S. Han and K.G. Shin. Fast restoration of real-time communication service from component failures in multihop networks. In Proc. of ACM SIGCOMM'97, pages 77--88, 1997.  Home/Search   Document Details and Download   Summary   ACM   TOC   Related Articles   Check

....maximized and then apply backup resource aggregation, similar to RAFT, along selected routes. The criteria for selecting primary and backup routes includes the aggregation information, among other factors, to determine relative importance of different routes. Dependable Real Time Protocol (DRTP) [22] provides a slightly different form of resource aggregation by overbooking backup reservations for flows along backup routes. Multiple backup routes can be set up to achive different levels of reliability. However, due to overbooking, there is a probability that some affected flows might be ....

S. Han and K.G. Shin. Fast restoration of real-time communication service from component failures in multihop networks. In Proc. of ACM SIGCOMM'97, pages 77--88, 1997.

....set has been discussed in general terms in [2, 7, 8] But we know of no previous theoretical or empirical investigation, and in particular no application to backup path establishment and restoration. Although there is a substantial literature on restoration in virtual circuit networks (see e.g. [3, 9, 13, 12]) our approach exploits speci c characteristics of MPLS. In considering the application of our restoration schemes to other technologies such as WDM and ATM, the trade o between the cost of setting up and tearing down virtual circuits versus the cost of path concatenation has to be evaluated. ....

S. Han and K. G. Shin. Fast restoration of real-time communication service from component failures in multi-hop networks. In Proc. of ACM SIGCOMM Symposium, September 1997.

....noncoincident component failures when more than four or five subchannels are used. In many networks however, it may be hard to find so many disjoint routes between two hosts. A global and proactive scheme that is similar to RAFT, called Backup Channel Protocol (BCP) has been recently suggested [9]. BCP supports different RoS classes, some of them for multiple failures, and it uses several backup channels (secondary paths) for every FT flow. The method used for resource aggregation is based on a heuristic algorithm. A difference with RAFT is that BCP is a statistical scheme, in the sense ....

S. Han and K. G. Shin, "Fast restoration of real-time communication service from component failures in multi-hop networks, " in Proceedings SIGCOMM Symposium, September 1997.

....may lead to service interruption too long to be acceptable, and in some cases the renewed reservation involved may not succeed at all because of resource constraint. The alternative solution to on demand recovery is preplanned failure restoration. For unicast, a technique called back up channel[3] can be used. In this technique, a backup (standby) route is setup in addition to the primary working route. The backup channel can be activated when a link or This work was supported in part by grants from NASA and NSF. node failure is detected on the primary route. The problem is more ....

S. Han and K. G. Shin. Fast restoration of real-time communication services from component failures in multi-hop networks. in Proc. ACM SIGCOMM'97 , September 1997

....user application. Otherwise, the performance of the corresponding real time application will suffer noticeably. Several schemes have been proposed to address this problem in the context of a network with heavy weight routers. Specifically, they make use of dispersity routing [1] or backup channels [7]. However, these schemes also reduce network utilization and increase the processing requirements of the routers. In summary, a heavy weight router can easily become a performance bottleneck due to a lack of processing power. The problem is compounded by the fact that integrated services will ....

Seungjae Han and Kang G. Shin. Fast restoration of real-time communication service from component failures in multi-hop networks. In Proceedings ACM SIGCOMM '97, pages 77--88, Cannes, France, September 1997.

....the end user. Otherwise, the performance of the corresponding real time application will suffer noticeably. Several schemes have been proposed to address this problem in the context of a network with heavy weight routers. Specifically, they make use of dispersity routing [1] or backup channels [7]. However, these schemes also reduce network utilization and increase the processing requirements of the routers. In summary, a heavy weight router can easily become a performance bottleneck due to a lack of processing power. The problem is compounded by the fact that integrated services will ....

Seungjae Han and Kang G. Shin. Fast restoration of real-time communication service from component failures in multi-hop networks. In Proceedings ACM SIGCOMM '97, pages 77--88, Cannes, France, September 1997.

....noncoincident component failures when more than four or five subchannels are used. In many networks however, it may be hard to find so many disjoint routes between two hosts. A global and proactive scheme that is similar to RAFT, called Backup Channel Protocol (BCP) has been recently suggested [9]. BCP supports different RoS classes, some of them for multiple failures, and it uses several backup channels (secondary paths) for every FT flow. The method used for resource aggregation is based on a heuristic algorithm. A difference with RAFT is that BCP is a statistical scheme, in the sense ....

S. Han and K. G. Shin, "Fast restoration of real-time communication service from component failures in multi-hop networks, " in Proceedings SIGCOMM Symposium, September 1997.

.... songkuk, dqao, skodase, kgshn eecs. mich. edu Abstract Dependability of Service (DOS) has become an important requirement for real time applications, such as remote medical services, business critical network meetings, and command 4 control applications. The Dependable Real Time Protocol (DRTP) [5, 6, 7] in which each dependable real time connection is realized with one primary and one or more backup channels, has been shown to be an effective way of providing DoS. How to route both primary and backup channels for each dependable real time connection is of vital importance to the success of ....

....[12] additional resources are reserved in the vicinity of each real time channel, and the failed components are de toured by using the reserved resources. In [13] the SFI method was extended to combat special patterns of multiple failures in a hexagonal mesh network. Spare resource allocation [4, 5, 6, 7] is another ap proach that employs failure detection and recovery for fault tolerance. Additional (called spare) resources are reserved a priori solely for the purpose of speedy service recovery from possible network failures. The key advantage of proactive schemes is that the latency and ....

[Article contains additional citation context not shown here]

S. Han and K. G. Shin. Fast restoration of real- time communication service fi'om component failures in multihop networks. In Proceedings of ACM SIGCOMM '97, pages 77-88, 1997.

....the minimum level of QoS, s he may receive only bare bone service even when there are plenty of resources available. One way of utilizing resources efficiently is to specify flexible (elastic) QoS requirements that can be adapted to the availability of network resources. Recently, Han and Shin [1] proposed to allocate one primary channel and one or more backup channels to each dependable real time (DR ) connection. One drawback of this scheme is the severe reduction in number of DR connections that can be accommodated, due mainly to the need for reserving resources for backups. This is ....

....made to provide timeliness guarantees necessary for the above mentioned applications. See the survey paper by Aras et al. 2] for a detailed account of many of the existing real time communication schemes. Most of the real time communication schemes known to date share three common properties [1]: QoS contracted, connection oriented, and reservation based. Before actually transferring any message, a contract must be established between a client and the network. In the contract, the client specifies his input traffic generation behavior and required QoS. The network then computes the ....

[Article contains additional citation context not shown here]

S. J. Han and K. G. Shin, "Fast restoration of real-time communication service from component failures," in Proc. of ACM SIGCOMM, pp. 77--89, 1997.

....fsongkuk,dqiao,skodase,kgshing eecs.umich.edu Abstract Dependability of Service (DoS) has become an important requirement for real time applications, such as remote medical services, business critical network meetings, and command control applications. The Dependable Real Time Protocol (DRTP) [5, 6, 7] in which each dependable real time connection is realized with one primary and one or more backup channels, has been shown to be an e ective way of providing DoS. How to route both primary and backup channels for each dependable real time connection is of vital importance to the success of ....

....[12] additional resources are reserved in the vicinity of each real time channel, and the failed components are detoured by using the reserved resources. In [13] the SFI method was extended to combat special patterns of multiple failures in a hexagonal mesh network. Spare resource allocation [4, 5, 6, 7] is another approach that employs failure detection and recovery for fault tolerance. Additional (called spare) resources are reserved a priori solely for the purpose of speedy service recovery from possible network failures. The key advantage of proactive schemes is that the latency and ....

[Article contains additional citation context not shown here]

S. Han and K. G. Shin. Fast restoration of realtime communication service from component failures in multihop networks. In Proceedings of ACM SIGCOMM '97, pages 77-88, 1997.

....resource exhaustion. Thus, some backups cannot be activated because other activated backups have already taken all spare resources. In such a case, a multiplexing failure is said to occur. We account for the probability that a backup suffers a multiplexing failure in the calculation of P r . See [18] for a detailed account of calculating P r . Thus far, we explained the procedure of allocating spare resources, assuming that backup paths are given. How to route backups greatly influences the resource efficiency, i.e. the amount of spare resources to achieve a certain P r . Our study shows ....

S. Han and K. G. Shin, "Fast restoration of real-time communication service from component failures in multi-hop networks," in Proc. ACM SIGCOMM, pp. 77--88, 1997. 20

....two major problems: i) backup establishment before a failure and (ii) failure handling upon occurrence of a failure. In this paper, we will focus only on the first problem without addressing the issues related to the second problem such as failure detection, reporting, and backup activation. See [13, 14] for solutions to these issues. Two key issues in backup establishment are backup route selection and spareresource calculation. In contrast to the existing endto end schemes for non dynamic environments, we separate these two issues from each other, so that backup channels can be routed ....

....link, we can calculate the probability denoted by S(B i ; B j ) of simultaneous activation of two backups, B i and B j , whose primaries are M i and M j , respectively; S(B i ; B j ) sc(M i ; M j ) Delta , where sc(M i ; M j ) is the number of components shared between M i and M j . See [14] for the derivation. Based on this probability, the set of backups to be multiplexed together is determined for each backup on a link. B i and B j are multiplexed if S(B i ; B j ) is smaller than a certain threshold , called the multiplexing degree. So, the rule to decide resource sharing in the ....

S. Han and K. G. Shin, "Fast restoration of realtime communication service from component failures in multi-hop networks," in Proc. ACM SIGCOMM, pp. 77--88, 1997.

No context found.

S. Han and K.G. Shin. Fast restoration of real-time communication service from component failures in multihop networks. In Proc. of ACM SIGCOMM'97, pages 77--88, 1997.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC