انسجام حافظه نهان یکپارچه مبتنی بر پروکسی و طرح مدیریت تحرک برای برنامه های خدمت گیرنده خدمت دهنده در سیستم های آی پی موبایل
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|28244||2009||14 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Parallel and Distributed Computing, Volume 69, Issue 6, June 2009, Pages 559–572
In this paper, we investigate a proxy-based integrated cache consistency and mobility management scheme for supporting client–server applications in Mobile IP systems with the objective to minimize the overall network traffic generated. Our cache consistency management scheme is based on a stateful strategy by which cache invalidation messages are asynchronously sent by the server to a mobile host (MH) whenever data objects cached at the MH have been updated. We use a per-user proxy to buffer invalidation messages to allow the MH to disconnect arbitrarily and to reduce the number of uplink requests when the MH is reconnected. Moreover, the user proxy takes the responsibility of mobility management to further reduce the network traffic. We investigate a design by which the MH’s proxy serves as a gateway foreign agent (GFA) as in the MIP Regional Registration protocol to keep track of the address of the MH in a region, with the proxy migrating with the MH when the MH crosses a regional area. We identify the optimal regional area size under which the overall network traffic cost, due to cache consistency management, mobility management, and query requests/replies, is minimized. The integrated cache consistency and mobility management scheme is demonstrated to outperform MIPv6, no-proxy and/or no-cache schemes, as well as a decoupled scheme that optimally but separately manages mobility and service activities in Mobile IPv6 environments.
With the advances of IP-based wireless networks, and the growth in the number of mobile hosts (MHs) carrying wireless devices, it is widely speculated that Mobile IPv6 (MIPv6)  will become prevalent in next generation all-IP networks to allow users to maintain service continuity while on the go . A major challenge is to maintain service continuity even if an MH disconnects and then reconnects at will. An MH may disconnect voluntarily simply to reduce the connection cost and/or to avoid power consumption. An MH may also disconnect involuntarily due to handoff or wireless link problems . Two major sources of traffic in MIPv6 systems are mobility management  and  and service management . Traditionally, mobility management has been considered separately from service management, as mobility management mainly deals with mobility handoff, location update and location search, while service management mainly deals with data delivery, and application servers can always query the underlying mobility management system to know the location of the MH. Over the years, many micro-mobility protocols extending MIP have been proposed with the goal to minimize mobility management overheads, including MIP Regional Registration  and  HMIPv6 , Cellular IP , IDMP , and HAWAII . In this paper we consider integrated mobility and cache management in the context of MIPv6 environments, considering cache management as a form of service management. In particular, we propose a Proxy-based Integrated Cache and Mobility Management (PICMM) scheme in MIPv6 networks to support mobile client–server database applications in which the MH queries the server for dynamic data. For example, an MH may query dynamic data such as stock prices, dynamic web pages, weather reports, or traffic information. To avoid sending a query to the server and receiving a reply through the expensive and often unreliable wireless communication network, an MH can cache data objects on the local storage and then answer queries for data that are up-to-date. Caching reduces the server access cost and improves the user-perceived response time . Cache management for such mobile client–server database applications essentially is service management since it deals with efficient data delivery issues. To process user queries correctly based on caching, an MH must ensure that its cached data are up-to-date. Our integrated cache consistency and mobility management scheme proposed in this paper is based on the stateful strategy  and  by which a cache invalidation message is asynchronously sent by the server to the MH, whenever there is an update to a data object cached at the MH. The MH uses invalidation reports received to determine the validity of its cache content before answering a query. If a query asks for a data object that has been invalidated, then a request is sent uplink to the sever to ask for a fresh copy of the data object accessed by the query before the query can be answered. Moreover, to support service continuity in cases the MH is disconnected and then reconnected again, we use a per-user proxy to buffer invalidation messages to allow the MH to disconnect arbitrarily and to reduce the number of uplink requests to check the cache status when the MH is reconnected. The generated network traffic cost associated with cache consistency management thus includes the cost of receiving and buffering invalidation messages at the proxy and the cost of forwarding them from the proxy to the MH, as well as the cost of sending requests to the server and receiving responses in case cached data objects are not up-to-date to answer queries. This cost is considered as part of the “service” management cost which we like to minimize. Here we note that a large body of research in the literature on cache invalidation for mobile environments is based on the stateless strategy , ,  and  by which the server has no knowledge of cache contents of MHs and will broadcast information on data objects that have been updated either periodically or asynchronously. However, the stateless strategy is not suitable for deployment over large wireless networks where users can roam from one network to another. Our approach utilizes a per-user proxy to keep invalidation reports of the MH’s cache. Whenever the proxy moves it informs the application servers of its whereabout, so invalidation reports can be sent directly from the application servers to the proxy. To minimize both the “service” and “mobility” network traffic costs, the proxy furthermore takes the responsibility of mobility management for integrated mobility and cache management. We investigate a design by which the MH’s proxy serves as the MH’s gateway foreign agent (GFA) as in the MIP Regional Registration micro-mobility protocol  and  to keep track of the address of the MH in a region. The proxy migrates with the MH when the MH crosses a regional area. We aim to identify the optimal regional area size under which the overall network traffic generated due to cache consistency management, mobility management, and query requests/replies is minimized. The idea can be extended to other micro-mobility management protocols such as HMIPv6 , Cellular IP , IDMP , and HAWAII , although in this paper we only consider the Regional Registration protocol. The basic idea of our approach is that we use a client-side proxy to support caching and mobility management in Mobile IPv6. The proxy has three functions: (1) working as a GFA as in regional registration to keep track of the MH’s location; (2) acting as a service proxy for services engaged by the MH; (3) allocating an extended cache space to store service context information including cache invalidation reports with timestamps for each MH. The proxy will receive invalidation reports from the server on behalf of the MH. If the MH is connected, the proxy will forward the invalidation report to the MH and then discard the invalidation report. If the MH is disconnected, the proxy will store invalidation reports in the proxy’s extended cache. Once the MH is reconnected, the MH will get the latest invalidation reports from the proxy. The benefit of the proxy-based scheme lies in the fact that the proxy as an integral part of mobility management knows the MH’s location information at all times and thus can efficiently manage the MH’s cached data and perform data delivery on behalf of server applications. The client-side proxy is created when the MH starts in MIPv6, acting as a GFA for the MH. We note that a cross-layer design such as MobileMAN  could be used for the implementation of the client-side proxy to run on the access router (AR). The cross-layer design has two components. The first component is at the network layer dealing with mobility management. The second component is at the application layer dealing with service management such as storing invalidation reports or service context information in its extended cache. We also note that Proxy Mobile IPv6 (PMIPv6)  and , a recent mobility management protocol for all-IP mobile networks, also proposes to run proxies on ARs for network-based mobility management. However only mobility management is addressed. The security issue of deploying proxies to run on ARs can be handled in a similar way as in PMIPv6. Finally, modern ARs are very powerful devices with ample memory space to run our proxies with extended cache functions. As the proxies of MHs (and hence their GFAs) will move from time to time as they cross regional areas, the load on ARs will be distributed according to the MHs’ mobility patterns. In the case of random mobility, the load would spread out to ARs evenly. The contributions of the paper are (1) the notion of integrated mobility and cache management to minimize the overall network traffic cost for supporting mobile client–server query applications in future MIPv6 systems; (2) identifying the optimal proxy setting including the region area size that will minimize the overall network traffic generated due to mobility and cache consistency management; and (3) demonstrating the benefit of integrated mobility and cache consistency management in MIPv6 compared with basic MIPv6, no-proxy and/or no-caching schemes, as well as a decoupled scheme under which mobility management and cache management are separate but optimally run. This work extends from our earlier preliminary work  to cover a comprehensive comparative analysis with existing cache invalidation and mobility management approaches in MIPv6 environments. The extension includes identifying conditions under which PICMM performs better than existing schemes, comparing PICMM with a decoupled scheme that optimally but separately manages cache and mobility activities, and analyzing the optimal regional area size as a function of a mobile user’s sleep pattern and the server database size. We note that the use of per-user proxies for integrated mobility and service management was discussed in . This work differs from  in that it specifically addresses cache consistency issues in query-based client–server mobile applications in MIPv6 environments. By integrating cache consistency management, query management and mobility management based on stateless cache invalidation protocols, it supports frequent MH disconnections and achieves cost minimization. The rest of the paper is organized as follows. Section 2 describes the system model and mobile database applications, Section 3 describes the proposed integrated cache and mobility management scheme. Section 4 develops a performance model to analyze the cost incurred under the proposed scheme. Section 5 identifies optimal conditions under which our proposed proxy-based algorithm performs the best in terms of network traffic minimization. We compare the performance of our proposed scheme with no-proxy and/or no-caching schemes, as well as with a decoupled scheme that separately manages mobility and service activities, to demonstrate the benefit of integrated cache and mobility management in Mobile IPv6 systems. Finally, Section 6 summarizes the paper and discusses the applicability.
نتیجه گیری انگلیسی
In this paper, we proposed a novel scheme for integrated mobility and service management in Mobile IPv6 systems. The core of the idea lies in leveraging a client-side proxy with duties for both cache and mobility management, and allowing intelligent MHs to determine the best service areas for service handoffs to minimize the network traffic cost. We devised a computational procedure to compute the optimal service area size that would minimize the overall network traffic cost, when given a set of parameters characterizing the MH’s mobility and service characteristics. We compared our scheme with several no-proxy and/or no-caching schemes and a decoupled scheme that optimally but separately manages mobility and service activities and concluded that our scheme outperforms these schemes in terms of the network traffic cost generated over a wide range of parameter values that typify practical mobile applications. To apply the analysis results presented in the paper, one can execute the computational procedure at static time to determine the optimal View the MathML sourceKopt over a possible range of parameter values for key parameters such as αα, ββ, nCTnCT, View the MathML sourceNdata, λq,iλq,i, σσ and μiμi. Then at runtime an MH can perform a simple look-up operation to determine View the MathML sourceKopt based on data collected at runtime. This allows each MH to dynamically determine the best service area size to minimize the overall network traffic generated. The performance gain is in the amount of network traffic communication cost saved per time unit per user, so the cost saving due to a proper selection of the best service area dynamically will have significant impacts since the cumulative effect for all mobile users over a long time period would be significant. Our approach requires cooperation between service providers and network operators because network operators own Mobile IP network routers which do not necessarily open to service providers. The current trend, however, is that network operators also wish to provide wireless services to end users, so the service provider and network operator are often integrated as well . More and more network operators also intend to offer their own proprietary services . An example is modern cell phone companies that own mobile networks and also provide bundled services to cell phone users . The tremendous performance gain in terms of reduced network traffic generated by our approach compared with a strict layering approach typified by the no-proxy scheme provides a strong justification of proxy-based integrated cache consistency and mobility service management for client–server applications in Mobile IP networks. In the future, we plan to investigate fault tolerance aspects of integrated service and mobility management for caching-based mobile applications. We also plan to investigate security issues including establishing and maintaining trust between the client and the proxy and preventing man-in-the-middle attacks.