### Table 1: Membrane topology predictions for the mouse tensin-related protein (gi|14787415) by several programs (available online at: http://www.expasy.org/tools/)

### Table 1. Results of topological searches for TIM barrels over the Atlas (3000 domains, left) and the the entire structural databank (15361 domains, right). Query names are explained in the main text. Nhits gives the total number of domains found containing the query topological pattern. Times exclude database load overhead.

### Table 1 compares the four algorithms in the CUCS domain. The SNMP algorithm runs the fastest by far, and, in this domain, discovers the entire network accurately. This is clearly the algorithm of choice. The other algorithms have nearly the same overhead, and all discover nearly the entire topology, except for the probing/traceroute algorithm, which takes significantly less time than the other two, but discovers only 65% of the hosts. Note that the time taken by any of the algorithms can be reduced by parallelizing it. Thus, for this domain, we recommend using SNMP, failing which, the second best choice is DNS zone transfer/ping broadcast, since it has less overhead than DNS zone transfer/traceroute, but still discovers nearly the entire topology.

"... In PAGE 10: ... However, we do feel that relying entirely on SNMP to discover network topology is a bad idea: though other algorithms take more time and more overhead, for a large topology, they are significantly more complete. CUCS Network Speed Overhead Completeness1 Accuracy Time (minutes) # pings # traces Normalized overhead Hosts Routers Subnets SNMP 11 5 0 5 482 (99%) 5 (100%) 7 (100%) 100% DNS zone transfer/ broadcast ping 148 1195 0 1195 485 (100%) 5 (100%) 6 (86%) 100% DNS zone transfer/ traceroute 128 840 480 1752 480 (99%) 5 (100%) 7 (100%) 99% Probing/traceroute 58 611 336 1249 317 (65%) 5 (100%) 7 (100%) 99% Table1 : Comparison of intra-domain discovery algorithms in the cs.... ..."

### Table 1: Parameters of topologies

"... In PAGE 3: ... GHITLE relies on a preferential attachment algorithm and assigns each edge in the topology a business rela- tionship (provider-to-customer or shared-cost). Table1 shows the number of ASes and the breakdown between Tier-1s/Transits/Stubs in each topology. The Large topology is an Internet-like topology with a small number of Tier-1s, a lot of stubs and a significant num- ber of transit domains in-between.... ..."

### Table 5: Drops in Topology 1

"... In PAGE 10: ...31 Table 4: Delay and Jitter of EF class in Topology 3 5.2 Discard on DiffServ Domain The Table5 shows the loss rate of the voice traffic on EF class and default traffic on BE... ..."

### Table 5: Drops in Topology 1

"... In PAGE 10: ...31 Table 4: Delay and Jitter of EF class in Topology 3 5.2 Discard on DiffServ Domain The Table5 shows the loss rate of the voice traffic on EF class and default traffic on BE... ..."

### Table 1: Simulated Topologies

1998

"... In PAGE 3: ... We simulated many net- works that di#0Ber in number of nodes, diameter, aver- age node degree, ratio of transit nodes to stub nodes, etc. Table1 summarizes the properties of the topolo- gies for whichwe present results. All of these graphs have 1500 nodes, of which 60 are transit nodes.... In PAGE 3: ... All of these graphs have 1500 nodes, of which 60 are transit nodes. In Table1 , SN, SD, TN and TD stand for #5Cstub node quot;, #5Cstub domain quot;, #5Ctransit node quot; and #5Ctransit domain quot;, respectively.Thus the Base graphs average four stub domains per transit node, and six stub nodes per stub... In PAGE 5: ...tencies and conserve bandwidth, but also reduce server load. 5 Simulation Results Wehave simulated cache performance for the topologies speci#0Ced in Table1 across a wide range of cache sizes, server distributions, and access parame- ters. We #0Crst summarize the general results, and then describe how the results are a#0Bected byvariations in the certain simulation parameters.... ..."

Cited by 56

### Table 1 Topological interpretation of the eight base relations of RCC-8. i( ) speci es the topological interior of a spatial region, the topological closure.

1999

"... In PAGE 3: ... Exactly one of these relations holds between any two spatial regions. These relations can be given a straightforward topological interpretation in terms of point-set topology (see Table1 ), which is almost the same as the semantics for the topological rela- tions given by Egenhofer [12] (though Egenhofer places stronger constraints on the domain of regions, e.g.... In PAGE 7: ...oth regions must not be empty, i.e., the complements of both X and Y are not equal to the universe. In the same way all topological constraints corresponding to the RCC-8 relations (see Table1 ) can be written as constraints of the form (m = U) and (e 6 = U), where m and e are set-theoretic expressions, denoted as model constraints and entailment constraints, respectively [2]. In the above example, X \ Y is the model constraint and X and Y are the entailment constraints.... ..."

Cited by 91

### Table 2: Rocketfuel ISP Topology Parameters

2006

"... In PAGE 3: ...48 Gb/sec and 2 ms delay Level 2 routers: 620 Mb/sec and 3 ms delay Level 3 routers: 155 Mb/sec and 50 ms delay Level 4 routers: 45 Mb/sec and 50 ms delay Level 5 routers and below: 1.55 Mb/sec and 50 ms delay Table2 outlines the details of the multiple AS topology. BGP routers within an AS are fully connected to form the iBGP domain.... ..."

Cited by 1

### Table 1. Topology configurations and characteristics

2003

"... In PAGE 8: ...onstant for the analysis, i.e. the number of bi-directional ports is 16, one traffic flow is associated with each input port, and the single SDRAM uses separate buses for read and write accesses. Topologies using eight processing elements are listed in Table1 together with static design characteristics. The first four topologies (I - IV) are also shown in Figure 1.... In PAGE 9: ...2. Results from analysis Given the topology configurations in Table1 analy- sis results from abstract benchmarking are given in Figure 5. The arrival rate of packets is fixed to the point where the modeled SDRAM reaches saturation in utiliza- tion.... In PAGE 10: ... The latter configurations more realistically model run-time jitter and arbitration effects as well as effects due to slightly unbalanced task graphs since a fully synchronous design is not feasible in our application domain. Note that the worst-case memory bounds given in Table1 for packet descriptors do not hold in the plain Round-Robin case in Figure 5. We would consider those designs to be unbal- anced for our application scenario since, for instance, in the pure pipeline case, up to 18 packets per flow could be in the network processor concurrently (as opposed to up to nine packets in Table 1).... In PAGE 10: ... Note that the worst-case memory bounds given in Table 1 for packet descriptors do not hold in the plain Round-Robin case in Figure 5. We would consider those designs to be unbal- anced for our application scenario since, for instance, in the pure pipeline case, up to 18 packets per flow could be in the network processor concurrently (as opposed to up to nine packets in Table1 ). Only under ideal assumptions the pure pipeline (con- figuration (1)) is able to match latency values with pool configurations (configuration (IV and V)) by over- provisioning the throughput of point-to-point (P-2-P) connections and thereby decreasing the transport delay.... ..."

Cited by 10