items attached with the related Semantic descriptions, which are called semantic regions in, semantic segments in and so on. While logically the cache is always organized using an index which maintains the semantic as well as physical storage information for every cached item, there are various ways to physically store the data. The commonly used cache replacement strategies are built on temporal locality, such as LRU, MRU and CLOCK. Semantic cache organizes data by semantic metrics; it makes cache management more flexible. Other Caching Techniques Yin and Cao [13] propose three schemes: CachePath, CacheData and HybridCache. In CacheData, intermediate nodes cache the data to serve future requests instead of fetching data from the data center. In CachePath, mobile nodes cache the data path and use it to redirect future requests to the nearby node which has the data instead of the faraway data center. To further improve the performance [10, 11] a hybrid approach (HybridCache), is given by taking advantage of CacheData and CachePath while avoiding their weaknesses. A cooperative caching scheme, called CoCa, was proposed by C.Y.Chow [17]. The CoCa framework facilitate mobile nodes to share their cached contents with each other in order to reduce the number of server requests and the number of access misses in a single hop wireless mobile network. The authors extended CoCa with a group-based cooperative caching scheme, called GroCoCa [18]. According to GroCoCa, the decision of whether a data item should be cached depends on two factors: Access affinity on the data items and The mobility of each node. Papadopouli suggested the 7DS architecture [16] in which a couple of protocols are defined to share and disseminate information among users. It operates either on a prefetches mode, based on the information and user s future needs or on an on-demand mode, which searches for data items in a singlehop multicast basis. Hassan Artail proposed a caching scheme COACS [11] which stands for Cooperative and Adaptive Caching System. The idea is to create a cooperative caching system that minimizes delay and maximizes the likelihood of finding data that is cached in the Ad hoc network, all without inducing excessively large traffic at the nodes. COACS is a distributed caching scheme that relies on the indexing of cached queries to make the task of locating the desired database data more efficient and reliable. Comparison of Cooperative Caching Schemes Table 1 Schemes Cache Resolution Cache Replacement Summary Directory-Based LRU Squirrel Hash-Based LRU COOP Cock Tail Scheme Inter-category and intra category Rules Yin CacheData, LRU CachePath, HybridCache N.Chand ZC LUV Semantic No Specification Temporal locality such as LRU, MRU Cooperative Directory-Based No Specification Proxy Caching S.Lin Aggregate Two factors: 1.Distance 2.Access frequency Caching In IMANETs To combine MANETs with the wired network or Internet, we consider the Imanet infrastructure and come across the problem of information search and access under this environment. In IMANETs the routing protocols such as Destination Sequenced Distance Vector (DSDV), Dynamic Source Routing (DSR), Ad hoc On Demand Distance Vector (AODV), Zone Routing Protocol (ZRP), and Temporally Ordered Routing Algorithm (TORA) are explored. They are similar as in MANETs. These protocols based on the model that a sender MT knows the location of receiver MT based on the route information, which is accumulated and analyzed by a route discovery or route maintenance algorithms. A Route discovery operation captures the current network topology and related information. It has to be executed when an MT needs to transmit a data item. To avoid repetitive route discovery, the MTs can cache the old route information. Why Caching in IMANETs? IMANETs [3, 8] has several constraints. First, all the MTs cannot access the Internet. Second, due to mobility, a set of MTs can be separated from the rest of the MTs and get disconnected from the Internet. Finally, an MT requiring multi-hop relay to access the Internet may incur longer access latency than those which have direct access to the Internet. To address these constraints, an aggregate caching mechanism for IMANETs is proposed. The basic idea is that by storing data items in the local cache of the MTs, members of the IMANETs can efficiently access the required information. Thus, the aggregated local cache of the MTs can be considered as a unified large cache for the Imanet. The proposed aggregate cache can alleviate the constraints of IMANETs discussed above. When an MT is blocked from direct access to the Internet, it may access the requested data items from the local cache of nearby MTs or via relays. If an MT is isolated from the Internet, it can search other reachable MTs for the requested data item. Finally, if an MT is located further from the Internet, it may request the data items from [Page No. 93]
5 th IEEE International Conference on Advanced Computing & Communication Technologies [ICACCT-2011] ISBN 81-87885-03-3 other close by MTs to reduce access latency. Here, two issues are addressed for implementation of an aggregate caching mechanism in IMANETs: Efficient search: An efficient information search algorithm is fundamental for locating the requested data in IMANETs. Cache management: To reduce the average access latency as well as enhance the data accessibility, efficient cache admission control and replacement policies are critical. The cache admission control policy determines whether a data item should be cached, while the cache replacement policy intelligently selects a victim data item to be replaced when a cache becomes full. Information search in IMANETs is different from the search engine based approach used in the wired Internet. An MT needs to broadcast its request to the possible data sources (including the Internet and other MTs within the Imanet) in order to retrieve the requested data efficiently. An aggregate cache for IMANETs is proposed to address the issues of accessibility and latency. Conclusion In this paper we have discussed cache sharing issues, schemes related to mobile Ad hoc network environment and give analysis of some popular cooperative caching schemes. These caching schemes are useful in MANETs environment. Here we present how these schemes are advantageous in order to find a data item in a MANETs by using less resources (e.g. network bandwidth, energy etc.) and improves the performance(data availability and latency time). We also discussed the limitations of these techniques. As the cooperative caching is a useful technique to improve the data availability in the MANETs so these analyses will be helpful for the future research. References [1] N.Chand, R.C.Joshi and Manoj Misra, Cooperative caching in mobile Ad hoc networks based on data utility,mobile Information Systems pp. 19 37,2007. [2] Yu Du and S.Gupta, COOP-A Cooperative Caching service in MANETs, Proceedings of the IEEE ICAS/ICNS (2005), pp.58-63, 2006. [3] S. Lim, W. Lee, G. Cao and C. Das, A Novel Caching Scheme for Internet based Mobile Ad hoc Network Performance, Ad hoc Networks, vol.4, no.2 pp.225-239,2006. [4] S.G.Dykes and K.A.Robbines, A Viability of Cooperative Proxy Caching, IEEE INFOCOM,pp 1205-1214,2001. [5] L.Fan, P. Cao, J. Almeida, and A. Z. Broder, Summary Cache: A Scalable Wide Area Web Cache Sharing Protocol, IEEE/ACM Transaction on Networking,vol 8,no3,June 2000. [6] Qun Renet, Using Semantic Caching to Manage Location Dependent Data in Mobile Computing, IEEE MOBOCOM, pp.210-220,2000. [7] T.Hara Data Replication for Improving Data Accessibility in Ad hoc Networks, IEEE Transaction on Mobile Computing,vol 5 no11,pp 1515-1532,Nov 2006. [8] D. Barbara and T. Imielinksi Sleepers and Workaholics: Caching Strategies for Mobile Environments. In Proc. ACM SIGMOD, pp1 12, 1994. [9] S. Lim, W. Lee, G. Cao, and C. R. Das. A Novel Caching Scheme for Internet based Mobile Ad hoc Networks.In Proc. 12th International Conference on Computer Communications and Networks (ICCCN), pp 38 43, 2003. [10] F.Sailhan and V. Issarny. Cooperative Caching in Ad hoc Networks,In Proc. 4th International Conference on Mobile Data Management (MDM), pp 13 28, 2003. [11] H.Artail, H.Safa, "COACS: A Cooperative and Adaptive Caching System for MANETs," IEEE Transactions On Mobile Computing, Vol. 7, no. 8, pp 961-977 Aug 2008. [12] N.Chand, R.C.Joshi and M.Misra. Cooperative caching strategy in mobile Ad hoc networks based on clusters, Wireless Personal Communications, 43(1):41 63, 2007. [13] L. Yin and G. Cao Supporting cooperative caching in adhoc networks,ieee Transactions on Mobile Computing, 5(1):77 89, 2006. [14] B.Tang, "Benefit-Based Data Caching in Ad hoc Networks" IEEE Transaction on Mobile Computing, vol 7, no3, pp289-304, March2008. [15] SitaramIyer, Antony Rowstron,and Peter Druschel. Squirrel: a decentralized peer-to-peer web cache,in PODC 02:Proceedings of the twenty-first annual symposium on Principles of distributed computing, pages 213 222. ACM Press, 2002. [16] M.Papadopouli and H. Schulzrinne Effects of Power Conservation,WirelessConverage and Cooperation on Data Dissemination among Mobile Devices,In Proceedings of MobiHoc, pages 117 127, 2001. [17] C.Y. Chow, H.V. Leong and A Chan, Peer-to-Peer Cooperative Caching in Mobile Environments Proceedings of 24 th International Conference on Distributed Computing Systems Workshop pp528-533,2004 [18] Chi-Yin Chow, Hong Va Leong "GroCoca: Groupbased Peer-to-Peer Cooperative Caching in Mobile Environment", IEEE Journal On Selected Areas In Communications, Vol. 25, no. 1, January 2007. [Page No. 94]