D. Dunn et. al., "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration," IEEE J. on Sel. Areas in Comm., Vol. 12, No. 1, pp. 88-99.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....to implement at the cost of requiring more spare capacity than failure dependent path restoration. In this paper, we use link disjoint backup paths in failure independent path restoration. A problem that arises in the failure independent path restoration scheme is the existence of trap topology [17], 4] In a trap topology, the working path may block all the possible linkdisjoint backup paths although the network topology is twoconnected. For example, when the traffic flow between node 13 and node 15 in Network 6 in Fig. 8 has working path routed via nodes 1 and 22, this path does not have ....

....hop paths, they are the shortest among all working paths protected by link disjoint backup paths. Although this method introduces longer working paths and disturbs traffic on the working paths already been routed, it is still a practical method thanks to the rare occurrence of the trap topology [17] and we use this method to deal with trap topologies. B. SCA Algorithms Previous research on spare capacity allocation of mesh type networks adopts the problem context above and uses either mathematical programming techniques or heuristics to determine the spare capacity assignment as well as ....

D. Anthony Dunn, Wayne D. Grover, and Mike H. MacGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration," IEEE Journal on Selected Areas of Communications,vol.2, no. 1, pp. 88--99, 1 1994.

....networks, which is suitable for the design of large scale self healing networks. Finding this heuristic algorithm is an important contribution of the paper. The problem of capacity allocation and flow assignment in self healing mesh type networks was studied before, e.g. in [9] 10] 11] 12] [19] for link restoration and in [13] 14] 16] for global reconfiguration, path restoration. Unlike prevoius work, here we mainly con sider failure oriented, path restoration. A comparison of link and path restorations was given in [17]in the context of STM networks. The remainder of the paper is ....

D. Dunn, W. Grover, M. McGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration", IEEE J. Select. Areas Uommun, vol. 12, 1994, pp. 88-89.

....of actual networks: the IP backbone network shown in Figure 2a, and the Kyoto University network shown in Figure 2b. The IP backbone network consists of 12 nodes and 42 links, all of the same capacity. The Kyoto network has 78 nodes and 1122 links, also of identical capacity. As shown in [14], the k shortest paths method for generating disjoint paths is fast and reasonably close to the optimum solution. For each source destination pair k, we ran Dijkstra s shortest path algorithm, and then iterated the procedure on the network with the newly discovered path pruned. The set of paths ....

D.A. Dunn, W.D. Grover and M.H. MacGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration", in IEEE Journal on Selected Areas in Communications, vol. 12, no. 1, January 1994.

....mH metric (each link has weight one) Fig. 3b shows that, once the shortest path has been allocated to the working lightpath between nodes A and D, there is no chance to setup the protection path because the network is disconnected. For this particular property this has been called trap network [24]. Finally, Fig. 3b shows that a pair of link disjoint paths actually exists between A and D, but the two step search can not find it. A technique able to overcome the two steps limitations has been proposed by Bhandari [23] It is called one step search, B C D E F working A B C D E F ....

D. A. Dunn, W. D. Grover, and M. H. Gregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration," IEEE Journal on Selected Areas in Communications, pp. 88--89, 1994.

....approaches is less (statistically significant decrease by 12 and 42 respectively) and rapidly decreases as the group size increases. The explanation for this difference in feasibility is the existance of multipoint traps . The term trap was first introduced in the point to point context in [4]. A trap is a topology where a corresponding set of backup routes is not available due to the disjointness constraint although disjoint working and backup routes may be available if selected simultaneously and not sequentially (working then disjoint backup) Although not always feasible on an ....

D. Dunn et. al., "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration", IEEE J. on Sel. Areas in Comm., Vol. 12, No. 1, pp. 88-99.

....to implement at the cost of requiring more spare capacity than failure dependent path restoration. In this paper, we use link disjoint backup paths in failure independent path restoration. A problem that arises in the failure independent path restoration scheme is the existence of trap topology [17], 4] In a trap topology, the working path may block all the possible linkdisjoint backup paths although the network topology is twoconnected. For example, when the traffic flow between node 13 and node 15 in Network 6 in Fig. 8 has working path routed via nodes 1 and 22, this path does not have ....

....hop paths, they are the shortest among all working paths protected by link disjoint backup paths. Although this method introduces longer working paths and disturbs traffic on the working paths already been routed, it is still a practical method thanks to the rare occurrence of the trap topology [17] and we use this method to deal with trap topologies. B. SCA Algorithms Previous research on spare capacity allocation of mesh type networks adopts the problem context above and uses either mathematical programming techniques or heuristics to determine the spare capacity allocation as well as ....

D. Anthony Dunn, Wayne D. Grover, and Mike H. MacGregor, "Com- parison of k-shortest paths and maximum flow routing for network facility restoration," IEEE Journal on Selected Areas of Communications, vol. 2, no. 1, pp. 88--99, 1 1994.

....path has at least one node disjoint backup path. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Fig. 1. Network 6 with 23 nodes and 33 links, a path is from node 8 to 11 A similar problem, called trap topology, has been discussed by Dunn, Grover and MacGregor [20]. In a trap topology, a working path may block all the possible linkdisjoint backup paths although the network topology is twolink connected. The problem to find a working path with linkdisjoint backup path can be solved by using augmenting path algorithm from [17] as we discussed in [16] ....

D. A. Dunn, W. D. Grover, and M. H. MacGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration, " IEEE Journal on Selected Areas of Communications, vol. 2, no. 1, pp. 88--99, 1 1994.

....While the hop limit bound is a factor in preventing available disjoint backup circuits, the trap has become the dominant factor for point to multipoint restoration after being a relatively rare event for point topoint backup restoration. Point to point traps are discussed in more detail in [7]. Up to this point we have focused on comparing the availability and cost of backup groups consisting exclusively of point to multipoint backup circuits (multicast) versus backup groups consisting exclusively of point topoint backup circuits (unicast) There is another alternative. Hybrid backup ....

D. Dunn, W. Grover, and M. MacGregor, "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration," IEEE Journal on Selected Areas in Communications, Vol. 12, No. 1, January 1994, pp. 88-99.

....a considerably large amount of time. Moreover, if the switching tables have to be generated for all the snapshots using this approach, then the time complexity involved would be unpredictably high. Thus, even for the multicommodity flow approach, approximations based on the shortest path solutions [3,5] that have lower time complexity will have to be considered. Therefore, the problem can be viewed as shortest path approximations. The transition between snapshots, as mentioned before, may result in jitters. Therefore, it is not sufficient if we calculate only the shortest path between every pair ....

D.A. Dunn, Wayne D.Grover and M.H. MacGregor, "Comparison of k Shortest Paths and Maximum Flow Routing for Network Facility Restoration", IEEE Journal on Selected Areas in Communication, Vol.12, pp.88-99, 1994.

....adaptive and potentially exploit spare capacity anywhere in the network. It has been shown that for span (also called link ) restorable networks, the restoration path set is closely modeled by the set of k shortest paths through the spare capacity graph between the end nodes of the failure [3]. Although giving algorithmic definition to the routing behaviour, the exact structure and lengths of the routes that arise under this model still depend on the distribution of spare capacity. And for optical networks it can be important to keep restoration paths as short as possible for optical ....

D. A. Dunn, W. D. Grover, M. H.MacGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration", IEEE J. Selected Areas Com., vol. 12, no. 1, pp. 88-99 (1994)

No context found.

D. A. Dunn, W. D. Grover and M. H. MacGregor, `Comparison of k-shortest paths and maximum flow routing for network facility restoration', IEEE JSAC Special Issue on Integrity of Public Telecommunications Networks, 12, (1), 88--99 (1994).

....paths infeasible. The basic issue can be appreciated from the previously studied problem of the trap topology in span restoration wherein a k shortest paths (ksp) routing process can sometimes have fewer members (PNE 1) in the presence of the trap than the corresponding max flow pathset [35]. This arises because a shortest path, greedily chosen early in building up a path set, makes one or more subsequent paths, which are present in the max flow solution, infeasible under the simpler ksp procedure. By avoiding the shortest path, a greater total number of paths are feasible. Although ....

....because a shortest path, greedily chosen early in building up a path set, makes one or more subsequent paths, which are present in the max flow solution, infeasible under the simpler ksp procedure. By avoiding the shortest path, a greater total number of paths are feasible. Although the work in [35] found that the effect of the trap topology is not a significant effect for span restoration, such phenomena are much more important in general in path restoration because the path subsets being formed extend end to end across the network on many O D pairs simultaneously. In addition, the total ....

D. A. Dunn, W. D. Grover, M. H. MacGregor, "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration," IEEE Journal on Selected Areas in Communications, Vol. 12, No. 1, pp. 88-99, January 1994.

....adaptive and potentially exploit spare capacity anywhere in the network. It has been shown that for span (also called link ) restorable networks, the restoration path set is closely modeled by the set of k shortest paths through the spare capacity graph between the end nodes of the failure [3]. Although giving algorithmic definition to the routing behaviour, the exact structure and lengths of the routes that arise under this model still depend on the distribution of spare capacity. And for optical networks it can be important to keep restoration paths as short as possible for optical ....

D. A. Dunn, W. D. Grover, M. H.MacGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration", IEEE J. Selected Areas Com., vol. 12, no. 1, pp. 88-99 (1994)

....path forming operations to supplement the useful paths available in pre configured patterns, using any other spare capacity remaining. In this regard, the k shortest paths (KSP) restorability of a network is the restorability when a k successively shortest link disjoint paths algorithm (such as in [33, 34]) calculates a restoration pathset for each failed span within the complete set of spare links present, without regard to any pre configuration plan. An unallocated link is a spare link that is not part of any pre configured pattern. The pre configured restorability is the average fraction (over ....

....the number of times the spare links on a span can contribute to the restoration of other spans. 4.1. Spanning Tree Pre configuration Heuristic An iterative procedure is used to develop a pre configuration plan based on spanning trees: at the start of each iteration, the KSP restoration pathset [33,34] is obtained for each span, as a fail Network Restorability Design Using Preconfiguration TR 1999 05 TRLabs Network Systems 11 31 ure span. The span weights are determined from the KSP phase by counting the number of times that each span is part of the restoration pathset of all other span ....

D. A. Dunn, W. D. Grover, M. H. MacGregor, "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration," IEEE Journal on Selected Areas in Communications, Vol. 12, No. 1, pp. 88-99, January 1994.

No context found.

D. Dunn et. al., "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration," IEEE J. on Sel. Areas in Comm., Vol. 12, No. 1, pp. 88-99.

No context found.

D. Dunn, W. Grover, and M. MacGregor, "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration," IEEE Journal on Selected Areas in Communications, Vol. 12, No. 1, January 1994, pp. 88-99.

No context found.

D. Dunn et. al., "Comparison of k-Shortest Paths and Maximum Flow Routing for Network Facility Restoration," IEEE J. on Sel. Areas in Comm., Vol. 12, No. 1, pp. 88-99.

No context found.

D.A. Dunn, W.D. Grover, M.H. MacGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration", IEEE Journal of Selected Areas in Communications, vol. 12, no. 1, pp. 88-99, January 1994.

No context found.

D.A. Dunn, W.D. Grover and M.H. MacGregor, "Comparison of k-shortest paths and maximum flow routing for network facility restoration", in IEEE Journal on Selected Areas in Communications, vol. 12, no. 1, January 1994.

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